Thermal management specialist MHS said on the 4th that it will strengthen cooperation on AI Semiconductor cooling technology with AI Semiconductor specialist HyperAccel.
HyperAccel is a startup that develops AI Semiconductors optimized for Generative AI inference based on its self-developed LPU (LLM Processing Unit) architecture. It is targeting the global AI inference semiconductor market with its flagship product "Bertha" and is securing next-generation semiconductor technology competitiveness for high-performance AI infrastructure.
The two companies have been continuing technology exchanges based on a memorandum of understanding (MOU) signed last year. MHS is collaborating from the early stages of product development by supplying heatsinks for evaluation boards (EVB) for Bertha's development.
MHS is also pushing to apply its self-developed MACS (Micro Aqua Cooling System) to improve Bertha's cooling performance. To that end, it plans to continuously expand related technology cooperation.
MACS is a next-generation thermal management solution that uses an ultra-thin microchannel-based liquid cooling technology to effectively control the heat of high-heat AI Semiconductors while simultaneously improving cooling performance and energy efficiency. As the drive for higher performance in AI Semiconductors brings heat issues to the fore, it is drawing attention as a next-generation cooling technology.