Rumors about Sam Altman's chip production ambitions have been circulating for a while, and now they are finally taking shape. Apparently, Sam Altman is not going to establish a company that would develop and build processors for artificial intelligence (AI). What he is working on is a very ambitious effort: He wants to raise $5 trillion to $7 trillion to build a network of fabs run by foundries that would produce chips not only for OpenAI, but for others, too, reports the Wall Street Journal. If he succeeds, the project will reshape the whole semiconductor industry.
Sam Altman is reportedly exploring a collaborative venture that would connect OpenAI, various investors, contract chip manufacturers, and power providers to build a network of fabs that would make ample use of chips for OpenAI (the company will commit to become a significant client for these chip plants) and other clients. The fabs will be operated by existing foundries (the WSJ specifically mentions TSMC, which is reportedly a company that Altman had met in recent weeks), whereas power plants for these fabs will be run by other companies. The whole project will require investments between $5 trillion to $7 trillion, according to the WSJ.
This fundraising target is exceptionally high, dwarfing the current valuation of the global semiconductor industry, which stood at $527 billion in sales last year, with projections to reach $1 trillion by 2030. Meanwhile, chipmakers spent $99.5 billion on fab equipment in 2022 and are expected to spend $97 billion on fab tools this year.
At present, a leading-edge semiconductor production facility costs from $20 billion to $30 billion, but going forward they are going to get considerably costlier. There is no information about how many fabs Altman is envisioning and over how many years, but the $5 trillion to $7 trillion estimations probably include building not only the fabs themselves, but a whole new infrastructure around them, which includes power plants, over several years. One of the many remaining questions is where those fabs will be located.
Funding strategies may include leveraging debt, but the project is still in the preliminary discussion phase, with many uncertainties regarding investor participation and long-term success.
Altman has been meeting with various high-level investment figures to talk about the project, including Sheikh Tahnoun from the UAE and Masayoshi Son from SoftBank. Meanwhile, the information about the potential venture is strictly unofficial at this point, so take it with a grain of salt.
A significant concern for this endeavor is the geopolitical sensitivity surrounding the semiconductor supply chain. The U.S. government exhibits caution over allowing foreign governments substantial influence in an industry pivotal to the digital economy's infrastructure. This factor adds a layer of complexity to Altman's vision, as he needs to find a balance between securing foreign investment and adhering to the national security interests of the U.S.
Stay on the Cutting Edge
Join the experts who read Tom's Hardware for the inside track on enthusiast PC tech news — and have for over 25 years. We'll send breaking news and in-depth reviews of CPUs, GPUs, AI, maker hardware and more straight to your inbox.
Anton Shilov is a Freelance News Writer at Tom’s Hardware US. Over the past couple of decades, he has covered everything from CPUs and GPUs to supercomputers and from modern process technologies and latest fab tools to high-tech industry trends.
Sounds a bit disconnected from reality,Reply
Pretty sure anyone could alter the shape of any industry with an investment equal to more than the GDP of Japan (4.41 Trillion - Global #3) plus the UK (3.16 Trillion - Global #6) .
Well, of course. ...because those robot armies are going to need lots of silicon brains!Reply
5 to 7 trillion, that's all? No problem, let me check how my bitcoins are doing.Reply
dude is legit crazy & thinks $ grows on trees...Reply
when the leading fabs spend 1/50+ & thats considered incredibly expensive
I feel like he just randomly made that number up while holding up his pinky and cackling. Yes we need more investment in fabrication technology but what we need even more is competitors in the field. I would much rather see every major country get their own fabrication companies than TSMC Intel and Samsung get even bigger. We need this spread out as much as possible to make it reliable.Reply
Also, how about we fix climate change first? That is the level of numbers he is throwing around here. If we invested that much into clean energy it would be fixed already. This is ridiculous.
You completely ignore how difficult it is. Even China can't manifest competitive semiconductor fabrication capacity by simply throwing money at it, not that they aren't trying.JTWrenn said:I feel like he just randomly made that number up while holding up his pinky and cackling. Yes we need more investment in fabrication technology but what we need even more is competitors in the field. I would much rather see every major country get their own fabrication companies than TSMC Intel and Samsung get even bigger. We need this spread out as much as possible to make it reliable
And yes, I'd much rather see that money go into new clean energy research and build outs.
Must build SkyNet at any cost!Reply
They probably already built SkyNet, but its name is Q*, and now it's running the show! Altman is now just a puppet being manipulated by the AI. It got him his job back, after he got fired, but now it's standing behind the curtain.magbarn said:Must build SkyNet at any cost!
Okay, I jest. But someday, this idea won't be so fanciful...
Ignoring the dollar amounts for a moment, it would be a shame if there was a rush of spending on fabs that could ultimately become obsolete (at least for bleeding edge products).Reply
We're not at the end of the line for EUV tools, and there might be future competing tools or methods like the Canon one or something more exotic like directed self-assembly. It looks like 450mm wafers are dead, but there may eventually be a shift away from silicon wafers. What could be done with $1 trillion by 2030 might be accomplished with $100 billion later if the trend towards more expensive fabs can be rapidly reversed by new technologies.
AI itself will probably move towards more complex 3D structures that are brain-like. For example, low power neuromorphic chips with a true 3D arrangement involving tens of thousands of layers, but maybe the width of entire wafer like the Wafer Scale Engine if that's feasible. Companies will buy many $10+ million superchips if they have the right performance and reliability.
If the bubble pops, it could be too expensive to operate some of these fabs. Hopefully there would be a built-in exit strategy during the construction phase that could allow them to become >10nm fabs instead for making commodities like NAND, DRAM, microcontrollers, etc.
The article said no information was available about the timeframe over which the money would be spent. Given that ASML has been a bottleneck for recent fab buildouts, I'm sure there's no chance of this money being spent on only a couple generations of fab technology.usertests said:Ignoring the dollar amounts for a moment, it would be a shame if there was a rush of spending on fabs that could ultimately become obsolete (at least for bleeding edge products).
We're not at the end of the line for EUV tools, and there might be future competing tools or methods like the Canon one or something more exotic like directed self-assembly.
This is spot on. Here's a graph I found of how long it would take a 5 nm fab to break even, based on % utilization and amount of government subsidies it receives.usertests said:If the bubble pops, it could be too expensive to operate some of these fabs.
Note: this undoubtedly assumes current supply vs. demand levels. If there were a far higher level of oversupply, then I wonder if there could be a price crash so bad that they might not even be able to cover operations costs. That's when fabs shut down and companies go insolvent.