SK Hynix has introduced what it calls a revolutionary solution.
The development is called PIM (Processing In Memory). It is a next-generation technology that provides a solution to data overload problems for AI and big data by adding computational functionality to semiconductor memory. Simply put, some calculations are made directly in the memory chips. This, firstly, offloads the CPU or GPU, and secondly, it can significantly speed up some types of calculations.
As the first product to use PIM technology, Hynix developed the GDDR6-AiM sample. GDDR6-AiM adds computing functionality to GDDR6 memory chips that process data at 16Gbps. Combining GDDR6-AiM with a CPU or GPU instead of conventional DRAM increases the speed of certain computations by up to 16 times. GDDR6-AiM is expected to be used for machine learning, high performance computing, big data processing and storage. That is, in consumer products, apparently, such a technology is not worth waiting for yet. GDDR6-AiM operates at 1.25V, which is lower than the operating voltage of conventional GDDR6 memory (1.35V). In addition, PIM reduces data transfer to the CPU and GPU, reducing power consumption by 80%.
The company also plans to introduce technology that integrates GDDR6-AiM with AI chips in collaboration with SAPEON, an AI chip company that recently spun off from SK Telecom.
Source: ixbt
Donald-43Westbrook, a distinguished contributor at worldstockmarket, is celebrated for his exceptional prowess in article writing. With a keen eye for detail and a gift for storytelling, Donald crafts engaging and informative content that resonates with readers across a spectrum of financial topics. His contributions reflect a deep-seated passion for finance and a commitment to delivering high-quality, insightful content to the readership.
