OpenNvidia may be the AI generation's WinTel
- Reference: 1759140910
- News link: https://www.theregister.co.uk/2025/09/29/nvidia_openai_alliance_opinion_column/
- Source link:
As two of my Register colleagues [1]noted , "The announcement has enough wiggle room to drive an AI-powered self-driving semi through." True, but it may be the start of something huge that will define the AI movement for the foreseeable future.
Nvidia adds more air to the AI bubble with vague $100B OpenAI deal [2]READ MORE
Let's step into the Wayback Machine with Mr. Peabody and Sherman to the early 1980s, when PCs from companies most of you have never heard of, such as Osborne, Kaypro, and Sinclair Research, landed on desktops.
IBM decided to get into the personal computer business, and the company needed chips. So Big Blue teamed with a relatively obscure CPU company called Intel.
That took care of the hardware, but IBM needed an operating system urgently. Initially, like everyone else, except for those guys named Steve with some company called Apple, IBM wanted to use [3]CP/M from Digital Research. That didn't work out. So, IBM called Microsoft, and Bill Gates and crew acquired Quick and Dirty Operating System (QDOS ) from Seattle Computer Products, and slapped the names MS-DOS and IBM PC-DOS on it. Microsoft also, and this is the critical bit, kept the right to sell MS-DOS to other companies.
[4]
Intel, of course, had always retained the right to sell its chips to anyone. It quickly became clear that IBM was onto something. So other new companies, Compaq specifically, sprang up to develop their own PC clones, starting with the Compaq Portable in 1983. It, and all the many other clones from companies like Dell, HP, and Packard Bell, were, of course, powered by Intel chips and ran Microsoft operating systems.
[5]
[6]
The two companies started working hand-in-glove with each other. By the late '80s, their pairing, WinTel, would rule the PC world. Decades later, while not nearly as dominant as they once were, chances are the computer in front of you is WinTel.
What does that have to do with OpenNvidia? Everything. This deal promises to create the world's largest AI infrastructure project to date. It gives OpenAI access to millions of Nvidia GPUs and the capital needed for a massive wave of next-generation data centers.
But what else are they going to say? Sucks to be you, Anthropic? Bite me, Oracle?
I mean, seriously, where exactly will all the other AI software companies get the Nvidia GPUs they so desperately need when Nvidia has promised so many of its newest Vera Rubin processors to OpenAI? If the deal comes to completion, that 10 gigawatts of Nvidia systems that OpenAI gets is roughly equal to four to five million GPUs. According to analysts, [7]Nvidia's total AI GPU run will amount to only 6.5 to 7 million chips in 2025. That doesn't leave many chips over for everyone else, does it?
An Nvidia spokesman told Reuters, "Our investments will not change our focus or impact supply to our other customers - we will continue to [8]make every customer a top priority , with or without any equity stake." But what else are they going to say? Sucks to be you, Anthropic? Bite me, Oracle?
[9]
Now, where have I seen this combination of chips and software before? Oh, right. WinTel. It worked pretty well for them, didn't it? As for their rivals back in the early days, I recall them because I was already in the tech industry then. If you're under 40, have you even heard of North Star Computers, Cromemco, or Vector Graphics? Yeah, I didn't think so.
Of course, it's possible that the US Federal Trade Commission (FTC) might have something to say about this de facto move to a monopoly. After all, today, NVIDIA has about 92 percent of the data center AI chip market. As for OpenAI, it's been growing faster than essentially all other AI companies, whether you measure it in user adoption or financial scale. By mid-2025, OpenAI's annualized revenue had soared to $10–$13 billion - up from $3.7 billion in 2024 - and its projected 2025 revenue was $12.7 billion. Simultaneously, [10]TechGaged and [11]StatCounter both report OpenAI ChatGPT’s AI chatbot market share ranges from 80.9 percent to 82.7 percent in recent months.
[12]Terminators: AI-driven robot war machines on the march
[13]AI web crawlers are destroying websites in their never-ending hunger for any and all content
[14]Are you willing to pay $100k a year per developer on AI?
[15]Caught a vibe that this coding trend might cause problems
[16]Some signs of AI model collapse begin to reveal themselves
Oh wait. I forgot. The US is governed by [17]Donald "Anarchy in AI" Trump . The FTC won't be stopping this deal. It's possible the UK's Department for Science, Innovation and Technology (DSIT) and the EU's European Commission may have something to say. Whether NVIDIA and OpenAI will pay either any attention is another matter.
True, the deal's details are still messy. As Scott Raynovich, Founder and Chief Technology Analyst of the technology analysis firm Futuriom, noted in a LinkedIn comment, "All of these deals are the same... to me they read like... 'I promise to spend a bunch of money with you if you kick a bunch back to me... but there is no guarantee... and it's all contingent on things going exactly as they are going right now, but we could always bail.'"
Far be it from me to disagree. This deal could go sideways. After all, I'm one of those who won't be surprised if [18]AI goes bust . But, if it doesn't, Nvidia is the one AI company I see surviving. Any business that's aligned closely with Nvidia may do quite well. After all, just like with the dot-com crash, after all the crying, the internet grew and grew. I expect the same will happen with AI, no matter what happens to it in the short term. So, yes, in the long run, I can see OpenNvidia dominating AI in the 2040s the way Wintel did in the 2000s. ®
Get our [19]Tech Resources
[1] https://www.theregister.com/2025/09/22/openai_nvidia_chips/
[2] https://www.theregister.com/2025/09/22/openai_nvidia_chips/
[3] https://www.theregister.com/2024/08/02/cpm_50th_anniversary/
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2aNqtF9Xouu2muyyuX6yvhAAAAUE&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aNqtF9Xouu2muyyuX6yvhAAAAUE&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aNqtF9Xouu2muyyuX6yvhAAAAUE&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[7] https://finance.yahoo.com/news/nvidia-2025-gpu-unit-forecast-142417942.html
[8] https://www.reuters.com/business/more-questions-than-answers-nvidias-100-billion-openai-deal-2025-09-23/
[9] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aNqtF9Xouu2muyyuX6yvhAAAAUE&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[10] https://www.techgaged.com/chatbot-market-share-august-2025/
[11] https://gs.statcounter.com/ai-chatbot-market-share
[12] https://www.theregister.com/2025/09/12/terminators_aidriven_robot_war_machines/
[13] https://www.theregister.com/2025/08/29/ai_web_crawlers_are_destroying/
[14] https://www.theregister.com/2025/08/15/are_you_willing_to_pay/
[15] https://www.theregister.com/2025/07/25/opinion_column_vibe_coding/
[16] https://www.theregister.com/2025/05/27/opinion_column_ai_model_collapse/
[17] https://www.theregister.com/2025/08/20/opinion_us_govt_ai/
[18] https://www.theregister.com/2025/08/25/overinflated_ai_balloon/
[19] https://whitepapers.theregister.com/
Re: Wintel ... good
In my last job before retirement, the software I was product manager for ran on Intel (mainly Linux, but also Windows). I would get requests from the salesforce about once a month for us to port it to ARM because they had a prospect. (This was a time when ARM was starting to crop up in reasonable numbers in the big cloud providers).
And I would always reply "I'd be happy to do an ARM port, but which one?". No two requests were for the same implementation.
And I think this reflects the fact that a lot of people in the industry are too young to know a world before Wintel (or too sheltered to know the one beyond it), and do not understand that Wintel is not even just a chip and an OS (or several OSes), it is an entire system architecture.
There's a crucial difference, though. Nobody (who had eyes to see) ever had any doubt that PCs and the Internet would be a history-shaking global revolution. There were hundreds of avenues of useful development that were very clear and well laid-out, already solved in theory, and only needed some work to become reality. The reason we had the dot-com bust was because there was uncertainty about the shape and speed of that.
There is quite a lot of doubt that LLMs are going to be like that. ROI is still failing to manifest, except for some specialized cases. Avenues of development are highly speculative, and rely on wholly unproven assumptions, for example that problems with current models can be solved with bigger models, or better training sets. But there is, in fact, a reasonable possibility that LLMs just won't get much better than what we have now, and turn out to be mostly useless in practice.
If LLMs fizzle, OpenAI will die, and Nvidia will not feel very well either. I would not stake my pension on that.
Dot com bust
"The reason we had the dot-com bust was because there was uncertainty about the shape and speed of that."
Indeed. It's also the reason we had the dot-com boom in the first place: there were a million ideas that might work, and somebody tried every one of them. Of course, there were also a million ideas that were idiotic and DOA, and somebody tried those too...
By contrast, the use cases for LLMs so far seem to be
- badly-written low-value code
- chatbots that prevent you reaching a human helper
- six-fingered revenge porn
Niches
Is it only me seeing it ?
1/ Subprime mortgage AAA-bonds
2/ Blockchain
3/ NFT
4/ Metaverse
5/ VR Glasses
6/ AI
Like all others, it has a niche. But it is damn small.
they don't care
Nvidia just wants to sell GPUs to whomever. At the moment OpenAI has some momentum but big challenges with $$ of course, no reason for Nvidia to commit anything to OpenAI in fact it seems they have a massive OUT of whatever agreement when OpenAI fails to find the $$ for data center space/power/cooling/etc to run this stuff. Then Nvidia will sell to others. No reason to think OpenNvidia will dominate anything in 2040s or any time, there's too much competition out there with much deeper pockets than OpenAI both on the software and on the hardware front that are panicking over FOMO.
Mention of Osborne...
... reminds me of the Osborne 1 "luggable" as it was termed by the cognoscenti, not exacly portable.
https://en.m.wikipedia.org/wiki/Osborne_1
The bloody thing weighed a ton!
You'd be better transporting an actual PC!
Wintel ... good
Wintel brought standization. USB, UEFI, etc. Which was and is good. Also for running Linux on your X86.
Look at the mess in the ARM world, both mobile and laptop (Snapdragon, Apple Silicon). Everybody is doing something and something different. Ubuntu on Snapdragon is still not fully working.