Comments on: Inside Tesla’s Innovative And Homegrown “Dojo” AI Supercomputer https://www.nextplatform.com/2022/08/23/inside-teslas-innovative-and-homegrown-dojo-ai-supercomputer/ In-depth coverage of high-end computing at large enterprises, supercomputing centers, hyperscale data centers, and public clouds. Thu, 01 Sep 2022 02:53:57 +0000 hourly 1 https://wordpress.org/?v=6.7.1 By: Anthony Holland https://www.nextplatform.com/2022/08/23/inside-teslas-innovative-and-homegrown-dojo-ai-supercomputer/#comment-196821 Thu, 01 Sep 2022 02:53:57 +0000 https://www.nextplatform.com/?p=141090#comment-196821 Hi Timothy

Thanks for the article.

Someone in the comments section of the most recent ‘Tesla Daily’ YouTube video has asked a curious question. Googling has left me suspecting the answer is almost definitely Fremont / San Jose / Santa Clara, but I haven’t seen explicit information.

The question is:Where is DOJO being built and assembled? Where is it going to be located when fully functional?

Can you help?

Thanks,
Anthony

]]>
By: Timothy Prickett Morgan https://www.nextplatform.com/2022/08/23/inside-teslas-innovative-and-homegrown-dojo-ai-supercomputer/#comment-196770 Tue, 30 Aug 2022 21:37:31 +0000 https://www.nextplatform.com/?p=141090#comment-196770 In reply to EC.

That’s now. Also, that GPU cluster may be used for lots of things, including CAD/CAE/CFD. As I pointed out, Dojo can’t really do that — yet.

]]>
By: EC https://www.nextplatform.com/2022/08/23/inside-teslas-innovative-and-homegrown-dojo-ai-supercomputer/#comment-196597 Sat, 27 Aug 2022 20:23:50 +0000 https://www.nextplatform.com/?p=141090#comment-196597 Ironic that Dojo is doing so well and seemingly has so much promise when Tesla felt the recent need to upgrade their A100 GPU count: “Tesla has boosted its in-house AI supercomputer with thousands of additional Nvidia A100 GPUs. The Tesla supercomputer had 5,760 A100 GPUs about a year ago, and that count has since risen to 7,360 A100 GPUs — that’s an additional 1,600 GPUs, or about a 28% increase.” tomshardware(dot)com/news/tesla-brags-about-in-house-supercomputer-now-with-7360-a100-gpus

]]>
By: Timothy Prickett Morgan https://www.nextplatform.com/2022/08/23/inside-teslas-innovative-and-homegrown-dojo-ai-supercomputer/#comment-196517 Fri, 26 Aug 2022 13:28:44 +0000 https://www.nextplatform.com/?p=141090#comment-196517 In reply to Timothy.

Wake up on the wrong side of the bed this morning?

First, it is silicon.

Second, you’re right. The system is not yet built. I was talking about the architecture. And yes, I should have been more clear about that. And I will fix that.

As for you, the world doesn’t need your hostility. It generates enough already.

Have a good day.

]]>
By: Timothy https://www.nextplatform.com/2022/08/23/inside-teslas-innovative-and-homegrown-dojo-ai-supercomputer/#comment-196508 Fri, 26 Aug 2022 11:08:00 +0000 https://www.nextplatform.com/?p=141090#comment-196508 During the Q&A it was stated that Tesla have silicone of the chip in their lab, not that they have a supercomputer up and running. And what state that silicone is currently in was not even addressed. The “Base Dojo V1 system” does not exist yet, is nowhere near deployment, and you got suckered into writing this fluff piece in present tense because you didn’t do your due diligence as a journalist. Also, Musk did not found Tesla; this has comprehensively been disproven time and time again.

]]>
By: Mich https://www.nextplatform.com/2022/08/23/inside-teslas-innovative-and-homegrown-dojo-ai-supercomputer/#comment-196473 Thu, 25 Aug 2022 07:36:42 +0000 https://www.nextplatform.com/?p=141090#comment-196473 Dojo will accelerate current development of FSD – no starting from scratch.
Also Dojo is not build specifically for autonomous driving – it’s generally specialized for training neural networks including FSD, Optimus bot or anything including AWS-like outsourcing.

]]>