- Intel extends its liquid cooling technology to support future NVIDIA ia chips
- Superfluid cooling tackles 1.5 kW heat loads for high performance systems
- Taiwan partnerships position Intel as a coolant for new generation fleas
It is not a secret for anyone that Intel, which recently named Lip-Bu Tan as a new CEO, faces difficult times.
The emblematic flea manufacturer will plan to run his foundry division in a joint venture with TSMC in order to overthrow the steam, but distinct from these plans, Intel also seeks to change his fortune by becoming a major player in the cooling of new generation AI equipment.
The company’s superfluous cooling solution would be able to manage heat production up to 1.5 kW per chip, a performance level that could be essential to cool the NVIDIA GB300 superchip and also for the next rack servers, such as those presented at GTC 2025.
Blow bubbles
During its event, Nvidia unveiled models of NVL576 Racks based in Kyber with GPU Rubin Ultra.
According to the co -founder, president and chief executive officer of Nvidia, Jensen Huang, these systems could draw up to 600 kW, with future racks potentially reaching energy demands in the megawatt. As energy needs are increasing, advanced cooling solutions such as Intel’s superfluous technology will become more and more essential.
The superfluid cooling was introduced for the first time by Intel in 2023 and used the injection of microbubbles to improve the flow of the coolant and the efficiency of the heat transfer.
mashdigi Note the technology is inspired by a method used by Mitsubishi Heavy Industries, “where bubbles are generated under the hull of ocean ships to reduce water resistance and improve propulsion efficiency”.
The site continues to explain that in the case of superfluous cooling, “a similar approach is applied by generating bubbles in the cooling liquid to increase the flow speed and improve heat elimination. This is combined with cold plates to improve thermal conductivity more.
In addition, the system uses a new type of non -conductive dielectric fluid to avoid damaging the servers submerged in the event of a leak. This approach makes it very suitable for dense calculation environments where traditional cooling methods fail.
Intel recently presented its progress to the Advanced Superfluid 2025 cooling forum in Taiwan, co-organized at the Industrial Technology Research Institute.
According to United Daily News Network (UDN)The event attracted more than 500 participants and more than ten local suppliers, highlighting a strong interest in industry. Taiwanese companies, including Maico, Yuanshan, Kuenling and Sun Max Tech, have presented equipment designed to take charge of the Intel cooling system, including servers, cooled liquid and thermal components chassis.
Intel would also have invested in advanced materials to reduce corrosion and mechanical wear, including conceptions based on liquid metals and electromagnetic pump systems, in order to improve long -term sustainability and reduce maintenance for large -scale deployments.
