- PakGazette has leaked details of SpaceX’s plans ahead of its IPO
- They include “make our own GPUs”
- It’s unclear whether that means chips in the current lineup of Tesla AI processors or a new category of GPU or AI accelerator.
SpaceX is apparently set to make its own GPUs, according to insights gleaned from its S-1 form ahead of the company’s IPO, although these won’t be products you’ll install on your PC.
Tom’s Hardware picked up the PakGazette report, and the news site claimed to have seen the S-1 form, which is filed by publicly traded companies. SpaceX is expected to be worth more than $1 trillion (and probably closer to $2 trillion) when its shares become available on the stock market (next month, in theory).
Part of the form explains the main spending plans to develop the AI, including a mention of “making our own GPUs” in the “substantial” capital expenditures listed.
Article continues below.
As noted, these would not be a rival to Nvidia’s GeForce GPUs, but rather graphics boards designed for AI tasks, commonly known as AI accelerators.
Nvidia also makes those solutions, of course, as does AMD, but SpaceX would apparently prefer to have its own homegrown AI accelerators, because, as noted in the S-1 filing, it doesn’t have “long-term contracts with many of our direct chip suppliers.” And clearly, that’s going to be problematic given the current RAM crisis and its knock-on effects.
What’s unclear is whether SpaceX’s mention of “GPU” in the presentation could refer more broadly to Tesla’s AI processors, which are currently in the AI4 generation, with the next generation AI5 set to provide a big jump in performance (of a factor of 40 times, it’s claimed). It’s possible that the company is referring to this series of AI chips, rather than a completely separate line of GPUs for AI tasks.
When it comes to manufacturing these chips, the ‘Terafab’ project, a huge chip manufacturing center in Texas involving SpaceX, xAI and Tesla, will presumably be involved, although GPUs have not been directly mentioned in relation to this facility.
Analysis: AI5 or something completely different, will not be on your PC
At least to me, this seems more like a play for a new product, rather than a mention of existing AI chips and calling them ‘GPU’ informally. That said, looking at the bigger picture, as Tom’s points out, Elon Musk has referred to Tesla’s AI5 chips as GPUs, and even noted that “it’s basically a GPU,” so you can see where the confusion comes from. All of this could simply boil down to a liberal use of the term GPU, and mean precisely nothing.
In any case, for the average consumer, the point to keep in mind is that whatever SpaceX is doing on the GPU front, if anything, won’t be relevant to the everyday computer user. This is silicon designed to boost AI performance specifically, and whether it’s technically an AI chip in the existing lineup, or something different in terms of a new GPU accelerator designed for AI tasks, it’s all semantics: it won’t be a graphics card as such (as in one designed to process 3D graphics).
Any confusion aside, Musk’s overall goal is pretty clear: ensure the supply of AI chips by producing them himself, given the aforementioned lack of “long-term contracts” with chip makers.
It remains to be seen how comforting this leak will be for potential investors in SpaceX, but Musk seems to have big plans here, and not for the first time.
The best laptops for all budgets
Follow TechRadar on Google News and add us as a preferred source to receive news, reviews and opinions from our experts in your feeds. Be sure to click the Follow button!
And of course, you can also follow TechRadar on YouTube and tiktok for news, reviews, unboxings in video form and receive regular updates from us on WhatsApp also.
