Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

No wonder Jensen got so defensive over $NVDA GPUs against custom chips. $GOOG will start using different custom chips for inference and training. $AMZN has already adopted this approach as they have Trainium for training and Inferentia for inference. Hyperscalers obviously see better price/performance in this approach, and this... show more

Oguz Erkan

117,402 subscribers

223,244 views • 4 months ago •via X (Twitter)

Science & Technology Finance

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

$AMZN CEO: “If you want to allow customers to use AI as expansively as they want, you need to bring down the cost of inference.” Hyperscalers are preparing for the age of inference. $AMZN has Trainium, $GOOG has TPU and $MSFT recently announced Maia 200. Once the bulk of AI workloads shifts from training to inference, they’ll depend less on $NVDA and margins will expand thanks to vertical integration. $AMZN is leading the race here as their chip business has already reached $10 billion annual run rate. Imagine being bearish on $AMZN.

$AMZN CEO: “If you want to allow customers to use AI as expansively as they want, you need to bring down the cost of inference.” Hyperscalers are preparing for the age of inference. $AMZN has Trainium, $GOOG has TPU and $MSFT recently announced Maia 200. Once the bulk of AI workloads shifts from training to inference, they’ll depend less on $NVDA and margins will expand thanks to vertical integration. $AMZN is leading the race here as their chip business has already reached $10 billion annual run rate. Imagine being bearish on $AMZN.

Oguz Erkan

130,592 views • 6 months ago

$AMZN CEO: “Customers badly want better price/performance. This is why we build custom chips. ” If $AMZN chip business was a standalone company, it would be at $50 billion ARR. Let that sink in. It’ll only accelerate from here as bulk of the AI workloads shift to inference and inference itself becomes more agentic. This will not only result in cheaper inference for customers, but it’ll also expand AWS margins by 7-8% as Jassy said in the earnings call. Imagine where $AMZN valuation would be if AWS growth keeps above 20% for the next decade and net margin stays above 30% thanks to custom chips. Long $AMZN.

$AMZN CEO: “Customers badly want better price/performance. This is why we build custom chips. ” If $AMZN chip business was a standalone company, it would be at $50 billion ARR. Let that sink in. It’ll only accelerate from here as bulk of the AI workloads shift to inference and inference itself becomes more agentic. This will not only result in cheaper inference for customers, but it’ll also expand AWS margins by 7-8% as Jassy said in the earnings call. Imagine where $AMZN valuation would be if AWS growth keeps above 20% for the next decade and net margin stays above 30% thanks to custom chips. Long $AMZN.

Oguz Erkan

61,999 views • 3 months ago

Groq founder Jonathan Ross explains why their architecture is better for inference than GPUs. Inference doesn't require as much compute power as training. What matters in inference is speed, cost, and energy consumption. $NVDA was going to lose market share as AI workloads shift to inference. They saw this and made a highly strategic move by striking a deal with Groq. $NVDA is already dominating training, and now it's also well-positioned to dominate inference.

Groq founder Jonathan Ross explains why their architecture is better for inference than GPUs. Inference doesn't require as much compute power as training. What matters in inference is speed, cost, and energy consumption. $NVDA was going to lose market share as AI workloads shift to inference. They saw this and made a highly strategic move by striking a deal with Groq. $NVDA is already dominating training, and now it's also well-positioned to dominate inference.

Oguz Erkan

97,186 views • 8 months ago

Peter Thiel says the quiet part loud: AI chips will get commoditized. This is why I avoid $NVDA. It made an immense amount of money marking up its GPUs by 10x and others had to pay as alternatives weren’t good enough. Now they are getting good enough. $AMD has already caught up with $NVDA in hardware performance, and ASICS are proving unexpectedly competitive due to their efficiency. $GOOG trained Gemini 3 solely on TPUs, and Anthropic is running a significant part of its training and inference workloads on Trainium clusters. As alternatives become even stronger, $NVDA margins and volumes will erode, and profits will shift from hardware to the application layer.

Peter Thiel says the quiet part loud: AI chips will get commoditized. This is why I avoid $NVDA. It made an immense amount of money marking up its GPUs by 10x and others had to pay as alternatives weren’t good enough. Now they are getting good enough. $AMD has already caught up with $NVDA in hardware performance, and ASICS are proving unexpectedly competitive due to their efficiency. $GOOG trained Gemini 3 solely on TPUs, and Anthropic is running a significant part of its training and inference workloads on Trainium clusters. As alternatives become even stronger, $NVDA margins and volumes will erode, and profits will shift from hardware to the application layer.

Oguz Erkan

460,614 views • 8 months ago

This is why Nvidia bought Groq. If Groq deploys a ton of low-cost inference chips, demand for training will increase massively. The more inference you have, the more training you need, and vice versa. Nvidia can keep its high margins selling chips for training forever.

This is why Nvidia bought Groq. If Groq deploys a ton of low-cost inference chips, demand for training will increase massively. The more inference you have, the more training you need, and vice versa. Nvidia can keep its high margins selling chips for training forever.

Lin

92,571 views • 8 months ago

$NVDA CEO: “Demand for inference will go up by a billion times.” $AMD will be one of the biggest winners as it has the upper hand in inference thanks to larger memory bandwith and capacity of its chips. Stock is still at 22x 2027 earnings. How can’t you be bullish on $AMD?

$NVDA CEO: “Demand for inference will go up by a billion times.” $AMD will be one of the biggest winners as it has the upper hand in inference thanks to larger memory bandwith and capacity of its chips. Stock is still at 22x 2027 earnings. How can’t you be bullish on $AMD?

Oguz Erkan

282,229 views • 4 months ago

$AMZN having a market cap $1.5T lower than $GOOGL makes no sense. $AMZN is the only big tech company that fully designs its own in-house ASIC chips, without relying on Broadcom $AVGO like $GOOGL does for its TPUs. On top of that, its Trainium chips are 40% more price-performance efficient than leading $NVDA GPUs. And let’s not forget — $AMZN owns more than 15% of Anthropic. $AMZN is such an obvious play.

$AMZN having a market cap $1.5T lower than $GOOGL makes no sense. $AMZN is the only big tech company that fully designs its own in-house ASIC chips, without relying on Broadcom $AVGO like $GOOGL does for its TPUs. On top of that, its Trainium chips are 40% more price-performance efficient than leading $NVDA GPUs. And let’s not forget — $AMZN owns more than 15% of Anthropic. $AMZN is such an obvious play.

Daniel

249,429 views • 7 months ago

Chips are the foundation of every AI experience. That's why understanding the hardware behind the software has never mattered more. When you ask a chatbot a question or generate an image, there's a physical chip somewhere doing the heavy lifting. For most of computing history, that was a CPU—the "brain" of a computer—great for general tasks needed to run software and operating systems. AI is more complex: for workloads called training and inference, AI needs to perform trillions of calculations in parallel. That's where AI accelerators come in. Purpose-built accelerators can deliver significantly better performance and efficiency than general-purpose chips. Amazon Web Services Trainium chips are an example—purpose-built for AI training and inference. New chips for a new era. ⬇️

Chips are the foundation of every AI experience. That's why understanding the hardware behind the software has never mattered more. When you ask a chatbot a question or generate an image, there's a physical chip somewhere doing the heavy lifting. For most of computing history, that was a CPU—the "brain" of a computer—great for general tasks needed to run software and operating systems. AI is more complex: for workloads called training and inference, AI needs to perform trillions of calculations in parallel. That's where AI accelerators come in. Purpose-built accelerators can deliver significantly better performance and efficiency than general-purpose chips. Amazon Web Services Trainium chips are an example—purpose-built for AI training and inference. New chips for a new era. ⬇️

Amazon

27,151 views • 3 months ago

$NVDA will lead a trillion-dollar TAM for AI chips by 2030. Here is how our Futurum team is breaking this down. AI chips at least $1 trillion by 2030 (conservative est) 25% Custom which will be led by $AVGO (~70-75% share) $MRVL number 2 in custom with around 15-20% 75% Merchant GPU led by NVIDIA (~80-85% share) $AMD will take around 15% with an upside of 20% share. Perhaps the biggest beneficiary will be $TSM, which will hold the keys to all production unless $INTC and Samsung make meaningful strides at the leading edge.

$NVDA will lead a trillion-dollar TAM for AI chips by 2030. Here is how our Futurum team is breaking this down. AI chips at least $1 trillion by 2030 (conservative est) 25% Custom which will be led by $AVGO (~70-75% share) $MRVL number 2 in custom with around 15-20% 75% Merchant GPU led by NVIDIA (~80-85% share) $AMD will take around 15% with an upside of 20% share. Perhaps the biggest beneficiary will be $TSM, which will hold the keys to all production unless $INTC and Samsung make meaningful strides at the leading edge.

Daniel Newman

42,074 views • 1 year ago

"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 views • 5 months ago

Jensen Huang $NVDA says the AI infrastructure buildout will continue to grow exponentially for the next 7–8 years before reaching a level that will need to be sustained. $AMZN, $META, $GOOGL, and $MSFT are expected to spend around ~$700B in capex in 2026, up ~60% vs 2025. Growth rates like this are likely to persist for several more years. Betting on chipmakers like $NVDA, $AMD, and $AVGO remains one of the safest and most compelling plays.

Jensen Huang $NVDA says the AI infrastructure buildout will continue to grow exponentially for the next 7–8 years before reaching a level that will need to be sustained. $AMZN, $META, $GOOGL, and $MSFT are expected to spend around ~$700B in capex in 2026, up ~60% vs 2025. Growth rates like this are likely to persist for several more years. Betting on chipmakers like $NVDA, $AMD, and $AVGO remains one of the safest and most compelling plays.

Daniel

31,591 views • 6 months ago

Jensen Huang: “We (Nvidia) are 100% out of China. We went from 95% market share to 0%. I can’t imagine any policymaker thinking that’s a good idea.” Chinese AI labs like DeepSeek now have to use Chinese chips for both inference and training. Chinese chips will inevitably rise if this Nvidia chip ban continues.

Jensen Huang: “We (Nvidia) are 100% out of China. We went from 95% market share to 0%. I can’t imagine any policymaker thinking that’s a good idea.” Chinese AI labs like DeepSeek now have to use Chinese chips for both inference and training. Chinese chips will inevitably rise if this Nvidia chip ban continues.

Yuchen Jin

2,362,286 views • 10 months ago

$NBIS and $CRWV growing like crazy shouldn't be surprising. Satya Nadella publicly explained 6 months ago on a podcast that $MSFT would be leasing a lot of capacity going forward rather than building all itself. Here is what I think is happening: These neo-cloud businesses, both $NBIS and $CRWV, exist solely because $NVDA wants them to exist. Nvidia has an equity stake in both companies, and they are both members of the Nvidia Partners Program. Nvidia grants early access to its chips to the partners. Nvidia does that because it knows all hyperscalers, especially $MSFT and $AMZN, are racing to develop their own cutting-edge GPUs, as this is their biggest cost in scaling cloud operations. Nvidia knows that if they are allowed to control the supply, it will prolong the upgrade cycles and reduce orders to gain time to develop its own chips. This is why it's providing early access to the neo-clouds that are 100% dependent on Nvidia chips. Hyperscalers know that if they delay orders for the fresh capacity, the demand will shift to the neo-clouds that can build capacity fast with the support of $NVDA. Thus, $NVDA corners hyperscalers and strategically compels them to put new orders and cripples their ability to focus on their own chips. $MSFT sees this and says, "If this is the case, I'll just lease the capacity from the neo-clouds." It's basically a strategic response to $NVDA. It leases the bulk of their capacity, making them dependent on it. So, if $MSFT backs down, they'll have a really hard time staying alive. There is a small window of opportunity here for neo-clouds. If they can scale beyond hyperscalers and form a fragmented customer base, they may keep thriving even if hyperscalers reduce leases. Currently, $NBIS has a better shot at it than $CRWV, as $MSFT already makes up 72% of its revenue. I am long $NBIS, and I have already made 4x of money in the stock. I'll remain long $NBIS and even grow my position, but I'll remain skeptical of $CRWV for its unfavorable strategic position, as I explained above.

$NBIS and $CRWV growing like crazy shouldn't be surprising. Satya Nadella publicly explained 6 months ago on a podcast that $MSFT would be leasing a lot of capacity going forward rather than building all itself. Here is what I think is happening: These neo-cloud businesses, both $NBIS and $CRWV, exist solely because $NVDA wants them to exist. Nvidia has an equity stake in both companies, and they are both members of the Nvidia Partners Program. Nvidia grants early access to its chips to the partners. Nvidia does that because it knows all hyperscalers, especially $MSFT and $AMZN, are racing to develop their own cutting-edge GPUs, as this is their biggest cost in scaling cloud operations. Nvidia knows that if they are allowed to control the supply, it will prolong the upgrade cycles and reduce orders to gain time to develop its own chips. This is why it's providing early access to the neo-clouds that are 100% dependent on Nvidia chips. Hyperscalers know that if they delay orders for the fresh capacity, the demand will shift to the neo-clouds that can build capacity fast with the support of $NVDA. Thus, $NVDA corners hyperscalers and strategically compels them to put new orders and cripples their ability to focus on their own chips. $MSFT sees this and says, "If this is the case, I'll just lease the capacity from the neo-clouds." It's basically a strategic response to $NVDA. It leases the bulk of their capacity, making them dependent on it. So, if $MSFT backs down, they'll have a really hard time staying alive. There is a small window of opportunity here for neo-clouds. If they can scale beyond hyperscalers and form a fragmented customer base, they may keep thriving even if hyperscalers reduce leases. Currently, $NBIS has a better shot at it than $CRWV, as $MSFT already makes up 72% of its revenue. I am long $NBIS, and I have already made 4x of money in the stock. I'll remain long $NBIS and even grow my position, but I'll remain skeptical of $CRWV for its unfavorable strategic position, as I explained above.

Oguz Erkan

180,198 views • 11 months ago

Lisa Su explains why, for at least the next few years, $AMD and $NVDA GPUs will continue to dominate the AI accelerator market. “ASICs like $GOOGL TPUs are great, but they don’t offer anywhere near the same programmability or model flexibility that GPUs provide. AI is at least a 10-year cycle, and we’re only in year two. Being so early, software developers want the flexibility to innovate on different algorithms.” Even as ASICs become competitive for inference, GPUs are on another level. $AMD and $NVDA are set to keep thriving in the years ahead.

Lisa Su explains why, for at least the next few years, $AMD and $NVDA GPUs will continue to dominate the AI accelerator market. “ASICs like $GOOGL TPUs are great, but they don’t offer anywhere near the same programmability or model flexibility that GPUs provide. AI is at least a 10-year cycle, and we’re only in year two. Being so early, software developers want the flexibility to innovate on different algorithms.” Even as ASICs become competitive for inference, GPUs are on another level. $AMD and $NVDA are set to keep thriving in the years ahead.

Daniel

125,735 views • 7 months ago

$AMD CEO Lisa Su said Helios has industry-leading specs and is on track to launch this year. Built with $META, Helios targets large-scale AI training & inference aiming for up to 10x performance where performance per dollar matters.

$AMD CEO Lisa Su said Helios has industry-leading specs and is on track to launch this year. Built with $META, Helios targets large-scale AI training & inference aiming for up to 10x performance where performance per dollar matters.

Shay Boloor

118,072 views • 7 months ago

AI answers your questions in seconds, but behind that speed is something called inference—the compute-intensive process where trained models generate responses. At AWS, we've built custom chips like Trainium, intelligent routing systems, and unified infrastructure to make inference faster and more affordable. As AI agents handle complex multi-step tasks, inference accounts for 80-90% of AI computing power. We're engineering at planetary scale to keep those milliseconds reliable.

AI answers your questions in seconds, but behind that speed is something called inference—the compute-intensive process where trained models generate responses. At AWS, we've built custom chips like Trainium, intelligent routing systems, and unified infrastructure to make inference faster and more affordable. As AI agents handle complex multi-step tasks, inference accounts for 80-90% of AI computing power. We're engineering at planetary scale to keep those milliseconds reliable.

Amazon Web Services

27,021 views • 6 months ago

Jensen just said he was reading a text from DeepMind’s Demis confirming it: pre-training + post-training scaling laws are fully intact, and Gemini 3 is the proof. A huge jump in quality exactly what Blackwell and Rubin were built for. This next curve is going to be wild. $NVDA

Jensen just said he was reading a text from DeepMind’s Demis confirming it: pre-training + post-training scaling laws are fully intact, and Gemini 3 is the proof. A huge jump in quality exactly what Blackwell and Rubin were built for. This next curve is going to be wild. $NVDA

Ben Pouladian

327,245 views • 9 months ago

#WATCH | Semicon India 2025 | Sandeep Kumar, CEO, L&T Semiconductor Technologies, says, "We are building a new security protocol chip. All the chips currently in use in passports, Aadhaar cards, and payment cards are imported. There are high chances of security and data breaches. At L&T, we will develop those chips for the security protocol. These chips will be released within 1.5 years. These chips will be used in passports, credit cards, and Aadhar cards, which will become electronic, and all payment gateways... The government already has plans to make new passports with electronic chips, so as soon as the chips are ready, passports will be made using these chips..."

#WATCH | Semicon India 2025 | Sandeep Kumar, CEO, L&T Semiconductor Technologies, says, "We are building a new security protocol chip. All the chips currently in use in passports, Aadhaar cards, and payment cards are imported. There are high chances of security and data breaches. At L&T, we will develop those chips for the security protocol. These chips will be released within 1.5 years. These chips will be used in passports, credit cards, and Aadhar cards, which will become electronic, and all payment gateways... The government already has plans to make new passports with electronic chips, so as soon as the chips are ready, passports will be made using these chips..."

ANI

54,188 views • 11 months ago