Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/sam-altman-seeks-trillions-of-dollars-to-reshape-business-of-chips-and-ai.19588/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Sam Altman Seeks Trillions of Dollars to Reshape Business of Chips and AI

Daniel Nenni

Admin
Staff member

OpenAI chief pursues investors including the U.A.E. for a project possibly requiring up to $7 trillion​

1707489838206.png


The technology behind generative AI has exploded, fueling a demand for chips that can handle the needed processing power.

Sam Altman was already trying to lead the development of human-level artificial intelligence. Now he has another great ambition: raising trillions of dollars to reshape the global semiconductor industry.

The OpenAI chief executive officer is in talks with investors including the United Arab Emirates government to raise funds for a wildly ambitious tech initiative that would boost the world’s chip-building capacity, expand its ability to power AI, among other things, and cost several trillion dollars, according to people familiar with the matter. The project could require raising as much as $5 trillion to $7 trillion, one of the people said.

The fundraising plans, which face significant obstacles, are aimed at solving constraints to OpenAI’s growth, including the scarcity of the pricey AI chips required to train large language models behind AI systems such as ChatGPT. Altman has often complained that there aren’t enough of these kinds of chips—known as graphics processing units, or GPUs—to power OpenAI’s quest for artificial general intelligence, which it defines as systems that are broadly smarter than humans.

 
I'm sure Sam and Pat will explain all about it at the Intel Foundry Event. Also note that the UMC President will be speaking as well.

Intel Foundry Services (IFS) welcomes you to IFS Direct Connect to hear exclusive insights from Intel Executives and ecosystem partners​

Hear from Intel leaders, technologists, and partners as they share details of our strategy, process technology, advanced packaging, and ecosystem. Learn how Intel Foundry Services can help you build your silicon designs leveraging Intel’s resilient, security, and sustainably oriented, source of supply.
Register now

Special Appearances By​

Sam Altman


Sam Altman
Co-founder and CEO
OpenAI

Secretary Gina M. Raimondo


Secretary Gina M. Raimondo
United States Secretary of Commerce

Satya Nadella


Satya Nadella
Chairman and Chief Executive Officer
Microsoft

Featured Speakers​

Pat Gelsinger


Pat Gelsinger
CEO
Intel

Keyvan Esfarjani


Keyvan Esfarjani
EVP, GM, Manufacturing, Supply Chain and Operations
Intel

Dr. Ann Kelleher


Dr. Ann Kelleher
EVP, GM, Technology Development
Intel

Stu Pann


Stu Pann
SVP, GM, Intel Foundry Services
Intel


Rene Hass


Rene Haas
Chief Executive Officer
Arm

Yan Xing Lee


Yuan Xing Lee
Vice President, Central Engineering
Broadcom

Eric Fisher


Eric Fisher
President
MediaTek North America

Jason Wang


Jason Wang
President
UMC

Ecosystem Spotlight​

John Lee


John Lee
General Manager and Vice President
Electronics, Semiconductors and Optics BU, Ansys

Anirudh Devgan


Anirudh Devgan
President and CEO
Cadence Design Systems

Mike Ellow


Mike Ellow
Executive Vice President
Siemens Digital Industries Software

Aart de Geus


Aart de Geus
Executive Chair and Founder
Synopsys
 
I sense hallucinogenic drugs might be a factor here. 😩 Along with some folks wanting to justify a massive fab build-out already in progress that is more likely to cause a glut than be insufficient. I can't believe Altman is serious.

Micro dosing IS common amongst AI people.
 
"In talks with TSMC, Altman has said he wants to build dozens of chip-fabrication plants in the next few years, the people said. His vision would be to raise the money from Middle East investors and have TSMC build and run them."

Isn't this going to increase the chip glut even more?
 
  • Like
Reactions: VCT
I sense hallucinogenic drugs might be a factor here. 😩 Along with some folks wanting to justify a massive fab build-out already in progress that is more likely to cause a glut than be insufficient. I can't believe Altman is serious.
Notice that the "press" reports are all completely empty of sources or details. Blogs stand far more to gain from hallucinating than SamA, and are routinely observable doing it.

We should not be echoing it. We need the practice, it is going to be a year full of hallucinations.
 
maybe Broadcom to be the another 18A customer. It's interesting to see Broadcom join the list of Arm and MediaTek.

Customer and Partner Panel with Rene Haas, Arm, Yuan Xing Lee, Broadcom, and Eric Fisher, MediaTek
 
maybe Broadcom to be the another 18A customer. It's interesting to see Broadcom join the list of Arm and MediaTek.

That is what I was thinking, also Microsoft. I am still hearing rumbles from Nvidia as well as Marvell. I know a LOT of people who will be at the IFS event and it is all day so the hallway conversations should be interesting.

I know Broadcom. They usually are not doing bleeding edge chips, they are generally a node behind with the big runners so that may work for 18A. Nvidia on the other hand is bleeding edge.

Very encouraging!
 
I think many of us here would be more qualified in modelling a fab build out plan than Sam. I get the feeling he’s really strayed from his lane here lol.

Sam is not going to build a fab. Sam is an AI visionary and can raise billions of dollars. Pat is a visionary and has a team of AI chip designers from the Habana acquisition. Pat also has Intel manufacturing. This can be a VERY powerful combination!

"Habana Labs is an Israeli semiconductor company known for its development of artificial intelligence (AI) processors and solutions. It was founded in 2016 and is headquartered in Israel. The company focuses on designing and producing high-performance AI processors for various applications, including data centers, cloud computing, and edge devices.

Habana's AI processors are designed to deliver efficient and scalable performance for deep learning workloads, enabling faster and more efficient training and inference tasks. The company's products typically feature specialized hardware accelerators optimized for neural network processing, allowing for high throughput and low latency.

In December 2019, Intel Corporation announced its acquisition of Habana Labs for approximately $2 billion. This acquisition was part of Intel's strategy to bolster its capabilities in the AI and data center markets. By integrating Habana's technology and expertise into its product portfolio, Intel aimed to strengthen its position in the rapidly growing AI hardware market and better compete with other semiconductor companies in the space."
 
I sense hallucinogenic drugs might be a factor here. 😩 Along with some folks wanting to justify a massive fab build-out already in progress that is more likely to cause a glut than be insufficient. I can't believe Altman is serious.

The guy looks and feels like Musk Elon

UAE had an embarrassing episode few years ago when public figures got onto the bitcoin hype train, and then got massively burned, with headlines like "founder of Bitcoin fund A disappeared without a trace after soliciting $$$$$$$ from B and C" coming weekly.

UAE did a 180° reversal from being one of the most crypto liberal jurisdictions in the world to the least.

One of the surest signs of a con is that the man massively overstates his wealth, and clout publicly.
 
Sam is not going to build a fab. Sam is an AI visionary and can raise billions of dollars. Pat is a visionary and has a team of AI chip designs from the Habana acquisition. Pat also has Intel manufacturing. This can be a VERY powerful combination!

"Habana Labs is an Israeli semiconductor company known for its development of artificial intelligence (AI) processors and solutions. It was founded in 2016 and is headquartered in Israel. The company focuses on designing and producing high-performance AI processors for various applications, including data centers, cloud computing, and edge devices.

Habana's AI processors are designed to deliver efficient and scalable performance for deep learning workloads, enabling faster and more efficient training and inference tasks. The company's products typically feature specialized hardware accelerators optimized for neural network processing, allowing for high throughput and low latency.

In December 2019, Intel Corporation announced its acquisition of Habana Labs for approximately $2 billion. This acquisition was part of Intel's strategy to bolster its capabilities in the AI and data center markets. By integrating Habana's technology and expertise into its product portfolio, Intel aimed to strengthen its position in the rapidly growing AI hardware market and better compete with other semiconductor companies in the space."
how does Intels AI revenue (say DC GPUs) compare to AMD and Nvidia? what is the revenue projected to be from Habana in 2024?

IMO, we need less fabs than companies have currently announced.... which is fine since they will be pushed out or cancelled. Sam will have his choice of many fab shells pretty soon. but lets see what happens
 
how does Intels AI revenue (say DC GPUs) compare to AMD and Nvidia? what is the revenue projected to be from Habana in 2024?

IMO, we need less fabs than companies have currently announced.... which is fine since they will be pushed out or cancelled. Sam will have his choice of many fab shells pretty soon. but lets see what happens
when talking about DC GPU, Intel really isn't competitive. It can be seen from number of public comparison between Gaudi 2 and H100. Intel is shy to talk about PVC probvably because of cost.
 
I'm just curious what he could actually spend $7tn on !

What's Intel's annual R&D spend or TSMC's ? Certainly well below $100bn. Raising $7tn implies spending an order of magnitude higher than that. I thought new fab equipment supply (EUV) was already very tight.

And if he's actually thinking about buying any chips, the current semi WW TAM is below $500bn if I remember correctly.

Or is he setting up a new SoftBank type fund ? You can spend unlimited $ if you do that !

PS - Just checked and it isn't April 1st yet.
 
"Altman has floated the idea of building dozens of chip-fabrication plants in the next few years with money from Middle East investors, and then paying TSMC to build and operate them, the Journal said."

The semiconductor industry is $600M and a big part of that is memory? :ROFLMAO:
 
"Altman has floated the idea of building dozens of chip-fabrication plants in the next few years with money from Middle East investors, and then paying TSMC to build and operate them, the Journal said."

The semiconductor industry is $600M and a big part of that is memory? :ROFLMAO:
That statement seems imply there will be more AI adoption in semiconductor manufacturing and provides standardized processes for all nodes, eventually tsmc licenses the know-how as AI database and then human being just copy them exactly for all over the world to build advanced Fabs. This Utopia would arrive some days but I believe will be no early than 2030.
 

OpenAI chief pursues investors including the U.A.E. for a project possibly requiring up to $7 trillion​

View attachment 1660

The technology behind generative AI has exploded, fueling a demand for chips that can handle the needed processing power.

Sam Altman was already trying to lead the development of human-level artificial intelligence. Now he has another great ambition: raising trillions of dollars to reshape the global semiconductor industry.

The OpenAI chief executive officer is in talks with investors including the United Arab Emirates government to raise funds for a wildly ambitious tech initiative that would boost the world’s chip-building capacity, expand its ability to power AI, among other things, and cost several trillion dollars, according to people familiar with the matter. The project could require raising as much as $5 trillion to $7 trillion, one of the people said.

The fundraising plans, which face significant obstacles, are aimed at solving constraints to OpenAI’s growth, including the scarcity of the pricey AI chips required to train large language models behind AI systems such as ChatGPT. Altman has often complained that there aren’t enough of these kinds of chips—known as graphics processing units, or GPUs—to power OpenAI’s quest for artificial general intelligence, which it defines as systems that are broadly smarter than humans.

US government needs to think ahead where these fabs will be located.
 
For under 3 trillion he can buy NVIDIA ARM TSMC Intel and maybe even Samsung
I'm just curious what he could actually spend $7tn on !

What's Intel's annual R&D spend or TSMC's ? Certainly well below $100bn. Raising $7tn implies spending an order of magnitude higher than that. I thought new fab equipment supply (EUV) was already very tight.

And if he's actually thinking about buying any chips, the current semi WW TAM is below $500bn if I remember correctly.

Or is he setting up a new SoftBank type fund ? You can spend unlimited $ if you do that !

PS - Just checked and it isn't April 1st yet.
For 3 trillion, he can buy NVIDIA, TSMC, Intel, ARM, maybe even fit Samsung in.
.
 
Back
Top