Charles Liang, CEO of Supermicro, predicts that “the AI revolution will be larger than the economic revolution,” and the astonishing progress of Supermicro simply up to now yr appears to bear out this prediction. Supermicro income grew by 200% yr over yr within the third fiscal quarter, and it’s anticipated to develop one other 152% by the top of the yr. Liang spoke at VB Remodel 2024 in regards to the large wave of profitability the corporate has skilled and the developments which might be underpinning investor confidence that the AI revolution, and its profitability for well-positioned firms, will proceed.
And Supermicro is healthier positioned than most firms. AI depends upon the GPUs that firms like Nvidia (largely Nvidia) are producing, however with out Supermicro, you solely have GPUs. Supermicro ships clusters and racks of networked, cooled and pooled AI infrastructure (together with GPUs) to gasoline steady demand for its best-in-class constructing block options that allow prospects customise each part of their platform.
“As a Silicon Valley-based firm, we’re capable of work with all of the know-how leaders from Nvidia, Intel, AMD and Broadcom and so forth, so our engineering crew works with their crew from early morning to midnight if their product must be developed,” Liang mentioned. “And we work with all the foremost software program firms very carefully and all our main prospects to brainstorm and design essentially the most optimized platforms for them, based mostly on our constructing block resolution, with the very best time to market.”
Getting ready for the way forward for AI
“This AI increase, I consider that’s simply at a starting stage,” Liang mentioned. “So, we’re making ready capability to help the entire trade world wide.”
In Silicon Valley alone, they’re constructing and transport 4,000 racks monthly. About 1,000 of that quantity affords a liquid cooling resolution – and so they’re ramping up proper now to ship about 2K liquid-cooled racks a month. Liquid cooling shall be a vital issue going ahead, Liang added, now that energy constraints are rearing their ugly head. Proper now, knowledge facilities eat large quantities of energy, and utilities are having a tough time maintaining. However liquid cooling is the reply shifting ahead, Liang says. In reality, Elon Musk is utilizing Supermicro’s liquid cooling resolution for xAI.
“With liquid cooling, our knowledge heart options eat 30 to 40% much less power, in order that not simply saves prospects power price, but in addition allows prospects to deploy 30% extra computing energy with the identical energy funds,” he defined. “Our purpose is to make liquid cooling at the very least 20% to 30% of the entire knowledge heart deployment in subsequent 12 months.”
RelatedPosts
Charles Liang, CEO of Supermicro, predicts that “the AI revolution will be larger than the economic revolution,” and the astonishing progress of Supermicro simply up to now yr appears to bear out this prediction. Supermicro income grew by 200% yr over yr within the third fiscal quarter, and it’s anticipated to develop one other 152% by the top of the yr. Liang spoke at VB Remodel 2024 in regards to the large wave of profitability the corporate has skilled and the developments which might be underpinning investor confidence that the AI revolution, and its profitability for well-positioned firms, will proceed.
And Supermicro is healthier positioned than most firms. AI depends upon the GPUs that firms like Nvidia (largely Nvidia) are producing, however with out Supermicro, you solely have GPUs. Supermicro ships clusters and racks of networked, cooled and pooled AI infrastructure (together with GPUs) to gasoline steady demand for its best-in-class constructing block options that allow prospects customise each part of their platform.
“As a Silicon Valley-based firm, we’re capable of work with all of the know-how leaders from Nvidia, Intel, AMD and Broadcom and so forth, so our engineering crew works with their crew from early morning to midnight if their product must be developed,” Liang mentioned. “And we work with all the foremost software program firms very carefully and all our main prospects to brainstorm and design essentially the most optimized platforms for them, based mostly on our constructing block resolution, with the very best time to market.”
Getting ready for the way forward for AI
“This AI increase, I consider that’s simply at a starting stage,” Liang mentioned. “So, we’re making ready capability to help the entire trade world wide.”
In Silicon Valley alone, they’re constructing and transport 4,000 racks monthly. About 1,000 of that quantity affords a liquid cooling resolution – and so they’re ramping up proper now to ship about 2K liquid-cooled racks a month. Liquid cooling shall be a vital issue going ahead, Liang added, now that energy constraints are rearing their ugly head. Proper now, knowledge facilities eat large quantities of energy, and utilities are having a tough time maintaining. However liquid cooling is the reply shifting ahead, Liang says. In reality, Elon Musk is utilizing Supermicro’s liquid cooling resolution for xAI.
“With liquid cooling, our knowledge heart options eat 30 to 40% much less power, in order that not simply saves prospects power price, but in addition allows prospects to deploy 30% extra computing energy with the identical energy funds,” he defined. “Our purpose is to make liquid cooling at the very least 20% to 30% of the entire knowledge heart deployment in subsequent 12 months.”
