Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

Running a single deep coding model at max context on Cerebras requires 24 systems ($24M Capex) just to support 256 concurrent users. At that scale, $100M gets you way more memory bandwidth in standard GB300 racks.

SemiAnalysis

109,405 subscribers

93,448 Aufrufe • vor 5 Monaten •via X (Twitter)

Bildung Wissenschaft & Technologie

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

How we deploy multi-trillion parameter models at scale Today, let's break down the math for how Cerebras: 1. Fits the full model into on-chip memory 2. Maps the model across multiple chips 3. Handles a high # of concurrent requests

How we deploy multi-trillion parameter models at scale Today, let's break down the math for how Cerebras: 1. Fits the full model into on-chip memory 2. Maps the model across multiple chips 3. Handles a high # of concurrent requests

Sean Lie

251,367 Aufrufe • vor 1 Tag

$MU $SKHY $DRAM Holy Shit $AMD's Lisa Su just spent two full minutes at the 2026 AI keynote raving about HBM4 capacity as the thing that separates Helios from the competition. Not compute. Memory. "When you compare Helios to the competition, we're delivering 15% more compute, 50% more HBM4 memory capacity and memory bandwidth, and 50% more scale-out bandwidth." Notice which number is bigger. "What that means is that every Helios can deliver more performance for the largest models, more capacity for longer context, and the bandwidth to scale across thousands of racks.... More than 18,000 CDNA five GPU compute units, over 4,600 Zen 6 CPU cores, and 31 terabytes of HBM4 memory, all in a single rack." "That's what it takes to run agentic AI at scale. Today, I'm excited to announce that Helios is in full production. We have shipments on track to start at the end of the third quarter and ramping into the fourth quarter in the second half of the year. I can tell you, customer demand for Helios is extremely strong, we're extremely proud of the work that we have done across the leading AI labs to adopt Helios."

$MU $SKHY $DRAM Holy Shit $AMD's Lisa Su just spent two full minutes at the 2026 AI keynote raving about HBM4 capacity as the thing that separates Helios from the competition. Not compute. Memory. "When you compare Helios to the competition, we're delivering 15% more compute, 50% more HBM4 memory capacity and memory bandwidth, and 50% more scale-out bandwidth." Notice which number is bigger. "What that means is that every Helios can deliver more performance for the largest models, more capacity for longer context, and the bandwidth to scale across thousands of racks.... More than 18,000 CDNA five GPU compute units, over 4,600 Zen 6 CPU cores, and 31 terabytes of HBM4 memory, all in a single rack." "That's what it takes to run agentic AI at scale. Today, I'm excited to announce that Helios is in full production. We have shipments on track to start at the end of the third quarter and ramping into the fourth quarter in the second half of the year. I can tell you, customer demand for Helios is extremely strong, we're extremely proud of the work that we have done across the leading AI labs to adopt Helios."

Trade Whisperer

180,543 Aufrufe • vor 3 Monaten

Cerebras Code: 20x faster than Claude, 1x the price Today we are launching two monthly coding plans: ➡️Cerebras Code Pro: $50/m – for indie developers ➡️Cerebras Code Max: $200/m – for power users with 5x rate limits Both plans get: Qwen3-Coder at 2,000 tokens/s, 131K context, and no weekly limits. Sign up now:

Cerebras Code: 20x faster than Claude, 1x the price Today we are launching two monthly coding plans: ➡️Cerebras Code Pro: $50/m – for indie developers ➡️Cerebras Code Max: $200/m – for power users with 5x rate limits Both plans get: Qwen3-Coder at 2,000 tokens/s, 131K context, and no weekly limits. Sign up now:

Cerebras

461,574 Aufrufe • vor 1 Jahr

New course on serving LLMs efficiently -- how do you serve models to many concurrent users at low latency and reasonable cost? This short course is built with Red Hat and taught by Cedric Clyburn. Efficient LLM serving requires efficient memory management. A 70B-parameter model takes ~140 GB just to load the weights. On top of that, every active request needs its own chunk of GPU memory, the KV cache, to store the token context it has built up so far. In this course, you'll learn to reduce a model's memory footprint with quantization and serve it using vLLM, which handles many concurrent requests efficiently through smart memory management. Skills you'll gain: - Quantize a model and measure the accuracy tradeoff - Serve a model with vLLM and watch it handle concurrent requests efficiently - Benchmark your deployment and make informed tradeoffs between speed, cost, and accuracy Join and learn to serve LLMs efficiently:

New course on serving LLMs efficiently -- how do you serve models to many concurrent users at low latency and reasonable cost? This short course is built with Red Hat and taught by Cedric Clyburn. Efficient LLM serving requires efficient memory management. A 70B-parameter model takes ~140 GB just to load the weights. On top of that, every active request needs its own chunk of GPU memory, the KV cache, to store the token context it has built up so far. In this course, you'll learn to reduce a model's memory footprint with quantization and serve it using vLLM, which handles many concurrent requests efficiently through smart memory management. Skills you'll gain: - Quantize a model and measure the accuracy tradeoff - Serve a model with vLLM and watch it handle concurrent requests efficiently - Benchmark your deployment and make informed tradeoffs between speed, cost, and accuracy Join and learn to serve LLMs efficiently:

Andrew Ng

136,970 Aufrufe • vor 4 Monaten

DS4 running on DGX Spark (GB10 / CUDA), private branch for now. 12 tokens/sec, the memory bandwidth is limited in this system, at 270GB/sec. But prefill is ways more alighed to M3 Max at ~200 t/s. I'll release when more mature, but it is almost sure that it will get merged.

DS4 running on DGX Spark (GB10 / CUDA), private branch for now. 12 tokens/sec, the memory bandwidth is limited in this system, at 270GB/sec. But prefill is ways more alighed to M3 Max at ~200 t/s. I'll release when more mature, but it is almost sure that it will get merged.

antirez

84,477 Aufrufe • vor 5 Monaten

Yall seen Qwen3.8 27B single-stream decode, but someone asked how far I can push concurrency on 1x6000, so I did then scale it up x 4. 8 x 70sec back-to-back cycle of 256 concurrent users PER 6000. You are watching the final cycle, running hottest I've ever seen these cards.

Yall seen Qwen3.8 27B single-stream decode, but someone asked how far I can push concurrency on 1x6000, so I did then scale it up x 4. 8 x 70sec back-to-back cycle of 256 concurrent users PER 6000. You are watching the final cycle, running hottest I've ever seen these cards.

Killy

20,896 Aufrufe • vor 2 Monaten

🎉 Day-0 support for in vLLM, available today in v0.23.0! Congrats to Z.ai on GLM-5.2, a flagship model built for long-horizon coding agents. ✨ 1M-token context, built to hold project-scale engineering work in a single run ✨ Tuned for long-horizon coding: large-scale implementation, automated research, and performance optimization ✨ One task can carry a full dev workflow, from requirements to a deployable product across platforms ✨ Client-side and mobile engineering, including an on-device debugging loop Try it out running it on vLLM today: 🔗

🎉 Day-0 support for in vLLM, available today in v0.23.0! Congrats to Z.ai on GLM-5.2, a flagship model built for long-horizon coding agents. ✨ 1M-token context, built to hold project-scale engineering work in a single run ✨ Tuned for long-horizon coding: large-scale implementation, automated research, and performance optimization ✨ One task can carry a full dev workflow, from requirements to a deployable product across platforms ✨ Client-side and mobile engineering, including an on-device debugging loop Try it out running it on vLLM today: 🔗

vLLM

36,413 Aufrufe • vor 4 Monaten

Etched is deploying two new technologies in chip design: low-voltage inference and cluster-scale memory. CEO Gavin Uberti says they'll make their chips much more power-efficient and way, way faster than today's leading GPUs. He breaks it down: "We looked at a lot of early research directions, and we realized the key things that models need are way more compute and way faster memory." "If you think about inference, there are two key parts: prefill and decode. For prefill, it's a compute-bound problem. You need to have more FLOPS, more operations per second on each of your chips." "On our GPU, the bottleneck's actually thermals. You can't really run a GPU at more than around 50% of what it could theoretically do, or it'll melt." "So we're using a new technology today called low-voltage inference to try to solve this problem. You bring the voltage of the chip down dramatically, which allows us to have way, way better efficiency in terms of how much power is drawn per unit of math, and thus fit way way more flops onto the chip..." "For decode, it's all about bandwidth. Not just bandwidth on a chip, but bandwidth across your cluster. That's why we have this technology we call cluster-scale memory. It reduces the amount of time it takes to communicate from one chip to another dramatically." "As a result we can go use all of our HBM, HBM bandwidth, SRAM, SRAM bandwidth, and our scale-up domain as a single coherent pool. And that means if you're a user, you can go get much faster tokens per second, while still keeping your costs low."

Etched is deploying two new technologies in chip design: low-voltage inference and cluster-scale memory. CEO Gavin Uberti says they'll make their chips much more power-efficient and way, way faster than today's leading GPUs. He breaks it down: "We looked at a lot of early research directions, and we realized the key things that models need are way more compute and way faster memory." "If you think about inference, there are two key parts: prefill and decode. For prefill, it's a compute-bound problem. You need to have more FLOPS, more operations per second on each of your chips." "On our GPU, the bottleneck's actually thermals. You can't really run a GPU at more than around 50% of what it could theoretically do, or it'll melt." "So we're using a new technology today called low-voltage inference to try to solve this problem. You bring the voltage of the chip down dramatically, which allows us to have way, way better efficiency in terms of how much power is drawn per unit of math, and thus fit way way more flops onto the chip..." "For decode, it's all about bandwidth. Not just bandwidth on a chip, but bandwidth across your cluster. That's why we have this technology we call cluster-scale memory. It reduces the amount of time it takes to communicate from one chip to another dramatically." "As a result we can go use all of our HBM, HBM bandwidth, SRAM, SRAM bandwidth, and our scale-up domain as a single coherent pool. And that means if you're a user, you can go get much faster tokens per second, while still keeping your costs low."

TBPN

20,404 Aufrufe • vor 4 Monaten

$200/month gets you the Ferrari of vibe coding. Dozens of tool calls, instantly handled with 1.5M TPM on Cerebras Code Max available now!

$200/month gets you the Ferrari of vibe coding. Dozens of tool calls, instantly handled with 1.5M TPM on Cerebras Code Max available now!

Daniel Kim

186,378 Aufrufe • vor 1 Jahr

Here's a look at GPT-5.6 Sol Ultrafast running on Cerebras, side by side with regular Sol. OpenAI has said Sol on Cerebras launches this month for select customers. h/t Max

Here's a look at GPT-5.6 Sol Ultrafast running on Cerebras, side by side with regular Sol. OpenAI has said Sol on Cerebras launches this month for select customers. h/t Max

M1

275,696 Aufrufe • vor 3 Monaten

"There's just nothing better than that. That’s baseball at its pinnacle... That's just a lifelong memory. That's a baseball memory." Max Scherzer speaks on his time in Toronto.

"There's just nothing better than that. That’s baseball at its pinnacle... That's just a lifelong memory. That's a baseball memory." Max Scherzer speaks on his time in Toronto.

Sportsnet

92,938 Aufrufe • vor 1 Monat

Cerebras just IPO’d and the stock already ran up over 100% (Save this). For the entire 70 year history of the semiconductor industry, every company on earth has followed the same process. You take a dinner plate sized silicon wafer, put hundreds of tiny chips onto it, and dice it up like a pizza. Nvidia does it this way, AMD does it this way, Intel has done it this way for six decades and everyone who tried to break that convention failed. Until Cerebras asked the most annoyingly obvious question in the industry’s history, what if you just didn’t cut it? The result is the Wafer Scale Engine, a single chip 56 times larger than Nvidia’s H100 and it fundamentally changes the physics of how AI inference works. The reason this matters is not the size, it’s the bandwidth. Every time an AI model generates a single word, it has to reach into memory, pull weights, multiply them together, and produce a prediction and when you’re running millions of concurrent sessions at once, the bottleneck is not raw processing power but how fast data moves between memory and compute. Nvidia’s H100 moves data at roughly 3 terabytes per second, while Cerebras’ WSE-3 moves data at 21 petabytes per second, roughly 7,000 times faster because memory and compute live on the same enormous piece of silicon and data barely has to travel at all. That gap is exactly why OpenAI went from 150 tokens per second on traditional GPUs to 2,000 tokens per second on Cerebras hardware, and why AWS integrated Cerebras into Bedrock to deliver roughly 5x more inference capacity in the same physical footprint. The macro setup is making the trade even more urgent. South Korea DRAM export prices recently jumped 35%, flash memory surged 47%, and SSD pricing spiked nearly 140% and every single one of those increases hits Nvidia-based infrastructure directly, because the H100 requires 80GB of the most expensive, most contested memory in the AI supply chain. Cerebras’ WSE-3 uses zero external HBM memory, baking 44GB of SRAM directly into the wafer itself which means as memory pricing goes parabolic, every CFO evaluating AI infrastructure is suddenly looking much more seriously at the architecture that sidesteps that cost entirely. The demand is already showing up in the backlog. Cerebras ended 2025 with $24.6 billion in remaining performance obligations for a company doing just over $500 million in annual revenue, that is a number that implies years of contracted growth already sitting on the books. The IPO was 20x oversubscribed, the price range was raised twice before listing, and shares opened 89% above their listing price on a $5.55 billion raise that made it the largest semiconductor IPO in history. The risks are real and worth naming. 86% of 2025 revenue came from two entities with UAE ties, U.S. revenue actually fell 34% to $187 million, and the $20 billion OpenAI contract is conditional, if Cerebras misses delivery milestones, OpenAI can terminate and trigger repayment demands on a $1 billion loan facility. And yet the market is valuing Cerebras at roughly 91x trailing revenue, richer than Nvidia, AMD, and Arm combined. What investors are betting on is not that Cerebras beats Nvidia, it is that the inference supercycle is large enough to support an entirely different architecture optimized for a different workload, and that $24.6 billion in contracted backlog converts to diversified revenue before the market starts asking harder questions. CEO Andrew Feldman said this took a decade of late nights to get right, everyone who tried to copy it failed and given that the entire inference economy is now running through exactly the bottleneck Cerebras was built to eliminate, the market is starting to believe him.

Cerebras just IPO’d and the stock already ran up over 100% (Save this). For the entire 70 year history of the semiconductor industry, every company on earth has followed the same process. You take a dinner plate sized silicon wafer, put hundreds of tiny chips onto it, and dice it up like a pizza. Nvidia does it this way, AMD does it this way, Intel has done it this way for six decades and everyone who tried to break that convention failed. Until Cerebras asked the most annoyingly obvious question in the industry’s history, what if you just didn’t cut it? The result is the Wafer Scale Engine, a single chip 56 times larger than Nvidia’s H100 and it fundamentally changes the physics of how AI inference works. The reason this matters is not the size, it’s the bandwidth. Every time an AI model generates a single word, it has to reach into memory, pull weights, multiply them together, and produce a prediction and when you’re running millions of concurrent sessions at once, the bottleneck is not raw processing power but how fast data moves between memory and compute. Nvidia’s H100 moves data at roughly 3 terabytes per second, while Cerebras’ WSE-3 moves data at 21 petabytes per second, roughly 7,000 times faster because memory and compute live on the same enormous piece of silicon and data barely has to travel at all. That gap is exactly why OpenAI went from 150 tokens per second on traditional GPUs to 2,000 tokens per second on Cerebras hardware, and why AWS integrated Cerebras into Bedrock to deliver roughly 5x more inference capacity in the same physical footprint. The macro setup is making the trade even more urgent. South Korea DRAM export prices recently jumped 35%, flash memory surged 47%, and SSD pricing spiked nearly 140% and every single one of those increases hits Nvidia-based infrastructure directly, because the H100 requires 80GB of the most expensive, most contested memory in the AI supply chain. Cerebras’ WSE-3 uses zero external HBM memory, baking 44GB of SRAM directly into the wafer itself which means as memory pricing goes parabolic, every CFO evaluating AI infrastructure is suddenly looking much more seriously at the architecture that sidesteps that cost entirely. The demand is already showing up in the backlog. Cerebras ended 2025 with $24.6 billion in remaining performance obligations for a company doing just over $500 million in annual revenue, that is a number that implies years of contracted growth already sitting on the books. The IPO was 20x oversubscribed, the price range was raised twice before listing, and shares opened 89% above their listing price on a $5.55 billion raise that made it the largest semiconductor IPO in history. The risks are real and worth naming. 86% of 2025 revenue came from two entities with UAE ties, U.S. revenue actually fell 34% to $187 million, and the $20 billion OpenAI contract is conditional, if Cerebras misses delivery milestones, OpenAI can terminate and trigger repayment demands on a $1 billion loan facility. And yet the market is valuing Cerebras at roughly 91x trailing revenue, richer than Nvidia, AMD, and Arm combined. What investors are betting on is not that Cerebras beats Nvidia, it is that the inference supercycle is large enough to support an entirely different architecture optimized for a different workload, and that $24.6 billion in contracted backlog converts to diversified revenue before the market starts asking harder questions. CEO Andrew Feldman said this took a decade of late nights to get right, everyone who tried to copy it failed and given that the entire inference economy is now running through exactly the bottleneck Cerebras was built to eliminate, the market is starting to believe him.

Milk Road AI

30,441 Aufrufe • vor 5 Monaten

Qwen3.8-27B running at full BF16 on a free Kaggle TPU is kind of ridiculous. No quantization. No tiny context window. No expensive GPU instance. Just Qwen3.8-27B running on a Kaggle TPU v5e-8. The reported numbers: ~130 tok/s decode ~10,000 tok/s prefill 262K context That prefill number is especially wild. You can throw a huge amount of code or context at the model and ingest it extremely quickly, while still getting around 130 tokens per second during generation. And because it’s running in full BF16, you’re not relying on an aggressive quant just to make the model fit. But the really interesting part isn’t even the raw throughput. You can expose it as an OpenAI-compatible endpoint. That means you can plug the model into tools that already understand OpenAI-style APIs. Claude Code. Codex. OpenCode. And other compatible clients. So the workflow becomes pretty simple: Spin up the Qwen3.8-27B endpoint on Kaggle. Point your coding tool at the API. And suddenly you have a 27B coding model sitting behind the same interface you’d normally use for hosted models. The 262K context is also a huge deal for agentic coding. Large repositories can fit into a single context. Long conversations don’t need to be constantly trimmed. And tools can feed much more information back to the model without hitting a tiny context ceiling. The fact that this can be built around a free TPU environment is what makes this especially interesting. We’re getting to a point where experimenting with serious open models doesn’t always require owning a $2,000 GPU or paying for a large cloud instance. Free compute + open weights + an OpenAI-compatible API + existing coding agents. That’s a pretty powerful combination. Qwen3.8-27B is already an interesting model. Running the full BF16 version at ~130 tok/s with 262K context on free Kaggle TPU compute makes it a lot more interesting.

Qwen3.8-27B running at full BF16 on a free Kaggle TPU is kind of ridiculous. No quantization. No tiny context window. No expensive GPU instance. Just Qwen3.8-27B running on a Kaggle TPU v5e-8. The reported numbers: ~130 tok/s decode ~10,000 tok/s prefill 262K context That prefill number is especially wild. You can throw a huge amount of code or context at the model and ingest it extremely quickly, while still getting around 130 tokens per second during generation. And because it’s running in full BF16, you’re not relying on an aggressive quant just to make the model fit. But the really interesting part isn’t even the raw throughput. You can expose it as an OpenAI-compatible endpoint. That means you can plug the model into tools that already understand OpenAI-style APIs. Claude Code. Codex. OpenCode. And other compatible clients. So the workflow becomes pretty simple: Spin up the Qwen3.8-27B endpoint on Kaggle. Point your coding tool at the API. And suddenly you have a 27B coding model sitting behind the same interface you’d normally use for hosted models. The 262K context is also a huge deal for agentic coding. Large repositories can fit into a single context. Long conversations don’t need to be constantly trimmed. And tools can feed much more information back to the model without hitting a tiny context ceiling. The fact that this can be built around a free TPU environment is what makes this especially interesting. We’re getting to a point where experimenting with serious open models doesn’t always require owning a $2,000 GPU or paying for a large cloud instance. Free compute + open weights + an OpenAI-compatible API + existing coding agents. That’s a pretty powerful combination. Qwen3.8-27B is already an interesting model. Running the full BF16 version at ~130 tok/s with 262K context on free Kaggle TPU compute makes it a lot more interesting.

FHILY👑

36,136 Aufrufe • vor 1 Monat

AI inference is memory-bound more often than compute-bound. High Bandwidth Compute (HBC) flips the traditional model by moving compute to the data, helping unlock dramatic gains in memory bandwidth and efficiency. Learn more:

AI inference is memory-bound more often than compute-bound. High Bandwidth Compute (HBC) flips the traditional model by moving compute to the data, helping unlock dramatic gains in memory bandwidth and efficiency. Learn more:

Qualcomm

11,901 Aufrufe • vor 2 Monaten

“We coined this term called the semi-async valley of death.” At Cerebras Supernova, Silas Alberti unpacked why coding tasks that take 1 to 3 minutes break flow, and how Cerebras powers Cognition’s SWE-1.7 in Devin at up to 1,000 tokens per second to keep developers in the loop.

“We coined this term called the semi-async valley of death.” At Cerebras Supernova, Silas Alberti unpacked why coding tasks that take 1 to 3 minutes break flow, and how Cerebras powers Cognition’s SWE-1.7 in Devin at up to 1,000 tokens per second to keep developers in the loop.

Cerebras

89,992 Aufrufe • vor 1 Monat

Max Thomas 🇺🇸 running a PB of 9.90s (1.2) in his 100m season opener at the Florida Relays!

Max Thomas 🇺🇸 running a PB of 9.90s (1.2) in his 100m season opener at the Florida Relays!

Track & Field Gazette

29,538 Aufrufe • vor 6 Monaten

Pipe (Pipe Network) is a decentralized content delivery network (dCDN) that accelerates delivery by leveraging users’ excess storage & bandwidth. More nodes = faster delivery. More from 🤖DR 🇺🇸 at Breakpoint '24

Pipe (Pipe Network) is a decentralized content delivery network (dCDN) that accelerates delivery by leveraging users’ excess storage & bandwidth. More nodes = faster delivery. More from 🤖DR 🇺🇸 at Breakpoint '24

Solana

79,060 Aufrufe • vor 1 Jahr

With TRAE Max Mode you can now scale your context window from 200k → 1M for deep coding, longer reasoning, and complex projects. You can stay in control with full transparency on token usage, cost, and fast requests right after each conversation.

With TRAE Max Mode you can now scale your context window from 200k → 1M for deep coding, longer reasoning, and complex projects. You can stay in control with full transparency on token usage, cost, and fast requests right after each conversation.

TRAE

14,411 Aufrufe • vor 1 Jahr

Qwen3.8-Flash-Next is still going strong at 364.7K tokens of context on an M5 Max. And this isn’t just a static long-context test. The model was reasoning about how to speed up its own workflow while using tools, and the tool calls kept working without misses. Setup: • Qwen3.8-Flash-Next • M5 Max • 128GB unified memory • MLX-Serve PR #363 • OpenCode 2 • 364.7K context The interesting part isn’t simply getting hundreds of thousands of tokens into memory. It’s what happens once the context gets this large. Long-context inference usually comes with a painful tradeoff. As the KV cache grows, memory pressure increases and generation can slow down. But this setup is still pushing through 364K tokens while maintaining a usable agent workflow. The model can reason, call tools, inspect results, continue working, and keep the session moving. And the tool calls reportedly haven’t missed so far. That’s important for agentic coding. A huge context window is only useful if the model can actually operate reliably inside it. A 400K-token context that constantly breaks tool calls isn’t very useful. A 364K session that can keep reasoning and executing tools is a different story. And the test isn’t finished yet. The current run is approaching 400K tokens, with the expectation that it can keep going. This is also another interesting example of why Apple Silicon keeps showing up in local LLM experiments. The M5 Max’s unified memory gives a large model and its growing KV cache access to one shared memory pool. With MLX-Serve continuing to improve, these machines are becoming surprisingly capable long-context inference boxes. The bigger takeaway: Context length is becoming a workload, not just a model specification. Running a model at 256K is one thing. Keeping an agent alive at 300K+ while it reasons and uses tools is much more interesting. And Qwen3.8-Flash-Next is showing that this can be pushed surprisingly far on a single 128GB Mac. 364.7K and counting. Next stop: 400K.

Qwen3.8-Flash-Next is still going strong at 364.7K tokens of context on an M5 Max. And this isn’t just a static long-context test. The model was reasoning about how to speed up its own workflow while using tools, and the tool calls kept working without misses. Setup: • Qwen3.8-Flash-Next • M5 Max • 128GB unified memory • MLX-Serve PR #363 • OpenCode 2 • 364.7K context The interesting part isn’t simply getting hundreds of thousands of tokens into memory. It’s what happens once the context gets this large. Long-context inference usually comes with a painful tradeoff. As the KV cache grows, memory pressure increases and generation can slow down. But this setup is still pushing through 364K tokens while maintaining a usable agent workflow. The model can reason, call tools, inspect results, continue working, and keep the session moving. And the tool calls reportedly haven’t missed so far. That’s important for agentic coding. A huge context window is only useful if the model can actually operate reliably inside it. A 400K-token context that constantly breaks tool calls isn’t very useful. A 364K session that can keep reasoning and executing tools is a different story. And the test isn’t finished yet. The current run is approaching 400K tokens, with the expectation that it can keep going. This is also another interesting example of why Apple Silicon keeps showing up in local LLM experiments. The M5 Max’s unified memory gives a large model and its growing KV cache access to one shared memory pool. With MLX-Serve continuing to improve, these machines are becoming surprisingly capable long-context inference boxes. The bigger takeaway: Context length is becoming a workload, not just a model specification. Running a model at 256K is one thing. Keeping an agent alive at 300K+ while it reasons and uses tools is much more interesting. And Qwen3.8-Flash-Next is showing that this can be pushed surprisingly far on a single 128GB Mac. 364.7K and counting. Next stop: 400K.

FHILY👑

39,982 Aufrufe • vor 1 Monat