正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Congrats on the TypeScript 7 release! 🎉 TS7 vs TS6 racing Hono RPC type inference - tsc is 3-4x faster!! (Not 10x: that's parse + multi-core over many files - Hono's inference is one deep sequential type)

Yusuke Wada

25,491 subscribers

25,247 次观看 • 1 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

🚀 A new era in Bitcoin mining is here 🚀 On Dec 12, 2024, Type One Mining solo mined block #874168 using DATUM—not just providing hash power but fully designing the block template. This marks a shift where miners reclaim creative control of the blockchain. Our miners call it a euphoric experience to design, solve, and imprint their name on a block. Congrats Type One Marc & the entire team at Type One Mining! 🎉

🚀 A new era in Bitcoin mining is here 🚀 On Dec 12, 2024, Type One Mining solo mined block #874168 using DATUM—not just providing hash power but fully designing the block template. This marks a shift where miners reclaim creative control of the blockchain. Our miners call it a euphoric experience to design, solve, and imprint their name on a block. Congrats Type One Marc & the entire team at Type One Mining! 🎉

OCEAN

15,779 次观看 • 1 年前

"Inference, if you look at it as a market, will be much, much bigger than cloud computing was pre-ChatGPT." Lightspeed’s Bucky Moore says inference is an underrated investment category in AI, and expects the market to break up into large, specialized platforms for each modality: "The GPU supply crunch that we're seeing right now is largely, as Dylan Patel has said on the show before, due to the fact that not only these consumer products, but also B2B products like Claude Code and Codex are just really taking off and creating insane demand for inference." "We're talking hundreds of billions in spend every year. And if that's true, I think there will be very, very large inference platforms built in each modality." "So there will be an inference platform for real-time video models, there will be an inference platform for open-source and custom language models, there will be an inference platform built specifically for long-running agents." "So I think we're just going to see that industry, which today looks like one industry, break up into many because of how big it is and how much room for specialization there is."

"Inference, if you look at it as a market, will be much, much bigger than cloud computing was pre-ChatGPT." Lightspeed’s Bucky Moore says inference is an underrated investment category in AI, and expects the market to break up into large, specialized platforms for each modality: "The GPU supply crunch that we're seeing right now is largely, as Dylan Patel has said on the show before, due to the fact that not only these consumer products, but also B2B products like Claude Code and Codex are just really taking off and creating insane demand for inference." "We're talking hundreds of billions in spend every year. And if that's true, I think there will be very, very large inference platforms built in each modality." "So there will be an inference platform for real-time video models, there will be an inference platform for open-source and custom language models, there will be an inference platform built specifically for long-running agents." "So I think we're just going to see that industry, which today looks like one industry, break up into many because of how big it is and how much room for specialization there is."

TBPN

33,278 次观看 • 5 个月前

PRESIDENT TRUMP ready to release all Epstein files. Trump warns that people's names just associated with the Epstein files without nefarious activity could hurt them and their reputation, especially throughout the media. Maybe they were always ready to release the files. They just had to get everyone else on board and to demand the release of the files. Remember, layers, optics, and timing. Military advantage perspective. No emotions, logical decisions only. This is not a game, yet it is played like one. This is known as the art of war. If you're still wondering, Trump is NOT in the Epstein files in ANY type of nefarious way whatsoever. As you will soon find out. Godspeed patriots.

PRESIDENT TRUMP ready to release all Epstein files. Trump warns that people's names just associated with the Epstein files without nefarious activity could hurt them and their reputation, especially throughout the media. Maybe they were always ready to release the files. They just had to get everyone else on board and to demand the release of the files. Remember, layers, optics, and timing. Military advantage perspective. No emotions, logical decisions only. This is not a game, yet it is played like one. This is known as the art of war. If you're still wondering, Trump is NOT in the Epstein files in ANY type of nefarious way whatsoever. As you will soon find out. Godspeed patriots.

The SCIF

366,590 次观看 • 1 年前

Scott Jennings on the Epstein Files: Well, look, they've released how many documents tonight? 200,000 or more? 400? So, a good faith effort is being made to release files tonight… Regarding Bill Clinton, I'm sorry to tell you, the reason that Bill Clinton's picture is in the Epstein files is because this has always been a story about Bill Clinton and not Donald Trump. We should call it the Clinton files, not the Epstein files.” “That's what we're learning tonight. This is the Clinton files. This is the blowing up in the face of the Democrats over this. It's Bill Clinton — this is a story about Democrats and Bill Clinton, and it's always been that way.”

Scott Jennings on the Epstein Files: Well, look, they've released how many documents tonight? 200,000 or more? 400? So, a good faith effort is being made to release files tonight… Regarding Bill Clinton, I'm sorry to tell you, the reason that Bill Clinton's picture is in the Epstein files is because this has always been a story about Bill Clinton and not Donald Trump. We should call it the Clinton files, not the Epstein files.” “That's what we're learning tonight. This is the Clinton files. This is the blowing up in the face of the Democrats over this. It's Bill Clinton — this is a story about Democrats and Bill Clinton, and it's always been that way.”

RedWave Press

753,600 次观看 • 8 个月前

$AMD MI400 series & Sam Altman OpenAI Sam Alman praised MI400 series GPUs and MI450, and stated “The memory architecture is great for inference [and] for training as well.” and a "Game changer" for frontier AI Models Ok we have Elon Musk xAI , Mark Zuckerberg $META, Sam Alman OpenAI , Tareq Amin( CEO of Saudi), Satya Nadella $MSFT, Andy Jassy $AMZN, Larry Ellison $ORCL and many CEOs from large to small provided positive feedbacks on MI400 series rack solution, which is expected to deliver up to 10x inference performance over piror generations. And short sellers/bears are still beating down on $AMD H2 2025 and 2026? AMD will hit new ATH before year end. Not Financial Advice!

$AMD MI400 series & Sam Altman OpenAI Sam Alman praised MI400 series GPUs and MI450, and stated “The memory architecture is great for inference [and] for training as well.” and a "Game changer" for frontier AI Models Ok we have Elon Musk xAI , Mark Zuckerberg $META, Sam Alman OpenAI , Tareq Amin( CEO of Saudi), Satya Nadella $MSFT, Andy Jassy $AMZN, Larry Ellison $ORCL and many CEOs from large to small provided positive feedbacks on MI400 series rack solution, which is expected to deliver up to 10x inference performance over piror generations. And short sellers/bears are still beating down on $AMD H2 2025 and 2026? AMD will hit new ATH before year end. Not Financial Advice!

Mike

89,052 次观看 • 11 个月前

$AMD $5 Trillion MC Is Inevitable Long Term👑 This thread will focus more on Inference! 2026 EPYC "Venice" $TSM 2nm to save Large GW Scale Inference by 40% more than Prior Turin gen. Context: EPYC Turin achieves ~$0.001 per million tokens for batch inference vs $0.02-$0.12/ million tokens as I wrote the thread below. Venice is going to lower cost down to $0.0005-$0.0006/Million Tokens. OpenAI spent roughly $20B on Inference and Training, where 80-90% of that was for Inference per Analysts. AKA Renting Compute is Expensive AF! In this thread, I want to focus on why most analysts and investors are underestimating the role EPYC "Venice" and future Gen on overall Data center revenue. And $TSM ramping up 2nm supply early is a confirmation that AMD will be a major buyer long term. I will also link the thread the Gap between AMD Analysts & Reality and 2nm Ramp Thread so you have more comprehensive view of what I'm writing here. Before I go into detail this is my 2026 Projection: AI GPUs: $35-$50B EPYC Data Center: $15B-$17B Client Segment: $12-$13B Gaming: $6B Embedded: $4B-$5B Total Revenue $70-$100B Non-GAAP net income $18B-$25B Non-GAAP EPS $10.97-$15.40 Foward P/E 55x-70x= $603-$1,078 AMD's Analysts are projecting $0 Revenue for MI450 and sluggish EPYC Growth. Meaning, all analysts are either full of 💩 or Sexist, you decide! Analysts are also projecting 0% growth on AMD "Secret Weapon" Chip as $MSFT said we are at significant Windows refresh and upgrade cycle. Do you think TSMC would allocate more 2nm supply to $AMD at $0 MI450 revenue and sluggish EPYC? 1. EPYC is going to be the leader in lowest Inference! Current Turin cost saving is 95% vs $NVDA or 98-99% on Inference cost when you factor in renting Inference compute from Amazon Web Services, Microsoft Azure, or $NVDA Neocloud pets. TSMC claimed: 10-15% higher performance at iso-power, 25-30% lower power at iso-speed, and ~15% higher transistor density compared to 3nm. This reduces operational expenses (energy, cooling) while increasing throughput per chip. EPYC Turin achieves ~$0.001 per million tokens for batch inference (via vLLM on models like Llama 3 70B), driven by high core counts and low hardware costs. EPYC Venice offers ~1.7x overall performance and up to 70% more compute capability per core, with up to 256 cores (512 threads). Enhanced vector/AI instructions and open-source firmware (openSIL) optimize for inference workloads. AMD Incorporates AI Engines (now part of AMD's XDNA) for on-chip acceleration, improving efficiency for low-latency and edge inference. This reduces reliance on discrete GPUs, lowering system complexity and TCO. Venice SKUs are projected at $3,000-$15,000 ($5,000 for 256-core flagship), far below NVIDIA Rubin ($50,000-$90,000) or AMD's own MI450 GPUs ($40,000-$50,000). High memory bandwidth (up to 1.6 TB/s) supports efficient batch inference. Venice is designed exactly for Large customers that want to lower Inference Cost and MI450 Helios is for Customers that want Training at lowest TCO, TDP as well as lower Upfront 1GW scale(Full build $35-$40B vs $NVDA $55B-$80B). 2. Real World Example: OpenAI's 2025 inference spend reached ~$20B, escalating to even higher total compute rental (mostly inference) amid token volume growth(from video generating). By 2026, with usage doubling (consistent with industry trends: token demand grows 2-5x YoY), assume OpenAI processes ~1,800 billion million-tokens annually $NVDA Blackwell at $0.02-$0.12 is $36B(most optimized) Rubin is projected to be at $0.01/million tokens or $18B annual Inference Cost vs $AMD Venice $0.0005/million tokens or $0.9B annual Inference Cost => Massive saving for OpenAI or anyone that are paying 80-90% Annual Bill for Inference compute. In short, it is unsustainable to pay this much rent vs owning for all current AI players for the medium to long term. Rubin excels in low-latency decode (if Groq integration from $20B deal in 2027-2028), but Venice dominates batch (80% of inference by 2030). Actual savings depend on deployment scale (OpenAI's 6GW AMD plans), electricity rates, and software maturity. If Rubin only hits $0.03, savings swell to $53.1B vs. $17.1B. 3. Will running Inference on Venice and future Gen slow down response generation in 2026 and beyond? Human perception of "fast enough" for chat, agents, search augmentation, summarization, coding assistance is roughly Meaning, EPYC may generate $100B a year on data center revenue, Hence $MSFT $AMZN $META $GOOGL OpenAI xAI and 42+ Countries are leaning AMD for Inference, because the cost saving is MASSIVE! 4. Regular users (you, me, people using ChatGPT, Claude, Gemini, Grok, Perplexity...) are extremely unlikely to notice any slowdown and in many cases might even experience slightly faster or more consistent response times if the industry heavily shifts toward AMD EPYC for inference. What actually happens when companies save massively on inference? When OpenAI , Anthropic , Gemini , Grok Meta .... save billions on the batch/enterprise/RAG layer using EPYC Venice, they typically do one or more of these things with the savings, none of which make your chat slower but enhancing their bottom line(Profit) ~Keep prices the same → make more profit ~Lower subscription prices / increase free tier limits ~Train bigger & better models more frequently ~Offer longer context windows ~Add more reasoning steps / tool calls / agents per query ~Improve multimodal capabilities ~Build more data centers / reduce throttling during peaks In practice the consumer experience usually gets better, not worse, when inference becomes dramatically cheaper. Prime example is $META leaning AMD heavily or currently AMD largest customer. or Grok 2 to Grok 3 heavily used AMD for Inference saving. And most Grok Users reported Groke responses snappier, not slower. 5. What does this mean for potential Revenue? Noted that TSMC is massively ramping 2nm supply for $AMD both MI450 and EPYC. EPYC Conservative projection: FY2025: $10.5B(best Est) FY2026: $16B FY2027: $29B FY2028: $49B FY2029: $75B FY2030: $100B Large customers: $META OpenAI $MSFT $AMZN $GOOGL xAI (Apple?) Smaller customer: $DELL $HPE $SMCI and 42+ other countries. The roadmap to $5 Trillion is very much inevitable as Inference Cost from Renting or owning $NVDA are too high, but $NVDA will still dominate Training market share, where MI families are likely to take 15-20% market share, but the TAM is also expanding Rapidly. Most Institutions are projecting $2-$3Trillion TAM by 2030. $NVDA said $4 Trillion. Dr. Lisa Su said $1 Trillion+ by 2030. So you decide on how much TAM. If you enjoy this kind of analysis, Slap the Like/Repost and Bookmark to please the X Algo as it is Free.99! If you want to support my work further, consider subscribe to see more in-depth analysis! Alright, that is it. Not Financial Advice!

$AMD $5 Trillion MC Is Inevitable Long Term👑 This thread will focus more on Inference! 2026 EPYC "Venice" $TSM 2nm to save Large GW Scale Inference by 40% more than Prior Turin gen. Context: EPYC Turin achieves ~$0.001 per million tokens for batch inference vs $0.02-$0.12/ million tokens as I wrote the thread below. Venice is going to lower cost down to $0.0005-$0.0006/Million Tokens. OpenAI spent roughly $20B on Inference and Training, where 80-90% of that was for Inference per Analysts. AKA Renting Compute is Expensive AF! In this thread, I want to focus on why most analysts and investors are underestimating the role EPYC "Venice" and future Gen on overall Data center revenue. And $TSM ramping up 2nm supply early is a confirmation that AMD will be a major buyer long term. I will also link the thread the Gap between AMD Analysts & Reality and 2nm Ramp Thread so you have more comprehensive view of what I'm writing here. Before I go into detail this is my 2026 Projection: AI GPUs: $35-$50B EPYC Data Center: $15B-$17B Client Segment: $12-$13B Gaming: $6B Embedded: $4B-$5B Total Revenue $70-$100B Non-GAAP net income $18B-$25B Non-GAAP EPS $10.97-$15.40 Foward P/E 55x-70x= $603-$1,078 AMD's Analysts are projecting $0 Revenue for MI450 and sluggish EPYC Growth. Meaning, all analysts are either full of 💩 or Sexist, you decide! Analysts are also projecting 0% growth on AMD "Secret Weapon" Chip as $MSFT said we are at significant Windows refresh and upgrade cycle. Do you think TSMC would allocate more 2nm supply to $AMD at $0 MI450 revenue and sluggish EPYC? 1. EPYC is going to be the leader in lowest Inference! Current Turin cost saving is 95% vs $NVDA or 98-99% on Inference cost when you factor in renting Inference compute from Amazon Web Services, Microsoft Azure, or $NVDA Neocloud pets. TSMC claimed: 10-15% higher performance at iso-power, 25-30% lower power at iso-speed, and ~15% higher transistor density compared to 3nm. This reduces operational expenses (energy, cooling) while increasing throughput per chip. EPYC Turin achieves ~$0.001 per million tokens for batch inference (via vLLM on models like Llama 3 70B), driven by high core counts and low hardware costs. EPYC Venice offers ~1.7x overall performance and up to 70% more compute capability per core, with up to 256 cores (512 threads). Enhanced vector/AI instructions and open-source firmware (openSIL) optimize for inference workloads. AMD Incorporates AI Engines (now part of AMD's XDNA) for on-chip acceleration, improving efficiency for low-latency and edge inference. This reduces reliance on discrete GPUs, lowering system complexity and TCO. Venice SKUs are projected at $3,000-$15,000 ($5,000 for 256-core flagship), far below NVIDIA Rubin ($50,000-$90,000) or AMD's own MI450 GPUs ($40,000-$50,000). High memory bandwidth (up to 1.6 TB/s) supports efficient batch inference. Venice is designed exactly for Large customers that want to lower Inference Cost and MI450 Helios is for Customers that want Training at lowest TCO, TDP as well as lower Upfront 1GW scale(Full build $35-$40B vs $NVDA $55B-$80B). 2. Real World Example: OpenAI's 2025 inference spend reached ~$20B, escalating to even higher total compute rental (mostly inference) amid token volume growth(from video generating). By 2026, with usage doubling (consistent with industry trends: token demand grows 2-5x YoY), assume OpenAI processes ~1,800 billion million-tokens annually $NVDA Blackwell at $0.02-$0.12 is $36B(most optimized) Rubin is projected to be at $0.01/million tokens or $18B annual Inference Cost vs $AMD Venice $0.0005/million tokens or $0.9B annual Inference Cost => Massive saving for OpenAI or anyone that are paying 80-90% Annual Bill for Inference compute. In short, it is unsustainable to pay this much rent vs owning for all current AI players for the medium to long term. Rubin excels in low-latency decode (if Groq integration from $20B deal in 2027-2028), but Venice dominates batch (80% of inference by 2030). Actual savings depend on deployment scale (OpenAI's 6GW AMD plans), electricity rates, and software maturity. If Rubin only hits $0.03, savings swell to $53.1B vs. $17.1B. 3. Will running Inference on Venice and future Gen slow down response generation in 2026 and beyond? Human perception of "fast enough" for chat, agents, search augmentation, summarization, coding assistance is roughly Meaning, EPYC may generate $100B a year on data center revenue, Hence $MSFT $AMZN $META $GOOGL OpenAI xAI and 42+ Countries are leaning AMD for Inference, because the cost saving is MASSIVE! 4. Regular users (you, me, people using ChatGPT, Claude, Gemini, Grok, Perplexity...) are extremely unlikely to notice any slowdown and in many cases might even experience slightly faster or more consistent response times if the industry heavily shifts toward AMD EPYC for inference. What actually happens when companies save massively on inference? When OpenAI , Anthropic , Gemini , Grok Meta .... save billions on the batch/enterprise/RAG layer using EPYC Venice, they typically do one or more of these things with the savings, none of which make your chat slower but enhancing their bottom line(Profit) ~Keep prices the same → make more profit ~Lower subscription prices / increase free tier limits ~Train bigger & better models more frequently ~Offer longer context windows ~Add more reasoning steps / tool calls / agents per query ~Improve multimodal capabilities ~Build more data centers / reduce throttling during peaks In practice the consumer experience usually gets better, not worse, when inference becomes dramatically cheaper. Prime example is $META leaning AMD heavily or currently AMD largest customer. or Grok 2 to Grok 3 heavily used AMD for Inference saving. And most Grok Users reported Groke responses snappier, not slower. 5. What does this mean for potential Revenue? Noted that TSMC is massively ramping 2nm supply for $AMD both MI450 and EPYC. EPYC Conservative projection: FY2025: $10.5B(best Est) FY2026: $16B FY2027: $29B FY2028: $49B FY2029: $75B FY2030: $100B Large customers: $META OpenAI $MSFT $AMZN $GOOGL xAI (Apple?) Smaller customer: $DELL $HPE $SMCI and 42+ other countries. The roadmap to $5 Trillion is very much inevitable as Inference Cost from Renting or owning $NVDA are too high, but $NVDA will still dominate Training market share, where MI families are likely to take 15-20% market share, but the TAM is also expanding Rapidly. Most Institutions are projecting $2-$3Trillion TAM by 2030. $NVDA said $4 Trillion. Dr. Lisa Su said $1 Trillion+ by 2030. So you decide on how much TAM. If you enjoy this kind of analysis, Slap the Like/Repost and Bookmark to please the X Algo as it is Free.99! If you want to support my work further, consider subscribe to see more in-depth analysis! Alright, that is it. Not Financial Advice!

Mike

102,223 次观看 • 8 个月前

SITUATION EXPLAINED: Cerebras raised $5.55 billion in their IPO and closing their first day of trading valued at $66 billion, making it the biggest US tech IPO since Snowflake in 2020. Cerebras makes Wafer-Scale Engine chips built for AI inference. We asked Sarah Fong the main difference between wafer-scale chips and traditional GPUs: - GPUs are great at parallel work (graphics, training) - AI inference is sequential, AKA one token at a time This causes the "memory wall" problem: - Every GPU core needs model weights, KV cache, and activations to do its math - On a GPU, that data lives in off-chip memory (HBM) - Cores constantly load and offload from off-chip memory, which is a huge bottleneck; hardware accounts for ~70% of inference latency Cerebras' chips: -Dinner-plate sized (vs. GPUs which are palm-sized) with tens of thousands of cores -Memory sits directly on top of the cores as distributed SRAM -Weights and KV cache can be accessed at on-chip speeds in the PB/s range, compared with off-chip speeds in the TB/s range achieved by GPUs with HBM.

SITUATION EXPLAINED: Cerebras raised $5.55 billion in their IPO and closing their first day of trading valued at $66 billion, making it the biggest US tech IPO since Snowflake in 2020. Cerebras makes Wafer-Scale Engine chips built for AI inference. We asked Sarah Fong the main difference between wafer-scale chips and traditional GPUs: - GPUs are great at parallel work (graphics, training) - AI inference is sequential, AKA one token at a time This causes the "memory wall" problem: - Every GPU core needs model weights, KV cache, and activations to do its math - On a GPU, that data lives in off-chip memory (HBM) - Cores constantly load and offload from off-chip memory, which is a huge bottleneck; hardware accounts for ~70% of inference latency Cerebras' chips: -Dinner-plate sized (vs. GPUs which are palm-sized) with tens of thousands of cores -Memory sits directly on top of the cores as distributed SRAM -Weights and KV cache can be accessed at on-chip speeds in the PB/s range, compared with off-chip speeds in the TB/s range achieved by GPUs with HBM.

MTS

44,057 次观看 • 3 个月前

THIS IS HOW A SENIOR ENGINEER ACTUALLY SCALES THEMSELVES WITH CLAUDE CODE the biggest change with AI isn't coding faster. it's where you actually spend your time now. more detailed prompts, more code review, more planning, less typing, etc. here's the workflow: this guy has been shipping code since the days of cgi and perl. he uses a compound engineering plugin that runs 5 separate agents on every task. one brainstorms, one plans the technical implementation, one executes, one reviews, one checks different verticals. every step is documented in markdown files. it's slow and way more waiting. but the output quality is way higher because each agent is focused on one thing. then the REAL multiplier is in git worktrees if Claude Code made you 10x faster, worktrees multiplies that again depending on how many agents you can manage in parallel his team runs 4-8 Claude Code sessions at the same time across different worktrees with each one working on a separate task. the skill is managing multiple AI agents in parallel without losing track, that's the next evolution of engineering

THIS IS HOW A SENIOR ENGINEER ACTUALLY SCALES THEMSELVES WITH CLAUDE CODE the biggest change with AI isn't coding faster. it's where you actually spend your time now. more detailed prompts, more code review, more planning, less typing, etc. here's the workflow: this guy has been shipping code since the days of cgi and perl. he uses a compound engineering plugin that runs 5 separate agents on every task. one brainstorms, one plans the technical implementation, one executes, one reviews, one checks different verticals. every step is documented in markdown files. it's slow and way more waiting. but the output quality is way higher because each agent is focused on one thing. then the REAL multiplier is in git worktrees if Claude Code made you 10x faster, worktrees multiplies that again depending on how many agents you can manage in parallel his team runs 4-8 Claude Code sessions at the same time across different worktrees with each one working on a separate task. the skill is managing multiple AI agents in parallel without losing track, that's the next evolution of engineering

Om Patel

159,590 次观看 • 4 个月前

Too many blockchain infra providers run in the cloud or on rented servers That's a problem is built different 3 reasons why you want to work with us: 1) Best in class, enterprise grade bare metal, all fully owned 2) SOC 2 Type II certified--not just the data centers, but our own internal team 3) Optimized performance on our infra for your staked assets Curious to learn more? Check out the video and short overview thread below!

Too many blockchain infra providers run in the cloud or on rented servers That's a problem is built different 3 reasons why you want to work with us: 1) Best in class, enterprise grade bare metal, all fully owned 2) SOC 2 Type II certified--not just the data centers, but our own internal team 3) Optimized performance on our infra for your staked assets Curious to learn more? Check out the video and short overview thread below!

Phunky (Ryan) 🌎🥩

32,792 次观看 • 1 年前

🚨 ALIBABA JUST OPEN-SOURCED THE MOST POWERFUL MODEL IN THE WORLD - Qwen3.5 plus And it changes the AI race forever. This is not another “bigger model” release. This is a complete architectural reset. 397B parameters → only 17B active at inference, Qwen3.5 plus delivers performance on par with GPT-5.2 and Gemini-3-Pro. A new Hybrid Sparse MoE + Linear Attention architecture + Multi-Token Prediction = near-instant long-context responses and massive speed gains. Built as a true native multimodal model, it doesn’t just chat it builds. With a single prompt: “Use the React framework to create a visually impressive Mapbox demo featuring locations in Beijing and Shanghai” it generates a live interactive 3D map UI, structured components, real layouts, and production-ready frontend logic. This is real agentic, multimodal coding not a static code block. MMLU-Pro: 87.8 GPQA: 88.4 IFBench: 76.5 ⚡ up to 8× faster inference And it’s fully open-sourced under Apache-2.0. No locked APIs. No restricted tiers. 201 languages supported bringing frontier AI to developers, startups, and researchers worldwide. The race is no longer about the biggest model. It’s about the most usable intelligence. Open source just entered the top tier.

🚨 ALIBABA JUST OPEN-SOURCED THE MOST POWERFUL MODEL IN THE WORLD - Qwen3.5 plus And it changes the AI race forever. This is not another “bigger model” release. This is a complete architectural reset. 397B parameters → only 17B active at inference, Qwen3.5 plus delivers performance on par with GPT-5.2 and Gemini-3-Pro. A new Hybrid Sparse MoE + Linear Attention architecture + Multi-Token Prediction = near-instant long-context responses and massive speed gains. Built as a true native multimodal model, it doesn’t just chat it builds. With a single prompt: “Use the React framework to create a visually impressive Mapbox demo featuring locations in Beijing and Shanghai” it generates a live interactive 3D map UI, structured components, real layouts, and production-ready frontend logic. This is real agentic, multimodal coding not a static code block. MMLU-Pro: 87.8 GPQA: 88.4 IFBench: 76.5 ⚡ up to 8× faster inference And it’s fully open-sourced under Apache-2.0. No locked APIs. No restricted tiers. 201 languages supported bringing frontier AI to developers, startups, and researchers worldwide. The race is no longer about the biggest model. It’s about the most usable intelligence. Open source just entered the top tier.

RAVI KUMAR SAHU

380,123 次观看 • 6 个月前

🚨 In the new episode of my AfterMath series, I ask: Can a computer reach the full capacity of the human mind? With the recent advances in AI, this has become a hot topic for debate. I focus on the area I know best: mathematics. And I further narrow it down to the question: Can a computer be as good as a human in dealing with natural numbers? (Numbers like 0,1,2,3,4, and so on.) I make an important distinction between Type I and Type II statements about natural numbers. Statements of Type I are about specific numbers (here computers excel), and statements of Type II are GENERAL statements that apply to ALL NUMBERS at once (such as Fermat's Last Theorem). Mathematics is about Type II statements, and in this domain computers hit a wall, which I'd like to call the "Turing Wall." The problem is that Type II statements can't be reduced to Type I statements as computer's memory is finite. To handle Type II statements, one has to use FORMAL SYSTEMS, in which we symbolically encode properties of natural numbers, so they could be programmed on a computer. Once we choose the axioms and the rules of inference, we can run a computer, and it will produce many true statements about natural numbers. However, Tarski's Undefinability Theorem, which I introduce in this episode (it is closely related to Gödel's Incompleteness Theorems but is even more relevant to these issues, in my opinion), shows that this way we can NOT get all true statements about natural numbers. In fact, the notion of "true statement" is not contained in a formal system. To introduce this notion in a given formal system, one has to choose a model of this formal system, and there are many inequivalent models. Large Language Models give mathematicians a great tool for research, but they can't be used effectively for the kinds of foundational questions of mathematics we are discussing here. At the end of the episode, I go back to the 3-dimensional sphere I talked about in Episode 2. I give a 4-dimensional spacetime demonstration of it, using... a balloon. 😃 In a future video, I am planning to discuss the 3-dimensional sphere in more detail with my friend . See the links to the entire episode below.⤵️⤵️

🚨 In the new episode of my AfterMath series, I ask: Can a computer reach the full capacity of the human mind? With the recent advances in AI, this has become a hot topic for debate. I focus on the area I know best: mathematics. And I further narrow it down to the question: Can a computer be as good as a human in dealing with natural numbers? (Numbers like 0,1,2,3,4, and so on.) I make an important distinction between Type I and Type II statements about natural numbers. Statements of Type I are about specific numbers (here computers excel), and statements of Type II are GENERAL statements that apply to ALL NUMBERS at once (such as Fermat's Last Theorem). Mathematics is about Type II statements, and in this domain computers hit a wall, which I'd like to call the "Turing Wall." The problem is that Type II statements can't be reduced to Type I statements as computer's memory is finite. To handle Type II statements, one has to use FORMAL SYSTEMS, in which we symbolically encode properties of natural numbers, so they could be programmed on a computer. Once we choose the axioms and the rules of inference, we can run a computer, and it will produce many true statements about natural numbers. However, Tarski's Undefinability Theorem, which I introduce in this episode (it is closely related to Gödel's Incompleteness Theorems but is even more relevant to these issues, in my opinion), shows that this way we can NOT get all true statements about natural numbers. In fact, the notion of "true statement" is not contained in a formal system. To introduce this notion in a given formal system, one has to choose a model of this formal system, and there are many inequivalent models. Large Language Models give mathematicians a great tool for research, but they can't be used effectively for the kinds of foundational questions of mathematics we are discussing here. At the end of the episode, I go back to the 3-dimensional sphere I talked about in Episode 2. I give a 4-dimensional spacetime demonstration of it, using... a balloon. 😃 In a future video, I am planning to discuss the 3-dimensional sphere in more detail with my friend . See the links to the entire episode below.⤵️⤵️

Edward Frenkel

107,086 次观看 • 11 个月前

In just one week, Binh and I trained a full-body Unitree G1. Here's a recap: 1. Secured a Unitree G1 humanoid through a LinkedIn post 2. Deployed TWIST2 full-body teleoperation pipelines 3. Adapted TWIST2 for Zed stereo camera & collected full-body teleoperation samples (carried by Binh ) 4. Adapted & fine-tuned NVIDIA Gr00T N1.5 VLA on the TWIST2 public datasets, which I fine-tuned on an 8xNVIDIA H100 Cluster. We picked Gr00T N1.5 as it was trained with Unitree G1 embodiment data. 5. Adapted the TWIST2 codebase to stream in the actions from Gr00T via ZMQ using a co-located NVIDIA H100 for ~200ms inference latency 6. Tested the model in sim, then deployed to the real-world Unitree G1. We streamed a training sample observation to the VLA (as we didn't want to break robot in case real observations were OOD) We were the first team in the world to deploy the full TWIST2 data collection pipeline to the unitree g1 :) Much more work ahead though, which I'll work on as a side-project over the next months: 1. Exploring the various types of 'world models': video backbones, dynamics models, v-jepa-2 models. I believe these will generalize better & train much more data-efficiently than VLM backbones 2. Speeding up inference - I believe low-latency robotics inference will be a big challenge. There are many works in video diffusion which I'd like to test (e.g. SageAttention, SparseAttention, Drifting Models). Perhaps also writing custom CUDA kernels. 3. Economics of inference scaling :) What will be the compute demands as we scale inference up to millions of humanoids? Will it run on edge or on distributed 'co-located' inference clusters? These are questions I'd like to answer. Adapted TWIST2 codebase: Adapted Gr00T-N1.5 codebase: The ETH Robotics Club are doing a cool GTC Golden ticket competition with NVIDIA , so this is my submission :) The DGX Spark compute will get me a long way with initial prototyping & especially working on inference optimization for next-gen Blackwell GPUs #NVIDIAGTC #GOLDENTICKET #ETHRC

In just one week, Binh and I trained a full-body Unitree G1. Here's a recap: 1. Secured a Unitree G1 humanoid through a LinkedIn post 2. Deployed TWIST2 full-body teleoperation pipelines 3. Adapted TWIST2 for Zed stereo camera & collected full-body teleoperation samples (carried by Binh ) 4. Adapted & fine-tuned NVIDIA Gr00T N1.5 VLA on the TWIST2 public datasets, which I fine-tuned on an 8xNVIDIA H100 Cluster. We picked Gr00T N1.5 as it was trained with Unitree G1 embodiment data. 5. Adapted the TWIST2 codebase to stream in the actions from Gr00T via ZMQ using a co-located NVIDIA H100 for ~200ms inference latency 6. Tested the model in sim, then deployed to the real-world Unitree G1. We streamed a training sample observation to the VLA (as we didn't want to break robot in case real observations were OOD) We were the first team in the world to deploy the full TWIST2 data collection pipeline to the unitree g1 :) Much more work ahead though, which I'll work on as a side-project over the next months: 1. Exploring the various types of 'world models': video backbones, dynamics models, v-jepa-2 models. I believe these will generalize better & train much more data-efficiently than VLM backbones 2. Speeding up inference - I believe low-latency robotics inference will be a big challenge. There are many works in video diffusion which I'd like to test (e.g. SageAttention, SparseAttention, Drifting Models). Perhaps also writing custom CUDA kernels. 3. Economics of inference scaling :) What will be the compute demands as we scale inference up to millions of humanoids? Will it run on edge or on distributed 'co-located' inference clusters? These are questions I'd like to answer. Adapted TWIST2 codebase: Adapted Gr00T-N1.5 codebase: The ETH Robotics Club are doing a cool GTC Golden ticket competition with NVIDIA , so this is my submission :) The DGX Spark compute will get me a long way with initial prototyping & especially working on inference optimization for next-gen Blackwell GPUs #NVIDIAGTC #GOLDENTICKET #ETHRC

Arnie Ramesh

23,236 次观看 • 6 个月前

50% of my consulting work right now is helping companies use open-source models at scale. Everyone knows how to use an open-source LLM on their computers, but it's really hard to do this at scale for thousands of users. Here is how this plays out: 1. A team builds a prototype using DeepSeek. 2. Everything looks good. It works! 3. They follow an online guide to deploy the model online. 4. They ask 10 users to try the app. 5. Latency spikes everywhere. 6. The entire system halts. 7. They blame DeepSeek and try again using a new model. The problem is always with scaling inference, not the model. Here is one recommendation I give companies: Check out Nebius Token Factory if you don't want to ever think about deploying an open-source model again. This is a managed inference platform for deploying open-source LLMs at scale. This is not for prototypes or research experiments. This is for when you have a real application with real users. Three important notes about Token Factory: • You have complete control over how inference runs. • You have predictable tail latency (P99, not averages). • No surprise costs when you scale up. You can preplan your budget. Check it out here: Here are two codes you can use to get 100 hours of GPU usage on me: ymJLFa2ARYSKEdqb AdckcZaYjm7KqYY7 Thanks to the Nebius team for their continuous partnership.

50% of my consulting work right now is helping companies use open-source models at scale. Everyone knows how to use an open-source LLM on their computers, but it's really hard to do this at scale for thousands of users. Here is how this plays out: 1. A team builds a prototype using DeepSeek. 2. Everything looks good. It works! 3. They follow an online guide to deploy the model online. 4. They ask 10 users to try the app. 5. Latency spikes everywhere. 6. The entire system halts. 7. They blame DeepSeek and try again using a new model. The problem is always with scaling inference, not the model. Here is one recommendation I give companies: Check out Nebius Token Factory if you don't want to ever think about deploying an open-source model again. This is a managed inference platform for deploying open-source LLMs at scale. This is not for prototypes or research experiments. This is for when you have a real application with real users. Three important notes about Token Factory: • You have complete control over how inference runs. • You have predictable tail latency (P99, not averages). • No surprise costs when you scale up. You can preplan your budget. Check it out here: Here are two codes you can use to get 100 hours of GPU usage on me: ymJLFa2ARYSKEdqb AdckcZaYjm7KqYY7 Thanks to the Nebius team for their continuous partnership.

Santiago

48,443 次观看 • 7 个月前

I pay Claude $20 a month. Most $TAO holders do too. There is a stack you can build in 15 minutes that fixes that completely. It runs on Bittensor. It costs $10. You do not write a single line of code. Here is how every AI chat product actually works under the hood. Three layers. Always three. The model. The brain. GPT, Claude, DeepSeek, Kimi, GLM. The inference layer. The GPU that runs the model when you hit send. The interface. The chat box you actually look at. ChatGPT and Claude bundle all three and hand you the result. You cannot change the model. You cannot change the inference. The interface is non-negotiable. Every prompt you type goes to a server run by a private company whose terms of service can quietly change next month. The anti-ChatGPT move is to pick each layer yourself. This is where $TAO comes in. Chutes is Subnet 64 on Bittensor. It is the inference layer. Open source models like DeepSeek, Kimi, GLM, and Llama get served by a global network of miner-operated GPUs. Validators score the output quality. The best inference wins the emissions. You hit send. A miner somewhere runs your prompt. You get the answer back. The TAO you hold is in part paying for the GPU you just used. The basic stack is one URL. chutes. ai/chat No account. No API key. No setup. Switch models mid-conversation. Web search built in. Image generation. File uploads. Free. The advanced stack is Chutes plus TypingMind. One-time license. No recurring fee. Plugins, agents, custom personas, a prompt library you build over months. Full model switching between Chutes, OpenAI, and Anthropic from the same window. Total cost: $10 a month to Chutes for inference. That $10 buys you $50 in actual usage. But here is the signal most people missed inside this story. Chutes ran a free tier until February. Then they killed it. Then they raised the minimum to $10 in May. Most people saw that as bad news. It is the opposite. Free things on the internet do not last. Real products do. Chutes is becoming a real product. A subnet that generates actual revenue from actual users paying actual money for actual AI inference. That is what $43 million in Q1 network revenue looks like at the individual subnet level. And there is one more thing ChatGPT and Claude cannot offer that Chutes already has. Trusted Execution Environments. Your prompt gets encrypted on your device, shipped to a confidential compute GPU, and the lock only breaks inside the chip. The miner running the model physically cannot read your prompt. ChatGPT cannot promise that. Claude cannot promise that. Bittensor already built it. You are holding a network where the subnets are generating real revenue, shipping real privacy infrastructure, and replacing $20 a month centralised subscriptions with $10 a month decentralised inference. The people who use the product always understand the investment better than the people who only watch the price.

I pay Claude $20 a month. Most $TAO holders do too. There is a stack you can build in 15 minutes that fixes that completely. It runs on Bittensor. It costs $10. You do not write a single line of code. Here is how every AI chat product actually works under the hood. Three layers. Always three. The model. The brain. GPT, Claude, DeepSeek, Kimi, GLM. The inference layer. The GPU that runs the model when you hit send. The interface. The chat box you actually look at. ChatGPT and Claude bundle all three and hand you the result. You cannot change the model. You cannot change the inference. The interface is non-negotiable. Every prompt you type goes to a server run by a private company whose terms of service can quietly change next month. The anti-ChatGPT move is to pick each layer yourself. This is where $TAO comes in. Chutes is Subnet 64 on Bittensor. It is the inference layer. Open source models like DeepSeek, Kimi, GLM, and Llama get served by a global network of miner-operated GPUs. Validators score the output quality. The best inference wins the emissions. You hit send. A miner somewhere runs your prompt. You get the answer back. The TAO you hold is in part paying for the GPU you just used. The basic stack is one URL. chutes. ai/chat No account. No API key. No setup. Switch models mid-conversation. Web search built in. Image generation. File uploads. Free. The advanced stack is Chutes plus TypingMind. One-time license. No recurring fee. Plugins, agents, custom personas, a prompt library you build over months. Full model switching between Chutes, OpenAI, and Anthropic from the same window. Total cost: $10 a month to Chutes for inference. That $10 buys you $50 in actual usage. But here is the signal most people missed inside this story. Chutes ran a free tier until February. Then they killed it. Then they raised the minimum to $10 in May. Most people saw that as bad news. It is the opposite. Free things on the internet do not last. Real products do. Chutes is becoming a real product. A subnet that generates actual revenue from actual users paying actual money for actual AI inference. That is what $43 million in Q1 network revenue looks like at the individual subnet level. And there is one more thing ChatGPT and Claude cannot offer that Chutes already has. Trusted Execution Environments. Your prompt gets encrypted on your device, shipped to a confidential compute GPU, and the lock only breaks inside the chip. The miner running the model physically cannot read your prompt. ChatGPT cannot promise that. Claude cannot promise that. Bittensor already built it. You are holding a network where the subnets are generating real revenue, shipping real privacy infrastructure, and replacing $20 a month centralised subscriptions with $10 a month decentralised inference. The people who use the product always understand the investment better than the people who only watch the price.

2xnmore

27,088 次观看 • 3 个月前

MEET THE NVIDIA KILLER: OpenAI bet $10 BILLION on this company that makes chips 20x faster than Nvidia's. If this plays out as expected, it’s over for Nvidia. Cerebras Systems just locked in 750 megawatts of computing power to OpenAI through 2028. For reference: that's equivalent to the annual power consumption of 600,000 US homes. The deal? Over $10 billion. Here's what nobody understands: Cerebras doesn't make normal chips. Nvidia sells you thousands of tiny chips that you connect together. Cerebras makes ONE chip. A single wafer-scale processor the size of a dinner plate. 900,000 AI cores. 4 trillion transistors. All on one piece of silicon. The result? When OpenAI tested it, Cerebras ran inference 20X FASTER than Nvidia GPUs. That's not incremental improvement. That's a different category of performance. But here's where the story gets wild: Four months ago, Cerebras was a struggling company. Their IPO filing revealed that 87% of their revenue came from ONE customer: G42, a UAE-based AI firm. The US government launched a national security review. G42 had ties to Huawei. Ties to China. The IPO collapsed. Investors panicked. Cerebras withdrew their filing in October 2025. Most startups would've been dead. Instead, Cerebras did the opposite. They raised $1.1 billion at an $8.1 billion valuation. Kicked G42 out of the cap table entirely. Got CFIUS clearance. Then landed the OpenAI deal. Now they're raising ANOTHER $1 billion at a $22 billion valuation. They more than DOUBLED their valuation in 4 months. From near-death to $22 billion. While getting rid of their biggest customer. Why OpenAI chose them: ChatGPT has 900 million weekly users. Sam Altman keeps saying they have a "severe shortage" of compute. They need SPEED, not just power. When you ask ChatGPT a question, there's a loop happening: You send request → model thinks → sends response back Nvidia chips are fast at training models. Cerebras chips are built specifically for inference. For real-time responses. For the exact bottleneck OpenAI is trying to solve. Sachin Katti from OpenAI said it best: "Cerebras adds a dedicated low-latency inference solution to our platform. That means faster responses, more natural interactions, and a stronger foundation to scale real-time AI to many more people." In other words: "We need this to scale ChatGPT." The competitive landscape just shifted: Nvidia announced a $100 billion deal with OpenAI in September. But it's still not finalized. Meanwhile, Cerebras closed their deal before Thanksgiving. And it's ALREADY being deployed. Here's the part that should terrify Nvidia: In December, Nvidia bought Groq for $20 billion. Groq makes fast inference chips. Just like Cerebras. So why would Nvidia spend $20 billion buying a competitor to something they supposedly already dominate? Because they know what's coming. Inference is the new battleground. And Cerebras is winning it. The IPO is coming Q2 2026. After this OpenAI deal, Cerebras now has: ✓ IBM contracts ✓ Department of Energy contracts ✓ OpenAI locked in for 3 years ✓ $22 billion valuation ✓ CFIUS clearance ✓ Zero customer concentration risk They went from 87% revenue dependency on one customer to the most diversified chip company outside Nvidia. In four months. The lesson? Smart money doesn't follow headlines. It follows where the AI leaders are actually spending. OpenAI didn't announce this deal for publicity. They need Cerebras hardware to scale ChatGPT. That's a $10 billion vote of confidence. While everyone's watching Nvidia stock, the real war is happening in inference. And the company with ONE giant chip just beat the company with thousands of tiny ones. What do you think happens when Cerebras IPOs?

MEET THE NVIDIA KILLER: OpenAI bet $10 BILLION on this company that makes chips 20x faster than Nvidia's. If this plays out as expected, it’s over for Nvidia. Cerebras Systems just locked in 750 megawatts of computing power to OpenAI through 2028. For reference: that's equivalent to the annual power consumption of 600,000 US homes. The deal? Over $10 billion. Here's what nobody understands: Cerebras doesn't make normal chips. Nvidia sells you thousands of tiny chips that you connect together. Cerebras makes ONE chip. A single wafer-scale processor the size of a dinner plate. 900,000 AI cores. 4 trillion transistors. All on one piece of silicon. The result? When OpenAI tested it, Cerebras ran inference 20X FASTER than Nvidia GPUs. That's not incremental improvement. That's a different category of performance. But here's where the story gets wild: Four months ago, Cerebras was a struggling company. Their IPO filing revealed that 87% of their revenue came from ONE customer: G42, a UAE-based AI firm. The US government launched a national security review. G42 had ties to Huawei. Ties to China. The IPO collapsed. Investors panicked. Cerebras withdrew their filing in October 2025. Most startups would've been dead. Instead, Cerebras did the opposite. They raised $1.1 billion at an $8.1 billion valuation. Kicked G42 out of the cap table entirely. Got CFIUS clearance. Then landed the OpenAI deal. Now they're raising ANOTHER $1 billion at a $22 billion valuation. They more than DOUBLED their valuation in 4 months. From near-death to $22 billion. While getting rid of their biggest customer. Why OpenAI chose them: ChatGPT has 900 million weekly users. Sam Altman keeps saying they have a "severe shortage" of compute. They need SPEED, not just power. When you ask ChatGPT a question, there's a loop happening: You send request → model thinks → sends response back Nvidia chips are fast at training models. Cerebras chips are built specifically for inference. For real-time responses. For the exact bottleneck OpenAI is trying to solve. Sachin Katti from OpenAI said it best: "Cerebras adds a dedicated low-latency inference solution to our platform. That means faster responses, more natural interactions, and a stronger foundation to scale real-time AI to many more people." In other words: "We need this to scale ChatGPT." The competitive landscape just shifted: Nvidia announced a $100 billion deal with OpenAI in September. But it's still not finalized. Meanwhile, Cerebras closed their deal before Thanksgiving. And it's ALREADY being deployed. Here's the part that should terrify Nvidia: In December, Nvidia bought Groq for $20 billion. Groq makes fast inference chips. Just like Cerebras. So why would Nvidia spend $20 billion buying a competitor to something they supposedly already dominate? Because they know what's coming. Inference is the new battleground. And Cerebras is winning it. The IPO is coming Q2 2026. After this OpenAI deal, Cerebras now has: ✓ IBM contracts ✓ Department of Energy contracts ✓ OpenAI locked in for 3 years ✓ $22 billion valuation ✓ CFIUS clearance ✓ Zero customer concentration risk They went from 87% revenue dependency on one customer to the most diversified chip company outside Nvidia. In four months. The lesson? Smart money doesn't follow headlines. It follows where the AI leaders are actually spending. OpenAI didn't announce this deal for publicity. They need Cerebras hardware to scale ChatGPT. That's a $10 billion vote of confidence. While everyone's watching Nvidia stock, the real war is happening in inference. And the company with ONE giant chip just beat the company with thousands of tiny ones. What do you think happens when Cerebras IPOs?

Ricardo

28,088 次观看 • 7 个月前

Ricky Ponting points out how good this England field is to Pat Cummins, saying... "They've got a field set for some shorter type bowling. Short leg, deep backward square, regulation fine leg. "We saw Cummins last over get unsettled by one that angled back up into the left armpit. He's not a great ducker of the ball, he tries to ride the bounce and that's why I like this field. "You got one back on the hook so you can't play that, you got one waiting under the helmet at short leg... *Cummins inside edges one to short leg* "For that one!" #Ashes

Ricky Ponting points out how good this England field is to Pat Cummins, saying... "They've got a field set for some shorter type bowling. Short leg, deep backward square, regulation fine leg. "We saw Cummins last over get unsettled by one that angled back up into the left armpit. He's not a great ducker of the ball, he tries to ride the bounce and that's why I like this field. "You got one back on the hook so you can't play that, you got one waiting under the helmet at short leg... Cummins inside edges one to short leg "For that one!" #Ashes

7Cricket

68,121 次观看 • 8 个月前

HERMES AGENT RUNS A POWERFUL OPS SYSTEM ON ONE $6/MO VPS. NO GPU. 90% of people use AI like a search box. Type, grab the answer, close the tab. Tomorrow it has no idea who they are. I stopped doing that 3 months ago. Now Hermes runs as a 24/7 operating system - no GPU. It curates my research before I wake up, drafts content, tracks its own tasks, and pings me on Telegram from anywhere. The math nobody runs: → Agents that wrote 20+ of their own skills finish tasks ~40% faster - not a better model, just accumulated procedure → The full content system burns ~10-11M tokens/mo - and model choice alone swings the bill 10x → Same crons, same goals, one orchestrator + 3 specialists = the output of a small content team, running while I sleep → Runtime is free (MIT). You pay for cheap inference + a $6 box. That’s it. → Replaces the work people pay a VA $500-$2,000/mo to do The catch? Compounding takes 2-4 weeks of consistent use before it’s obvious. Most people quit before day 14. The operator doesn’t prompt. The operator builds the system. Reply “Hermes” and I’ll send you the step-by-step Playbook in your DMs 👇

HERMES AGENT RUNS A POWERFUL OPS SYSTEM ON ONE $6/MO VPS. NO GPU. 90% of people use AI like a search box. Type, grab the answer, close the tab. Tomorrow it has no idea who they are. I stopped doing that 3 months ago. Now Hermes runs as a 24/7 operating system - no GPU. It curates my research before I wake up, drafts content, tracks its own tasks, and pings me on Telegram from anywhere. The math nobody runs: → Agents that wrote 20+ of their own skills finish tasks ~40% faster - not a better model, just accumulated procedure → The full content system burns ~10-11M tokens/mo - and model choice alone swings the bill 10x → Same crons, same goals, one orchestrator + 3 specialists = the output of a small content team, running while I sleep → Runtime is free (MIT). You pay for cheap inference + a $6 box. That’s it. → Replaces the work people pay a VA $500-$2,000/mo to do The catch? Compounding takes 2-4 weeks of consistent use before it’s obvious. Most people quit before day 14. The operator doesn’t prompt. The operator builds the system. Reply “Hermes” and I’ll send you the step-by-step Playbook in your DMs 👇

ZEUS⚡️

10,647 次观看 • 2 个月前

Introducing A tool that allows you compute anything. Think of how programming languages work. You are not restricted to a certain type of program rather you are given the building blocks to build whatever you want. That's the goal for A turing-complete math platform that gives you the building blocks to calculate, experiment and build deep computational flows. This is the first of many demos. Try it out and join me on this journey as I build the best math and computation platform.

Introducing A tool that allows you compute anything. Think of how programming languages work. You are not restricted to a certain type of program rather you are given the building blocks to build whatever you want. That's the goal for A turing-complete math platform that gives you the building blocks to calculate, experiment and build deep computational flows. This is the first of many demos. Try it out and join me on this journey as I build the best math and computation platform.

Godspower Eze

19,676 次观看 • 4 个月前

$NVDA $GFS NVIDIA’s reported agreement to acquire Groq for $20B in cash (per CNBC, amplified via Reuters and other wire coverage) represents a materially different strategic posture than NVIDIA’s prior M&A pattern, given both the headline size (largest reported NVIDIA acquisition to date) and the unusual carve-out that Groq’s early-stage cloud business would not be included. Public reporting indicates the information originated from Alex Davis, CEO of Disruptive (lead investor in Groq’s latest financing), and that neither NVIDIA nor Groq had issued an immediate confirmation at the time of publication. The same reporting frames the transaction as coming together quickly, only months after Groq raised $750M at a ~$6.9B valuation, and highlights Groq’s positioning as a high-performance inference chip vendor founded by ex-Google TPU engineers. Groq is best understood as a vertically integrated inference acceleration company whose core asset is an application-specific processor optimized for deterministic, low-latency execution of transformer-style workloads, paired with a compiler-led software stack and a distribution layer (GroqCloud) designed to reduce developer friction via OpenAI-compatible APIs and integrations. Groq brands its architecture as a Language Processing Unit (LPU) and consistently emphasizes that the design target is inference, not training. The company’s own architecture description centers on 1-core execution, large on-chip SRAM used as primary storage (explicitly not cache), a custom compiler that statically schedules compute and communication, and direct chip-to-chip connectivity intended to coordinate multi-chip execution without relying on conventional caching hierarchies or dynamic runtime scheduling. The technical premise is a deliberate inversion of the conventional GPU approach. GPUs deliver throughput via massively parallel, multi-core execution with dynamic scheduling, complex memory hierarchies, and heavy reliance on off-chip HBM bandwidth and sophisticated runtime/kernel optimization. Groq instead argues that inference bottlenecks are driven by latency variance (tail latency), synchronization overhead, and memory access unpredictability inherent in dynamically scheduled, cache-heavy architectures, particularly when workloads are latency sensitive and batch sizes cannot be inflated. Groq’s solution is to move “control” into the compiler: the full execution graph and inter-chip communication schedule are computed ahead of time down to clock-cycle granularity, with deterministic execution designed to reduce run-to-run variance. In Groq’s framing, the removal of caches, reorder buffers, speculative execution overhead, and other sources of contention enables predictable latency and high utilization without per-model kernel engineering typical of GPU tuning cycles. A critical nuance is that Groq’s determinism is not merely a software claim; it is tightly coupled to architectural constraints and system design choices that trade flexibility for predictability. Third-party technical commentary indicates Groq’s chip uses a fully deterministic VLIW-style approach with minimal buffering, no external memory, and heavy dependence on sharding models across many chips because on-chip SRAM capacity is limited. SemiAnalysis describes a ~725 mm^2 die on GlobalFoundries 14nm with ~230MB of SRAM and notes that “no useful models” fit on a single chip, forcing multi-chip partitioning for modern LLMs and driving a system-level design where networking and compilation are first-class scheduling problems rather than ancillary infrastructure. This is consistent with Groq’s own messaging that tensor parallelism across chips is a primary design goal, enabled by large on-chip SRAM and compile-time coordination of compute plus interconnect. The on-chip SRAM emphasis is central to Groq’s latency story and also its most constraining trade-off. Groq claims on-chip SRAM bandwidth “upwards of 80 TB/s” and contrasts that with off-chip HBM bandwidth “about 8 TB/s,” asserting a potential 10x advantage from bandwidth plus reduced trips across chip-to-memory boundaries. While these comparisons are marketing-oriented and depend on workload specifics, the architectural implication is clear: Groq prioritizes ultra-fast local weight/activation access and then scales capacity by adding chips, not by attaching large off-chip memory pools. This design can reduce latency for sequential inference layers and minimize unpredictable stalls, but it pushes complexity into partitioning strategy, interconnect topology, and compiler scheduling, and it increases the number of chips needed for very large parameter counts and large KV-cache footprints. Groq also highlights numeric formats and compiler-driven precision management as a performance lever. In its 2025 technical blog, Groq describes “TruePoint numerics,” including 100-bit intermediate accumulation and selective quantization choices (FP32 for attention-sensitive operations, block floating point for MoE weights, FP8 storage in error-tolerant layers), and claims 2-4x speedups versus BF16 without measurable accuracy degradation on benchmarks such as MMLU and HumanEval. Even if the absolute uplift is workload dependent, the strategic point is that Groq is pursuing performance via end-to-end co-design: precision policy is not just hardware capability (FP8/BF16) but compiler-enforced mapping of precision to error sensitivity, which can matter materially for inference cost-per-token if it reduces memory traffic and boosts throughput without forcing aggressive, accuracy-damaging quantization. Independent performance datapoints indicate Groq has been credible on latency-oriented inference speed, at least for certain regimes. EE Times reported in 2023 that Groq demonstrated Llama-2 70B inference at ~240 tokens/s per user on a cloud-based dev system described as 10 racks and 64 chips, using the company’s 1st-gen silicon introduced several years earlier. Separate Groq commentary around independent benchmarking cites results showing ~241 tokens/s throughput and ~0.8s time to receive 100 output tokens for a Llama-2 70B API configuration, positioning the platform as a step-change in “available speed” for certain interactive use cases. These figures do not settle total cost-of-ownership versus GPUs or hyperscaler ASICs, but they establish that Groq’s system-level architecture can deliver strong single-user throughput and latency on large models when properly partitioned and scheduled. GroqCloud is the commercial wrapper that packages this hardware/software stack as “tokens-as-a-service,” aiming to make Groq adoption feel like switching API endpoints rather than adopting new silicon. Groq’s documentation states its API is designed to be “mostly compatible” with OpenAI client libraries, and its pricing page provides model-specific token rates, published speeds (tokens/s), prompt caching discounts, and batch processing discounts. For example, pricing lists inputs as low as $0.05 per 1M tokens and outputs as low as $0.08 per 1M tokens for certain smaller LLM configurations, with higher prices for larger models and long-context or MoE variants; it also advertises prompt caching with a 50% discount on cached input tokens for certain models and a batch API offering 50% lower cost for asynchronous processing windows. These mechanics are economically important because they demonstrate Groq’s go-to-market is not simply “sell chips,” but “sell predictable unit economics per token,” with tooling (batch, caching) that directly targets inference cost drivers (reused prompts, throughput smoothing, and asynchronous workloads). The cloud footprint and distribution partnerships indicate Groq has been building an inference-native “edge within the cloud” strategy rather than competing head-on with hyperscalers on breadth of services. A 2025 Groq newsroom release describes a European deployment in Helsinki with Equinix, positioned as latency reduction and data governance for European customers, and explicitly references Equinix Fabric enabling private connectivity to GroqCloud over public, private, or sovereign infrastructure. The same release enumerates additional capacity in the U.S. (Equinix, DataBank), Canada (Bell Canada), and Saudi Arabia (HUMAIN), and states these sites collectively served more than 20M tokens/s across Groq’s global network at that time. That supply-side metric matters because it provides a directional sense that Groq is scaling capacity as a network, not merely as a chip vendor. Customer disclosure is inherently limited because Groq is private and many enterprise deployments are not public, but Groq’s marketing materials and partnerships provide signals about demand vectors. The company’s public website displays logos of large consumer and enterprise brands (e.g., Dropbox, Vercel, Chevron, Volkswagen, Canva, Robinhood, Riot Games, Workday, Ramp) and includes a published customer quote claiming a 7.41x chat speed increase and an 89% cost reduction after moving to GroqCloud, followed by a tripling of token consumption. While marketing claims should be treated as case-specific and not generalized, they indicate that Groq is targeting both AI-native developers (who measure success by latency and cost-per-token) and enterprise buyers (who care about predictable performance and governance). Supplier and dependency mapping for Groq spans 3 layers: silicon production, system integration, and cloud infrastructure. On silicon, third-party analysis indicates GlobalFoundries 14nm for the 1st-gen Groq chip, implying a supply chain less constrained by the most capacity-tight leading-edge nodes and advanced packaging bottlenecks that dominate high-end GPU supply (HBM stacks, CoWoS-type packaging constraints). If accurate, this is strategically meaningful because it suggests Groq capacity expansion could be gated more by conventional wafer supply, board assembly, and data center power than by the same HBM/advanced packaging scarcity that has constrained top-tier GPU ramp cycles. On systems and cloud, Groq’s own releases identify colocation and connectivity partners (Equinix, DataBank, Bell Canada) and a Middle East partner (HUMAIN), implying dependencies on data center real estate, power availability, and network connectivity, alongside procurement of standard server components, NICs/switching, racks, and cooling infrastructure. The Groq design narrative also emphasizes air cooling and reduced need for complex power/cooling infrastructure, which—if realized in deployments—can widen the set of feasible hosting locations and lower deployment friction relative to liquid-cooled, very high power density GPU racks. Against that backdrop, the strategic rationale for NVIDIA acquiring Groq can be framed as a set of overlapping objectives: inference silicon optionality, architectural hedging, competitive defense, and supply chain diversification, with the carve-out of GroqCloud signaling a preference to avoid direct cloud competition and to focus on IP and product portfolio control rather than operating a capital-intensive token-serving business. The deal, if confirmed, would occur at a valuation step-up of ~190% versus Groq’s reported ~$6.9B private valuation in the September $750M round, reinforcing that any acquisition logic would be predominantly strategic rather than a conventional financial multiple arbitrage. The most compelling strategic driver is inference. Training has historically been the center of gravity for cutting-edge GPU demand, but inference volume is structurally larger and more distributed as deployments scale, with economics dominated by cost-per-token, latency guarantees, and utilization under spiky demand. Inference workloads also create a strategic vulnerability for NVIDIA: hyperscalers and large platforms can justify bespoke ASICs (TPU, Trainium/Inferentia, Maia-class efforts) because inference is stable, repeatable, and can amortize software investment at massive scale. Groq’s core proposition—deterministic, compiler-scheduled inference with predictable latency—aligns directly with the segment where GPU generality is least valued and where “good enough” programmability plus superior unit economics can win share. Acquiring Groq would allow NVIDIA to own a credible inference-native architecture rather than relying solely on GPUs and software optimization to defend that segment. Competitive defense logic is also plausible. Groq occupies a specific competitive wedge: low-latency, high-throughput interactive inference, delivered via a simple API abstraction that reduces switching cost. That wedge directly pressures GPU inference margins in the long run because it makes inference price/performance comparisons more transparent at the token level, and it targets a developer persona that historically defaulted to CUDA-first ecosystems. Even if NVIDIA’s current-generation systems can achieve very high tokens/s per user with extensive optimization, the strategic risk is that competing architectures normalize the idea that inference is best served by special-purpose silicon with a simpler programming model, weakening CUDA lock-in at the application layer. NVIDIA has actively demonstrated that Blackwell-era systems can exceed 1,000 tokens/s per user in benchmarked configurations, but that performance leadership does not automatically translate to lowest cost-per-token across the full range of batch sizes, latency targets, and deployment environments. Groq’s existence as a credible alternative architecture forces NVIDIA to keep defending inference economics rather than only raw performance leadership. The “technology acquisition” rationale is unusually strong in this specific case because Groq’s differentiator is not a single block of silicon IP but an end-to-end methodology: compiler-led static scheduling, deterministic networking, and a system architecture designed around tensor-parallel inference rather than throughput-maximizing batch inference. NVIDIA’s stack is already compiler-heavy (TensorRT, Triton, CUDA graphs, kernel fusion, speculative decoding techniques), but GPUs remain dynamically scheduled devices with complex memory hierarchies and stochastic latency behaviors under contention. Groq’s approach provides an alternate design point: treating the entire inference execution (compute plus communication) as a statically schedulable program. In principle, that IP could be valuable even if Groq silicon itself is not adopted at massive scale, because it can inform how NVIDIA builds future inference-optimized products, compilers, and networking fabrics, especially as distributed inference with large models makes communication a first-order performance determinant. Supply chain diversification is a non-obvious but potentially important driver. If Groq’s mainstream product generation is truly based on a mature process node and avoids HBM, then the scaling constraints look different than those of state-of-the-art GPUs. NVIDIA’s ability to meet incremental demand has been tightly coupled to advanced packaging and HBM supply, and those constraints can remain binding even when wafer supply is available. An inference ASIC architecture that relies primarily on on-chip SRAM and scales by adding chips—while not costless—could reduce dependence on HBM availability and advanced packaging capacity, enabling NVIDIA to ship “inference capacity” in higher absolute volumes or into geographies and customer segments where the highest-end GPUs are economically or logistically difficult to deploy. This could be particularly relevant for latency-sensitive inference deployed in regional colocation footprints rather than centralized hyperscale campuses. The carve-out of GroqCloud, if accurate, is itself a strategic signal about NVIDIA’s priorities. Operating a token-serving cloud at scale is capital intensive, structurally lower margin than silicon IP rents, and creates channel conflict with hyperscalers and CSP partners who are core NVIDIA customers. NVIDIA has generally positioned its cloud offerings through partnerships rather than as a direct hyperscale competitor. Excluding GroqCloud would preserve neutrality with CSPs and avoid inheriting multi-region data residency obligations and partner contracts, while still allowing NVIDIA to acquire Groq’s silicon, compiler technology, and engineering talent. At the same time, excluding GroqCloud would also mean NVIDIA would not automatically acquire the commercial proof-point of Groq’s unit economics or the customer contracts that validate product-market fit at scale, increasing the importance of diligence on whether Groq’s cloud pricing is structurally profitable or partially subsidized by fundraising. There is also a “preemptive acquisition” angle. The reporting identifies recent investors in Groq’s latest round including large financial institutions and strategic/industry players. In that context, Groq represents an asset that could plausibly have been acquired by a competitor (AMD/Intel) or by a hyperscaler seeking to accelerate inference independence. NVIDIA acquiring Groq could be a defensive move to prevent a credible inference-native architecture from being weaponized by a rival with deep distribution. Even if GroqCloud is carved out, controlling the silicon roadmap and compiler IP would meaningfully constrain Groq’s ability to evolve into a standalone competitor, unless the carved-out entity retains long-term rights to the hardware and software stack. However, the strategic case is not one-sided; there are meaningful risks and potential contradictions that would need to be reconciled for the transaction to be value-accretive on a multi-year horizon. 1st, Groq’s architecture appears to rely on scaling out chip count to achieve capacity, which introduces system cost, networking complexity, and physical footprint considerations. The absence of external memory and limited on-chip SRAM implies very large models require substantial chip parallelism, and the economics then depend heavily on chip cost, yield, power efficiency, and interconnect overhead. SemiAnalysis explicitly frames Groq as trading space for time and raises questions about token economics and whether publicly advertised pricing reflects fully loaded costs or market share capture. 2nd, integration risk is non-trivial. Groq’s compiler-led deterministic model is philosophically and practically different from CUDA’s dominant programming and execution model. A poorly executed integration could create internal product confusion, dilute engineering focus, or alienate developers if the combined stack fragments. 3rd, there is cannibalization risk. If Groq-class inference silicon undercuts GPU inference economics, NVIDIA could face internal margin trade-offs, even if the goal is to defend share against hyperscaler ASICs. Cannibalization can still be rational if it prevents larger share loss, but it would require crisp portfolio segmentation and go-to-market discipline. The presence of NVIDIA’s own rapidly improving inference performance complicates the “need” for Groq but does not eliminate the “option value.” NVIDIA has demonstrated benchmark-leading tokens/s per user on Blackwell-based systems, suggesting that raw interactive throughput is not necessarily the limiting factor for NVIDIA’s product line. The more enduring strategic question is unit economics and architectural control: whether future inference demand is better monetized through general-purpose GPUs plus software optimization, or whether a bifurcated product portfolio (training GPUs plus inference-native ASICs) becomes necessary to defend total AI compute wallet share as hyperscaler ASIC penetration increases. Acquiring Groq could be a decisive move to ensure NVIDIA participates in both regimes rather than betting exclusively on GPUs to win inference forever. What is “special” about Groq’s technology relative to a typical accelerator roadmap is the tight coupling of determinism, compilation, and networking into a single scheduling problem. The LPU narrative emphasizes deterministic compute and networking, static scheduling, and direct chip-to-chip coordination that allows “hundreds” (more precisely, 100s) of chips to behave like a single scheduled resource. The architecture also explicitly targets tensor-parallel, latency-optimized distribution rather than pure data-parallel throughput scaling, which matters for real-time applications where a single response must arrive quickly rather than many requests being processed in bulk. The implication is that Groq is optimized for the time-to-first-token and steady token streaming behavior that defines user experience in interactive LLMs, and it attempts to achieve that without relying on large batch sizes that can degrade latency. From a portfolio manager’s perspective, the most important interpretation is that an NVIDIA-Groq combination would likely be less about “NVIDIA needs more inference speed” and more about controlling the architectural trajectory of inference acceleration and removing a fast-improving, developer-friendly competitor from the market. The carve-out of GroqCloud would reinforce that the transaction is aimed at IP, talent, and product optionality, not acquiring a cloud revenue stream. The valuation step-up implied by $20B versus $6.9B would therefore be justified only if the acquired assets materially reduce long-term competitive risk (hyperscaler ASIC displacement, inference margin compression) or enable new monetization vectors (inference ASIC product line, supply chain de-bottlenecking, improved software determinism) that would be difficult to achieve on a comparable timeline via internal R&D.

$NVDA $GFS NVIDIA’s reported agreement to acquire Groq for $20B in cash (per CNBC, amplified via Reuters and other wire coverage) represents a materially different strategic posture than NVIDIA’s prior M&A pattern, given both the headline size (largest reported NVIDIA acquisition to date) and the unusual carve-out that Groq’s early-stage cloud business would not be included. Public reporting indicates the information originated from Alex Davis, CEO of Disruptive (lead investor in Groq’s latest financing), and that neither NVIDIA nor Groq had issued an immediate confirmation at the time of publication. The same reporting frames the transaction as coming together quickly, only months after Groq raised $750M at a ~$6.9B valuation, and highlights Groq’s positioning as a high-performance inference chip vendor founded by ex-Google TPU engineers. Groq is best understood as a vertically integrated inference acceleration company whose core asset is an application-specific processor optimized for deterministic, low-latency execution of transformer-style workloads, paired with a compiler-led software stack and a distribution layer (GroqCloud) designed to reduce developer friction via OpenAI-compatible APIs and integrations. Groq brands its architecture as a Language Processing Unit (LPU) and consistently emphasizes that the design target is inference, not training. The company’s own architecture description centers on 1-core execution, large on-chip SRAM used as primary storage (explicitly not cache), a custom compiler that statically schedules compute and communication, and direct chip-to-chip connectivity intended to coordinate multi-chip execution without relying on conventional caching hierarchies or dynamic runtime scheduling. The technical premise is a deliberate inversion of the conventional GPU approach. GPUs deliver throughput via massively parallel, multi-core execution with dynamic scheduling, complex memory hierarchies, and heavy reliance on off-chip HBM bandwidth and sophisticated runtime/kernel optimization. Groq instead argues that inference bottlenecks are driven by latency variance (tail latency), synchronization overhead, and memory access unpredictability inherent in dynamically scheduled, cache-heavy architectures, particularly when workloads are latency sensitive and batch sizes cannot be inflated. Groq’s solution is to move “control” into the compiler: the full execution graph and inter-chip communication schedule are computed ahead of time down to clock-cycle granularity, with deterministic execution designed to reduce run-to-run variance. In Groq’s framing, the removal of caches, reorder buffers, speculative execution overhead, and other sources of contention enables predictable latency and high utilization without per-model kernel engineering typical of GPU tuning cycles. A critical nuance is that Groq’s determinism is not merely a software claim; it is tightly coupled to architectural constraints and system design choices that trade flexibility for predictability. Third-party technical commentary indicates Groq’s chip uses a fully deterministic VLIW-style approach with minimal buffering, no external memory, and heavy dependence on sharding models across many chips because on-chip SRAM capacity is limited. SemiAnalysis describes a ~725 mm^2 die on GlobalFoundries 14nm with ~230MB of SRAM and notes that “no useful models” fit on a single chip, forcing multi-chip partitioning for modern LLMs and driving a system-level design where networking and compilation are first-class scheduling problems rather than ancillary infrastructure. This is consistent with Groq’s own messaging that tensor parallelism across chips is a primary design goal, enabled by large on-chip SRAM and compile-time coordination of compute plus interconnect. The on-chip SRAM emphasis is central to Groq’s latency story and also its most constraining trade-off. Groq claims on-chip SRAM bandwidth “upwards of 80 TB/s” and contrasts that with off-chip HBM bandwidth “about 8 TB/s,” asserting a potential 10x advantage from bandwidth plus reduced trips across chip-to-memory boundaries. While these comparisons are marketing-oriented and depend on workload specifics, the architectural implication is clear: Groq prioritizes ultra-fast local weight/activation access and then scales capacity by adding chips, not by attaching large off-chip memory pools. This design can reduce latency for sequential inference layers and minimize unpredictable stalls, but it pushes complexity into partitioning strategy, interconnect topology, and compiler scheduling, and it increases the number of chips needed for very large parameter counts and large KV-cache footprints. Groq also highlights numeric formats and compiler-driven precision management as a performance lever. In its 2025 technical blog, Groq describes “TruePoint numerics,” including 100-bit intermediate accumulation and selective quantization choices (FP32 for attention-sensitive operations, block floating point for MoE weights, FP8 storage in error-tolerant layers), and claims 2-4x speedups versus BF16 without measurable accuracy degradation on benchmarks such as MMLU and HumanEval. Even if the absolute uplift is workload dependent, the strategic point is that Groq is pursuing performance via end-to-end co-design: precision policy is not just hardware capability (FP8/BF16) but compiler-enforced mapping of precision to error sensitivity, which can matter materially for inference cost-per-token if it reduces memory traffic and boosts throughput without forcing aggressive, accuracy-damaging quantization. Independent performance datapoints indicate Groq has been credible on latency-oriented inference speed, at least for certain regimes. EE Times reported in 2023 that Groq demonstrated Llama-2 70B inference at ~240 tokens/s per user on a cloud-based dev system described as 10 racks and 64 chips, using the company’s 1st-gen silicon introduced several years earlier. Separate Groq commentary around independent benchmarking cites results showing ~241 tokens/s throughput and ~0.8s time to receive 100 output tokens for a Llama-2 70B API configuration, positioning the platform as a step-change in “available speed” for certain interactive use cases. These figures do not settle total cost-of-ownership versus GPUs or hyperscaler ASICs, but they establish that Groq’s system-level architecture can deliver strong single-user throughput and latency on large models when properly partitioned and scheduled. GroqCloud is the commercial wrapper that packages this hardware/software stack as “tokens-as-a-service,” aiming to make Groq adoption feel like switching API endpoints rather than adopting new silicon. Groq’s documentation states its API is designed to be “mostly compatible” with OpenAI client libraries, and its pricing page provides model-specific token rates, published speeds (tokens/s), prompt caching discounts, and batch processing discounts. For example, pricing lists inputs as low as $0.05 per 1M tokens and outputs as low as $0.08 per 1M tokens for certain smaller LLM configurations, with higher prices for larger models and long-context or MoE variants; it also advertises prompt caching with a 50% discount on cached input tokens for certain models and a batch API offering 50% lower cost for asynchronous processing windows. These mechanics are economically important because they demonstrate Groq’s go-to-market is not simply “sell chips,” but “sell predictable unit economics per token,” with tooling (batch, caching) that directly targets inference cost drivers (reused prompts, throughput smoothing, and asynchronous workloads). The cloud footprint and distribution partnerships indicate Groq has been building an inference-native “edge within the cloud” strategy rather than competing head-on with hyperscalers on breadth of services. A 2025 Groq newsroom release describes a European deployment in Helsinki with Equinix, positioned as latency reduction and data governance for European customers, and explicitly references Equinix Fabric enabling private connectivity to GroqCloud over public, private, or sovereign infrastructure. The same release enumerates additional capacity in the U.S. (Equinix, DataBank), Canada (Bell Canada), and Saudi Arabia (HUMAIN), and states these sites collectively served more than 20M tokens/s across Groq’s global network at that time. That supply-side metric matters because it provides a directional sense that Groq is scaling capacity as a network, not merely as a chip vendor. Customer disclosure is inherently limited because Groq is private and many enterprise deployments are not public, but Groq’s marketing materials and partnerships provide signals about demand vectors. The company’s public website displays logos of large consumer and enterprise brands (e.g., Dropbox, Vercel, Chevron, Volkswagen, Canva, Robinhood, Riot Games, Workday, Ramp) and includes a published customer quote claiming a 7.41x chat speed increase and an 89% cost reduction after moving to GroqCloud, followed by a tripling of token consumption. While marketing claims should be treated as case-specific and not generalized, they indicate that Groq is targeting both AI-native developers (who measure success by latency and cost-per-token) and enterprise buyers (who care about predictable performance and governance). Supplier and dependency mapping for Groq spans 3 layers: silicon production, system integration, and cloud infrastructure. On silicon, third-party analysis indicates GlobalFoundries 14nm for the 1st-gen Groq chip, implying a supply chain less constrained by the most capacity-tight leading-edge nodes and advanced packaging bottlenecks that dominate high-end GPU supply (HBM stacks, CoWoS-type packaging constraints). If accurate, this is strategically meaningful because it suggests Groq capacity expansion could be gated more by conventional wafer supply, board assembly, and data center power than by the same HBM/advanced packaging scarcity that has constrained top-tier GPU ramp cycles. On systems and cloud, Groq’s own releases identify colocation and connectivity partners (Equinix, DataBank, Bell Canada) and a Middle East partner (HUMAIN), implying dependencies on data center real estate, power availability, and network connectivity, alongside procurement of standard server components, NICs/switching, racks, and cooling infrastructure. The Groq design narrative also emphasizes air cooling and reduced need for complex power/cooling infrastructure, which—if realized in deployments—can widen the set of feasible hosting locations and lower deployment friction relative to liquid-cooled, very high power density GPU racks. Against that backdrop, the strategic rationale for NVIDIA acquiring Groq can be framed as a set of overlapping objectives: inference silicon optionality, architectural hedging, competitive defense, and supply chain diversification, with the carve-out of GroqCloud signaling a preference to avoid direct cloud competition and to focus on IP and product portfolio control rather than operating a capital-intensive token-serving business. The deal, if confirmed, would occur at a valuation step-up of ~190% versus Groq’s reported ~$6.9B private valuation in the September $750M round, reinforcing that any acquisition logic would be predominantly strategic rather than a conventional financial multiple arbitrage. The most compelling strategic driver is inference. Training has historically been the center of gravity for cutting-edge GPU demand, but inference volume is structurally larger and more distributed as deployments scale, with economics dominated by cost-per-token, latency guarantees, and utilization under spiky demand. Inference workloads also create a strategic vulnerability for NVIDIA: hyperscalers and large platforms can justify bespoke ASICs (TPU, Trainium/Inferentia, Maia-class efforts) because inference is stable, repeatable, and can amortize software investment at massive scale. Groq’s core proposition—deterministic, compiler-scheduled inference with predictable latency—aligns directly with the segment where GPU generality is least valued and where “good enough” programmability plus superior unit economics can win share. Acquiring Groq would allow NVIDIA to own a credible inference-native architecture rather than relying solely on GPUs and software optimization to defend that segment. Competitive defense logic is also plausible. Groq occupies a specific competitive wedge: low-latency, high-throughput interactive inference, delivered via a simple API abstraction that reduces switching cost. That wedge directly pressures GPU inference margins in the long run because it makes inference price/performance comparisons more transparent at the token level, and it targets a developer persona that historically defaulted to CUDA-first ecosystems. Even if NVIDIA’s current-generation systems can achieve very high tokens/s per user with extensive optimization, the strategic risk is that competing architectures normalize the idea that inference is best served by special-purpose silicon with a simpler programming model, weakening CUDA lock-in at the application layer. NVIDIA has actively demonstrated that Blackwell-era systems can exceed 1,000 tokens/s per user in benchmarked configurations, but that performance leadership does not automatically translate to lowest cost-per-token across the full range of batch sizes, latency targets, and deployment environments. Groq’s existence as a credible alternative architecture forces NVIDIA to keep defending inference economics rather than only raw performance leadership. The “technology acquisition” rationale is unusually strong in this specific case because Groq’s differentiator is not a single block of silicon IP but an end-to-end methodology: compiler-led static scheduling, deterministic networking, and a system architecture designed around tensor-parallel inference rather than throughput-maximizing batch inference. NVIDIA’s stack is already compiler-heavy (TensorRT, Triton, CUDA graphs, kernel fusion, speculative decoding techniques), but GPUs remain dynamically scheduled devices with complex memory hierarchies and stochastic latency behaviors under contention. Groq’s approach provides an alternate design point: treating the entire inference execution (compute plus communication) as a statically schedulable program. In principle, that IP could be valuable even if Groq silicon itself is not adopted at massive scale, because it can inform how NVIDIA builds future inference-optimized products, compilers, and networking fabrics, especially as distributed inference with large models makes communication a first-order performance determinant. Supply chain diversification is a non-obvious but potentially important driver. If Groq’s mainstream product generation is truly based on a mature process node and avoids HBM, then the scaling constraints look different than those of state-of-the-art GPUs. NVIDIA’s ability to meet incremental demand has been tightly coupled to advanced packaging and HBM supply, and those constraints can remain binding even when wafer supply is available. An inference ASIC architecture that relies primarily on on-chip SRAM and scales by adding chips—while not costless—could reduce dependence on HBM availability and advanced packaging capacity, enabling NVIDIA to ship “inference capacity” in higher absolute volumes or into geographies and customer segments where the highest-end GPUs are economically or logistically difficult to deploy. This could be particularly relevant for latency-sensitive inference deployed in regional colocation footprints rather than centralized hyperscale campuses. The carve-out of GroqCloud, if accurate, is itself a strategic signal about NVIDIA’s priorities. Operating a token-serving cloud at scale is capital intensive, structurally lower margin than silicon IP rents, and creates channel conflict with hyperscalers and CSP partners who are core NVIDIA customers. NVIDIA has generally positioned its cloud offerings through partnerships rather than as a direct hyperscale competitor. Excluding GroqCloud would preserve neutrality with CSPs and avoid inheriting multi-region data residency obligations and partner contracts, while still allowing NVIDIA to acquire Groq’s silicon, compiler technology, and engineering talent. At the same time, excluding GroqCloud would also mean NVIDIA would not automatically acquire the commercial proof-point of Groq’s unit economics or the customer contracts that validate product-market fit at scale, increasing the importance of diligence on whether Groq’s cloud pricing is structurally profitable or partially subsidized by fundraising. There is also a “preemptive acquisition” angle. The reporting identifies recent investors in Groq’s latest round including large financial institutions and strategic/industry players. In that context, Groq represents an asset that could plausibly have been acquired by a competitor (AMD/Intel) or by a hyperscaler seeking to accelerate inference independence. NVIDIA acquiring Groq could be a defensive move to prevent a credible inference-native architecture from being weaponized by a rival with deep distribution. Even if GroqCloud is carved out, controlling the silicon roadmap and compiler IP would meaningfully constrain Groq’s ability to evolve into a standalone competitor, unless the carved-out entity retains long-term rights to the hardware and software stack. However, the strategic case is not one-sided; there are meaningful risks and potential contradictions that would need to be reconciled for the transaction to be value-accretive on a multi-year horizon. 1st, Groq’s architecture appears to rely on scaling out chip count to achieve capacity, which introduces system cost, networking complexity, and physical footprint considerations. The absence of external memory and limited on-chip SRAM implies very large models require substantial chip parallelism, and the economics then depend heavily on chip cost, yield, power efficiency, and interconnect overhead. SemiAnalysis explicitly frames Groq as trading space for time and raises questions about token economics and whether publicly advertised pricing reflects fully loaded costs or market share capture. 2nd, integration risk is non-trivial. Groq’s compiler-led deterministic model is philosophically and practically different from CUDA’s dominant programming and execution model. A poorly executed integration could create internal product confusion, dilute engineering focus, or alienate developers if the combined stack fragments. 3rd, there is cannibalization risk. If Groq-class inference silicon undercuts GPU inference economics, NVIDIA could face internal margin trade-offs, even if the goal is to defend share against hyperscaler ASICs. Cannibalization can still be rational if it prevents larger share loss, but it would require crisp portfolio segmentation and go-to-market discipline. The presence of NVIDIA’s own rapidly improving inference performance complicates the “need” for Groq but does not eliminate the “option value.” NVIDIA has demonstrated benchmark-leading tokens/s per user on Blackwell-based systems, suggesting that raw interactive throughput is not necessarily the limiting factor for NVIDIA’s product line. The more enduring strategic question is unit economics and architectural control: whether future inference demand is better monetized through general-purpose GPUs plus software optimization, or whether a bifurcated product portfolio (training GPUs plus inference-native ASICs) becomes necessary to defend total AI compute wallet share as hyperscaler ASIC penetration increases. Acquiring Groq could be a decisive move to ensure NVIDIA participates in both regimes rather than betting exclusively on GPUs to win inference forever. What is “special” about Groq’s technology relative to a typical accelerator roadmap is the tight coupling of determinism, compilation, and networking into a single scheduling problem. The LPU narrative emphasizes deterministic compute and networking, static scheduling, and direct chip-to-chip coordination that allows “hundreds” (more precisely, 100s) of chips to behave like a single scheduled resource. The architecture also explicitly targets tensor-parallel, latency-optimized distribution rather than pure data-parallel throughput scaling, which matters for real-time applications where a single response must arrive quickly rather than many requests being processed in bulk. The implication is that Groq is optimized for the time-to-first-token and steady token streaming behavior that defines user experience in interactive LLMs, and it attempts to achieve that without relying on large batch sizes that can degrade latency. From a portfolio manager’s perspective, the most important interpretation is that an NVIDIA-Groq combination would likely be less about “NVIDIA needs more inference speed” and more about controlling the architectural trajectory of inference acceleration and removing a fast-improving, developer-friendly competitor from the market. The carve-out of GroqCloud would reinforce that the transaction is aimed at IP, talent, and product optionality, not acquiring a cloud revenue stream. The valuation step-up implied by $20B versus $6.9B would therefore be justified only if the acquired assets materially reduce long-term competitive risk (hyperscaler ASIC displacement, inference margin compression) or enable new monetization vectors (inference ASIC product line, supply chain de-bottlenecking, improved software determinism) that would be difficult to achieve on a comparable timeline via internal R&D.

TheValueist

102,145 次观看 • 8 个月前