理解頂級(jí)算力,才能提供頂級(jí)散熱。AI的競(jìng)賽,也是散熱技術(shù)的競(jìng)賽,當(dāng)前正處于從風(fēng)冷到液冷歷史性轉(zhuǎn)換的臨界點(diǎn)。GB200/300作為高度集成的系統(tǒng),其散熱需求已從單一的芯片級(jí),擴(kuò)展到芯片、內(nèi)存、NVLink互連模塊、供電模塊等多個(gè)高熱源的整體協(xié)同散熱。
AI算力(GB200/300)AI computing power
理解頂級(jí)算力,才能提供頂級(jí)散熱。AI的競(jìng)賽,也是散熱技術(shù)的競(jìng)賽,當(dāng)前正處于從風(fēng)冷到液冷歷史性轉(zhuǎn)換的臨界點(diǎn)。GB200/300作為高度集成的系統(tǒng),其散熱需求已從單一的芯片級(jí),擴(kuò)展到芯片、內(nèi)存、NVLink互連模塊、供電模塊等多個(gè)高熱源的整體協(xié)同散熱。
Only by understanding top-tier computing power can we deliver top-tier thermal management. The race for AI supremacy is also a race for thermal management technology, and we are currently at a historic turning point in the transition from air cooling to liquid cooling. As highly integrated systems, the GB200/300 series now require thermal management that extends beyond a single chip to encompass the coordinated cooling of multiple high- heat-generating components, including the chip, memory, NVLink interconnect modules and power delivery modules.

