Elon Musk's Texas Tesla Gigafactory is expanding to include an AI supercomputer cluster, and Supermicro's CEO is a big fan of the cooling solution. Charles Liang, founder and CEO of SuperMicro, X (formerly Twitter) to celebrate Musk's use of SuperMicro's liquid cooling technology for both Tesla's new cluster and xAI's similar supercomputer Went to
Pictured together between server racks, Liang and Musk are “trying to bring liquid cooling technology to large AI data centers.” Liang estimates that Musk's move to liquid-cooling AI data centers “could save 20 billion trees for our planet,” apparently referring to the improvements that would occur if all of the world's Liquid cooling could have been adopted at data centers.
AI data centers are notorious for massive power draw, and Supermicro hopes to alleviate that strain by advancing liquid cooling. The company claims that direct liquid cooling can offer up to 89% reduction in cooling infrastructure electricity costs compared to air cooling.
Thanks to @elonmusk for bringing liquid cooling technology to large AI data centers! This could save 20 billion trees for our planet❤️ pic.twitter.com/oJ48Dw3YVFJuly 2, 2024
In a previous tweet, Liang clarified that Super Micro's goal is to “promote DLC. [direct liquid cooling] Adoption from <1% to 30%+ in a year. Musk is widely deploying Supermicro's cooling for his Tesla Gigafactory's supercomputer cluster. The new expansion to the existing Gigafactory will have 50,000 Nvidia GPUs and more Tesla AI hardware to train Tesla's. Full self-driving feature.
The expansion is turning heads thanks to supermassive fans under construction for liquid cooling, which Musk also recently highlighted in one of his X posts (expand the tweet below). Musk estimates that the Gigafactory supercomputer will yield 130 MW at deployment, with an expected growth of 500MW once Tesla's proprietary AI hardware is installed. Musk claims that construction of the facility is almost complete, and it plans to be ready for deployment in the next few months.
Size of ~130MW power and cooling this year, but will increase to 500MW in the next 18 months. About half Tesla AI hardware, half Nvidia/other. Play to win or don't play at all.June 20, 2024
Tesla's Gigafactory supercomputer cluster should not be confused with Elon's other multibillion-dollar supercomputer cluster, the X/xAI supercomputer, which is currently under construction. That's right: Elon Musk is building not one, but two of the world's largest GPU-powered AI supercomputer clusters. The xAI supercomputer is slightly more popular than Tesla, with Musk already ordering 100,000 of Nvidia's H100 GPUs. xAI will use its supercomputer to train GrokAI, X's quirky AI chatbot alternative available to X Premium subscribers.
Also expected to be ready “within a few months”, the xAI supercomputer will also be liquid-cooled by SuperMicro and already planned to upgrade to 300,000 Nvidia B200 GPUs next summer. Is. According to recent reports, getting the XAI cluster online is a slightly higher priority for Musk than for Tesla, as Musk reportedly ordered Nvidia to ship thousands of GPUs that Tesla originally ordered for the X in June. were done. The move was reported to delay the construction of Tesla's supercomputer cluster by months, but like the Musk-based news, there is a high probability of exaggeration.