正在加载视频...

视频加载失败

返回首页

Sam Altman on why he likes hard problems (like making his own chip) “Philosophically I'm very inclined to try to find the high-leverage, difficult problem that will continue the exponential. And for us this is models and compute. Those are problems that naturally suit me.” “To scale compute in...

davidsenra's profile picture

David Senra

238,374 subscribers

27,700 次观看 • 2 个月前 •via X (Twitter)

科学技术财经
LIVE

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

This is the best take I have heard around the existing AI landscape today Fantastic explanation around the constraints at OpenAI and Anthropic, focus on enterprise business models, and the doomerism narrative in AI From former AI czar David Sacks "Even though Anthropic's revenue has followed this exponential graph very predictably, it cannot do that ever. The reason is that as you hit new levels of scale, you encounter new problems. You are simply going to run out of compute, electricity, data centers, and infrastructure. There are limits to the physical world you are going to hit. And there is already evidence that Anthropic is hitting some of those limits."
4:27

This is the best take I have heard around the existing AI landscape today Fantastic explanation around the constraints at OpenAI and Anthropic, focus on enterprise business models, and the doomerism narrative in AI From former AI czar David Sacks "Even though Anthropic's revenue has followed this exponential graph very predictably, it cannot do that ever. The reason is that as you hit new levels of scale, you encounter new problems. You are simply going to run out of compute, electricity, data centers, and infrastructure. There are limits to the physical world you are going to hit. And there is already evidence that Anthropic is hitting some of those limits."

Boring_Business

236,523 次观看 • 6 个月前

.Shaun Maguire says SpaceX's data center deal with Google may suggest there is something fundamentally different about the company's terrestrial compute. "I think a lot of people were surprised by the Anthropic and Google announcements. And I think a lot of people are getting the wrong takeaway there." "The way a lot of people are interpreting that is it's like, 'oh, we're just in this temporary compute shortage. And so anyone's willing to work with anyone, and they're willing to pay any price for compute.'" "I think that might be a reasonable way to look at the Anthropic announcement. I don't think that makes sense for looking at the Google announcement." "To me, I think an alternative hypothesis is that SpaceX's terrestrial compute is not actually a commodity. There's something that's better about it than other people's terrestrial compute."
2:22

.Shaun Maguire says SpaceX's data center deal with Google may suggest there is something fundamentally different about the company's terrestrial compute. "I think a lot of people were surprised by the Anthropic and Google announcements. And I think a lot of people are getting the wrong takeaway there." "The way a lot of people are interpreting that is it's like, 'oh, we're just in this temporary compute shortage. And so anyone's willing to work with anyone, and they're willing to pay any price for compute.'" "I think that might be a reasonable way to look at the Anthropic announcement. I don't think that makes sense for looking at the Google announcement." "To me, I think an alternative hypothesis is that SpaceX's terrestrial compute is not actually a commodity. There's something that's better about it than other people's terrestrial compute."

Jack Kuhr

44,998 次观看 • 5 个月前

#WATCH | Hyderabad, Telangana: On reports of commercial LPG cylinder shortage, Mohammad Rabbani, owner of Shah Gouse Hotel, says, "There are a lot of problems with gas here in Hyderabad... There is almost a disruption in the gas supply. Because of this, there has been a slight impact, and there is a lot of difficulty in cooking... Today they are saying that it will be difficult to supply gas from tomorrow... I want to request the state government and central government to supply gas to us and if they supply gas, then we will do business and many people are facing problems in this... Riders of Swiggy and Zomato are also very helpless; these days, they too are facing a lot of problems..."
2:29

#WATCH | Hyderabad, Telangana: On reports of commercial LPG cylinder shortage, Mohammad Rabbani, owner of Shah Gouse Hotel, says, "There are a lot of problems with gas here in Hyderabad... There is almost a disruption in the gas supply. Because of this, there has been a slight impact, and there is a lot of difficulty in cooking... Today they are saying that it will be difficult to supply gas from tomorrow... I want to request the state government and central government to supply gas to us and if they supply gas, then we will do business and many people are facing problems in this... Riders of Swiggy and Zomato are also very helpless; these days, they too are facing a lot of problems..."

ANI

49,058 次观看 • 8 个月前

🚨 BLACKROCK CEO LARRY FINK ON THE AI BUBBLE DEBATE: "The demand for compute is growing faster than supply. I don't see anything but a bubble. I see greater shortages." What the shortages look like: "Shortages of skilled trade, shortages of power and energy, shortages of chips, shortages of memory. The whole supply chain from the base of people working up until the most advanced computer chip." His actual concern: "I hear a lot of people fearful about a bubble. I'm more fearful that we're not gonna be able to execute and fulfill the opportunity to make sure that America is the winner in this AI race."
0:42

🚨 BLACKROCK CEO LARRY FINK ON THE AI BUBBLE DEBATE: "The demand for compute is growing faster than supply. I don't see anything but a bubble. I see greater shortages." What the shortages look like: "Shortages of skilled trade, shortages of power and energy, shortages of chips, shortages of memory. The whole supply chain from the base of people working up until the most advanced computer chip." His actual concern: "I hear a lot of people fearful about a bubble. I'm more fearful that we're not gonna be able to execute and fulfill the opportunity to make sure that America is the winner in this AI race."

WOLF Terminal

36,442 次观看 • 23 天前

OpenAI and Anthropic are currently taking a third of the incremental world compute supply, and Dylan thinks that this will go up to half next year. At the current rate of physical compute scaling and algorithmic progress, we're a single-digit number of years away from having individual AI companies with a greater effective workforce than there are humans on the planet. So many forces are barrelling us towards centralization in AI - the economies of scale in training (where you get to amortize any learnings across billions of sessions), the fact that being slightly ahead allows you to better economize on ever-more-expensive compute. And this isn’t even accounting for the fact that eventually you’ll have models that are capable of learning from deployment and doing proper RSI One of the key questions of our age, in my view, is how we can have broad empowerment and control over AI despite the structural factors favoring centralization.
1:59

OpenAI and Anthropic are currently taking a third of the incremental world compute supply, and Dylan thinks that this will go up to half next year. At the current rate of physical compute scaling and algorithmic progress, we're a single-digit number of years away from having individual AI companies with a greater effective workforce than there are humans on the planet. So many forces are barrelling us towards centralization in AI - the economies of scale in training (where you get to amortize any learnings across billions of sessions), the fact that being slightly ahead allows you to better economize on ever-more-expensive compute. And this isn’t even accounting for the fact that eventually you’ll have models that are capable of learning from deployment and doing proper RSI One of the key questions of our age, in my view, is how we can have broad empowerment and control over AI despite the structural factors favoring centralization.

Dwarkesh Patel

118,073 次观看 • 2 个月前

ELON: AI WILL HARNESS THE POWER OF THE SUN AND GALAXY FOR COMPUTE "I think we'll see intelligence continue to scale all the way up to where most of the power of the sun is harnessed for compute and then ultimately most of the power of the galaxy, you know, sort of cloud Kardashev two, Kardashev three scale compute. So I guess once you think of artificial intelligence not as a destination that you reach, but really as part of the overall escalation of intelligence that we are aware of." Source: Jon Hernandez
1:36

ELON: AI WILL HARNESS THE POWER OF THE SUN AND GALAXY FOR COMPUTE "I think we'll see intelligence continue to scale all the way up to where most of the power of the sun is harnessed for compute and then ultimately most of the power of the galaxy, you know, sort of cloud Kardashev two, Kardashev three scale compute. So I guess once you think of artificial intelligence not as a destination that you reach, but really as part of the overall escalation of intelligence that we are aware of." Source: Jon Hernandez

Mario Nawfal

1,668,224 次观看 • 1 年前

THIS ❤️ "Everyone in the world is experiencing stuff like this. In a time in the world that is relatively dark, outside of this paddock. You know, there's a lot of negativity. There's wars absolutely everywhere. There's conflicts going on everywhere. There's issues in all these governments and people are struggling. There's a bigger disparity between poor and rich. There's just so many problems." "But through this little experience, seeing that there's a lot of empathy. There are people that are empathetic out there. There's a lot of love, a lot of caring people out there in the world. It gives me a lot of hope for mankind, if I'm really honest. To see that experience in that weekend, that's what Roscoe has taught me." ❤️
0:46

THIS ❤️ "Everyone in the world is experiencing stuff like this. In a time in the world that is relatively dark, outside of this paddock. You know, there's a lot of negativity. There's wars absolutely everywhere. There's conflicts going on everywhere. There's issues in all these governments and people are struggling. There's a bigger disparity between poor and rich. There's just so many problems." "But through this little experience, seeing that there's a lot of empathy. There are people that are empathetic out there. There's a lot of love, a lot of caring people out there in the world. It gives me a lot of hope for mankind, if I'm really honest. To see that experience in that weekend, that's what Roscoe has taught me." ❤️

sim

250,797 次观看 • 1 年前

Elon just gave one of the clearest explanations yet for why Terafab has to exist There are basically two problems: • Geopolitical risk — at some point, advanced chips from Taiwan may not be available • Scale — even if supply remains stable, existing fabs eventually run out of capacity as AI, humanoid robots and autonomous cars keep scaling And chips are the foundation underneath all of it Server AI, edge compute, Optimus, Tesla vehicles....everything needs enormous and reliable semiconductor supply Elon: “So it’s either build Terafab or fail to scale” That pretty much sums up the entire strategy Secure the chip supply....then scale AI and robotics without a ceiling
1:39

Elon just gave one of the clearest explanations yet for why Terafab has to exist There are basically two problems: • Geopolitical risk — at some point, advanced chips from Taiwan may not be available • Scale — even if supply remains stable, existing fabs eventually run out of capacity as AI, humanoid robots and autonomous cars keep scaling And chips are the foundation underneath all of it Server AI, edge compute, Optimus, Tesla vehicles....everything needs enormous and reliable semiconductor supply Elon: “So it’s either build Terafab or fail to scale” That pretty much sums up the entire strategy Secure the chip supply....then scale AI and robotics without a ceiling

X Freeze

816,977 次观看 • 1 个月前

ROBA Compute is officially LIVE. This isn't a feature drop. This is ROBA expanding into the AI infrastructure market - hundreds of billions in scale, and we're staking our claim. Rent GPUs on-demand. Enterprise-grade allocation. Transparent USDC pricing. Zero hidden fees and all on-chain. We're already in talks with enterprise clients for compute at scale. 15% of all compute revenue → straight to $ROBA buybacks & burns. GPU capacity is limited at launch. First come, first served. Over the next weeks we will begin accepting new compute partners to provide more power to the marketplace in addition to ROBA's own in-house GPUs and integrated robotics stack. Platform fees are discounted for a limited time only. The infrastructure layer powering physical AI and beyond starts here.
0:41

ROBA Compute is officially LIVE. This isn't a feature drop. This is ROBA expanding into the AI infrastructure market - hundreds of billions in scale, and we're staking our claim. Rent GPUs on-demand. Enterprise-grade allocation. Transparent USDC pricing. Zero hidden fees and all on-chain. We're already in talks with enterprise clients for compute at scale. 15% of all compute revenue → straight to $ROBA buybacks & burns. GPU capacity is limited at launch. First come, first served. Over the next weeks we will begin accepting new compute partners to provide more power to the marketplace in addition to ROBA's own in-house GPUs and integrated robotics stack. Platform fees are discounted for a limited time only. The infrastructure layer powering physical AI and beyond starts here.

Roba Labs

44,588 次观看 • 3 个月前

Interesting to hear OpenAI Head of Compute Sachin Katti Sachin Katti explain why inference workloads will still be served from giant data centers in Texas rather than smaller edge locations.. The inherent scale benefits of large DCs + lack of available labor makes building smaller DCs closer to major metros inefficient and uneconomic. "There are still significant benefits to scale on building this compute. Building 50MW of compute is far more expensive per MW than building a GW of compute at one location.. Labor is a big bottleneck around the world today especially in the US. We just don't have enough people to build these things. So getting the kind of critical human mass you need to build, you would much rather do it for a bigger scale than for little bits of 50MW spread around the country." Seems like the robust growth for DC developers like Core Scientific, Galaxy, TeraWulf, Cipher, etc. will persist for the foreseeable future, despite initial concerns that the current large scale build-outs in places like West Texas were a one-off occurrence. $CORZ $GLXY $WULF $CIFR $HUT
1:34

Interesting to hear OpenAI Head of Compute Sachin Katti Sachin Katti explain why inference workloads will still be served from giant data centers in Texas rather than smaller edge locations.. The inherent scale benefits of large DCs + lack of available labor makes building smaller DCs closer to major metros inefficient and uneconomic. "There are still significant benefits to scale on building this compute. Building 50MW of compute is far more expensive per MW than building a GW of compute at one location.. Labor is a big bottleneck around the world today especially in the US. We just don't have enough people to build these things. So getting the kind of critical human mass you need to build, you would much rather do it for a bigger scale than for little bits of 50MW spread around the country." Seems like the robust growth for DC developers like Core Scientific, Galaxy, TeraWulf, Cipher, etc. will persist for the foreseeable future, despite initial concerns that the current large scale build-outs in places like West Texas were a one-off occurrence. $CORZ $GLXY $WULF $CIFR $HUT

Rittenhouse Research

43,687 次观看 • 5 个月前

OpenAI and Anthropic can take compute which costs them $15b a GW, and turn it into $50b+ of revenue. In order to get the credit to build their datacenters, most neoclouds have already signed long term contracts to supply compute at these $15b/GW rates. However, Meta and SpaceX, who have a GW+ of discretionary compute supported by their own balance sheet, can take advantage of this surge in compute prices and sell short-term contracts at the much higher rate that the leading labs (who are best able to monetize this compute) are willing to bid for.
1:37

OpenAI and Anthropic can take compute which costs them $15b a GW, and turn it into $50b+ of revenue. In order to get the credit to build their datacenters, most neoclouds have already signed long term contracts to supply compute at these $15b/GW rates. However, Meta and SpaceX, who have a GW+ of discretionary compute supported by their own balance sheet, can take advantage of this surge in compute prices and sell short-term contracts at the much higher rate that the leading labs (who are best able to monetize this compute) are willing to bid for.

Dwarkesh Patel

53,905 次观看 • 2 个月前

At AIPCon 10, Palantir CEO Alex Karp shares our secret to sales: “We’re hoping that you’ll go to a large language model company and learn that they don’t care about you at all.” “What you will find is there are a myriad of problems that these very important models solve, and there are even bigger problems that they create.” “We’re in the business of giving you the ability to solve those problems for yourself and own the means of production.”
1:20

At AIPCon 10, Palantir CEO Alex Karp shares our secret to sales: “We’re hoping that you’ll go to a large language model company and learn that they don’t care about you at all.” “What you will find is there are a myriad of problems that these very important models solve, and there are even bigger problems that they create.” “We’re in the business of giving you the ability to solve those problems for yourself and own the means of production.”

Palantir

59,048,607 次观看 • 5 个月前

Cerebras just had the biggest IPO of the year. Founder Andrew Feldman says the 3 most important things he had to convince investors of while doing the roadshow were that demand for inference is going to 1,000,000x, the GPU isn't the only way to do compute, and that the CUDA moat is overstated. What he said: "Jensen said some time ago on Brad Gerstner's podcast that the demand for inference will grow by a 1,000,000x, and nobody believed him. And at the same time, you saw Sam Altman displaying real vision and going out and trying to lock up huge amounts of compute, memory, data centers, and power, because he saw it too." "[We tried] to share what that means — what exponential demand means. And that we're still so early, and yet the demand for AI compute is overwhelming." "The other thing is that there are lots of ways to do this. The GPU isn't the only way. You've got TPUs, Trainium, and us. There are lots of different ways to build a solution here." "And finally — the notion that CUDA is this grand lock-in is overplayed. Gemini 3, which is an excellent model, was trained on TPUs with no CUDA. The Anthropic models were trained on Trainium with no CUDA. Some of the best models, some of the most interesting things are being done without CUDA. And that lock-in might be overplayed." $CBRS
1:46

Cerebras just had the biggest IPO of the year. Founder Andrew Feldman says the 3 most important things he had to convince investors of while doing the roadshow were that demand for inference is going to 1,000,000x, the GPU isn't the only way to do compute, and that the CUDA moat is overstated. What he said: "Jensen said some time ago on Brad Gerstner's podcast that the demand for inference will grow by a 1,000,000x, and nobody believed him. And at the same time, you saw Sam Altman displaying real vision and going out and trying to lock up huge amounts of compute, memory, data centers, and power, because he saw it too." "[We tried] to share what that means — what exponential demand means. And that we're still so early, and yet the demand for AI compute is overwhelming." "The other thing is that there are lots of ways to do this. The GPU isn't the only way. You've got TPUs, Trainium, and us. There are lots of different ways to build a solution here." "And finally — the notion that CUDA is this grand lock-in is overplayed. Gemini 3, which is an excellent model, was trained on TPUs with no CUDA. The Anthropic models were trained on Trainium with no CUDA. Some of the best models, some of the most interesting things are being done without CUDA. And that lock-in might be overplayed." $CBRS

TBPN

36,474 次观看 • 6 个月前

Sam Altman says the AI infrastructure race is not about bigger chatbots. It is about shifting the planet's intellectual output from human brains to AI brains. "If people knew what we could do with compute, they would want way, way more." "The thing I am personally most excited about is to use AI and lots of compute to discover new science." "A recent cool example here is we built the Sora Android app using Codex... in less than a month." "You can imagine that way much further, where entire companies can build their products using lots of compute." Then he gives the scale: "Let's say that an AI company today might be generating something on the order of 10 trillion tokens a day out of frontier models." "We're going to have these models at a company be outputting more tokens per day than all of humanity put together, and then 10 times that, and then 100 times that." The bottleneck is moving from software talent to compute supply, inference speed, product integration, and where all that machine intelligence gets pointed.
3:31

Sam Altman says the AI infrastructure race is not about bigger chatbots. It is about shifting the planet's intellectual output from human brains to AI brains. "If people knew what we could do with compute, they would want way, way more." "The thing I am personally most excited about is to use AI and lots of compute to discover new science." "A recent cool example here is we built the Sora Android app using Codex... in less than a month." "You can imagine that way much further, where entire companies can build their products using lots of compute." Then he gives the scale: "Let's say that an AI company today might be generating something on the order of 10 trillion tokens a day out of frontier models." "We're going to have these models at a company be outputting more tokens per day than all of humanity put together, and then 10 times that, and then 100 times that." The bottleneck is moving from software talent to compute supply, inference speed, product integration, and where all that machine intelligence gets pointed.

Karl Mehta

16,619 次观看 • 4 个月前

Why would Elon cap usage of Grok Bot when its traction is going so well? Dylan Patel explains: "You have to remember what Elon's trying to do right now. He's taking SpaceX and turning it into a neocloud, in some sense. He's building ridiculous amounts of compute." "If you take his balance sheet, he's got like $80B of cash. $20B is going to go to rockets. The other $60B he can spend on compute." "He's raising another $40B right now. And at the same time he claims he wants to build 13 gigawatts. If you build all of that out, it's $400B+." "So where does he get the cash? He has to monetize his compute as effectively as possible. To be able to pay for all the capex that he wants to do." "And given what's going on in the spot compute market — being able to sell GPUs for $15 an hour, when the TCO of them is only $4 an hour, you have this massive delta. So he's exploiting that." "He needs to focus on, 'How do I build the most compute, period?' That's step 1. Step 2 is, 'How do I use that compute most effectively to dominate the universe?'" "The only way to build 13 GW financially is if he makes as much money as possible from the compute he has. So if that means he ripped some compute out of xAI's team to sell it to Anthropic and Google, he had to do that. Because ultimately a year from now, his xAI team will have a lot more compute to deliver by doing that." "So he needs to be able to generate all this cash flow, and bridge the financial gap, to be able to spend $300B, $400B, $500B on capex over the next year, year and a half."
2:38

Why would Elon cap usage of Grok Bot when its traction is going so well? Dylan Patel explains: "You have to remember what Elon's trying to do right now. He's taking SpaceX and turning it into a neocloud, in some sense. He's building ridiculous amounts of compute." "If you take his balance sheet, he's got like $80B of cash. $20B is going to go to rockets. The other $60B he can spend on compute." "He's raising another $40B right now. And at the same time he claims he wants to build 13 gigawatts. If you build all of that out, it's $400B+." "So where does he get the cash? He has to monetize his compute as effectively as possible. To be able to pay for all the capex that he wants to do." "And given what's going on in the spot compute market — being able to sell GPUs for $15 an hour, when the TCO of them is only $4 an hour, you have this massive delta. So he's exploiting that." "He needs to focus on, 'How do I build the most compute, period?' That's step 1. Step 2 is, 'How do I use that compute most effectively to dominate the universe?'" "The only way to build 13 GW financially is if he makes as much money as possible from the compute he has. So if that means he ripped some compute out of xAI's team to sell it to Anthropic and Google, he had to do that. Because ultimately a year from now, his xAI team will have a lot more compute to deliver by doing that." "So he needs to be able to generate all this cash flow, and bridge the financial gap, to be able to spend $300B, $400B, $500B on capex over the next year, year and a half."

TBPN

113,955 次观看 • 7 天前

ELON MUSK: "We're starting off with an advanced technology fab here in Austin, and I'd like to thank Greg Abbott and the state of Texas for the support. So in the advanced technology fab, we will have all of the equipment necessary to make a chip of any kind logical memory, and we will also have all of the equipment necessary to make the masks. So in a single building, we can create a mask, make the chip, test the chip, make another mask, and have an incredibly fast recursive loop for improving the chip design. To the best of my knowledge, this doesn't exist anywhere in the world. We're really going to push the limit of physics in compute, and we're going to try a bunch of wild and crazy things, which you can do if you've got that fast iteration loop that I can't emphasize enough the importance of being able to make it, to test it and and then make and then change the design, do another one, and have that in a single building."
2:03

ELON MUSK: "We're starting off with an advanced technology fab here in Austin, and I'd like to thank Greg Abbott and the state of Texas for the support. So in the advanced technology fab, we will have all of the equipment necessary to make a chip of any kind logical memory, and we will also have all of the equipment necessary to make the masks. So in a single building, we can create a mask, make the chip, test the chip, make another mask, and have an incredibly fast recursive loop for improving the chip design. To the best of my knowledge, this doesn't exist anywhere in the world. We're really going to push the limit of physics in compute, and we're going to try a bunch of wild and crazy things, which you can do if you've got that fast iteration loop that I can't emphasize enough the importance of being able to make it, to test it and and then make and then change the design, do another one, and have that in a single building."

DogeDesigner

16,956,306 次观看 • 7 个月前

Dylan Patel (Dylan Patel) says 43 gigawatts of new AI compute comes online next year. That's more than all the AI compute in the world today. Brad Gerstner asked Jensen Huang if it can actually get built: "I think the demand is definitely there and more. The challenge of course, is that there are a lot of obstacles. When you move from digital to the physical, there are a lot of obstacles. There are land permitting, there's power generation that you got to bring online. Of course, you need construction workers and outfitters and plumbers and all kinds of-" "We created a million new jobs just building these things." "We created a million new jobs. And their skills are not fungible... particularly the ones that are building these $50 billion data centers, you're going to want some people with great skills... And then there's, of course, whatever moratoriums that are happening, whatever issues that we're seeing around the country. And so I think my guess is that land, power, shell, the build-out resources, will be one of the most important obstacles to reaching that." — Jensen Huang, CEO of NVIDIA, with Brad Gerstner on Bg2 Pod
4:20

Dylan Patel (Dylan Patel) says 43 gigawatts of new AI compute comes online next year. That's more than all the AI compute in the world today. Brad Gerstner asked Jensen Huang if it can actually get built: "I think the demand is definitely there and more. The challenge of course, is that there are a lot of obstacles. When you move from digital to the physical, there are a lot of obstacles. There are land permitting, there's power generation that you got to bring online. Of course, you need construction workers and outfitters and plumbers and all kinds of-" "We created a million new jobs just building these things." "We created a million new jobs. And their skills are not fungible... particularly the ones that are building these $50 billion data centers, you're going to want some people with great skills... And then there's, of course, whatever moratoriums that are happening, whatever issues that we're seeing around the country. And so I think my guess is that land, power, shell, the build-out resources, will be one of the most important obstacles to reaching that." — Jensen Huang, CEO of NVIDIA, with Brad Gerstner on Bg2 Pod

bg2clips

17,157 次观看 • 26 天前

The TeraFab Scale To reach a terawatt of AI compute output per year, SpaceX needs a chip factory at a scale that does not exist. That is the TeraFab. Elon Musk on the size: ~ 100 million square feet of factory floor ~ 10x larger than the entire Tesla Gigafactory Texas ~ Designed to produce one billion full-reticle chip equivalents per year ~ Each chip producing roughly one kilowatt of compute For context, current global AI chip manufacturing tops out at around 100 gigawatts per year of total compute. The TeraFab is built for 1000 gigawatts. That is 10 times the current global capacity, from a single facility.
0:28

The TeraFab Scale To reach a terawatt of AI compute output per year, SpaceX needs a chip factory at a scale that does not exist. That is the TeraFab. Elon Musk on the size: ~ 100 million square feet of factory floor ~ 10x larger than the entire Tesla Gigafactory Texas ~ Designed to produce one billion full-reticle chip equivalents per year ~ Each chip producing roughly one kilowatt of compute For context, current global AI chip manufacturing tops out at around 100 gigawatts per year of total compute. The TeraFab is built for 1000 gigawatts. That is 10 times the current global capacity, from a single facility.

Tesla Owners Silicon Valley

24,375 次观看 • 3 个月前