{"id":54012,"date":"2022-08-23T16:33:48","date_gmt":"2022-08-23T16:33:48","guid":{"rendered":"https:\/\/harchi90.com\/1074mm2-on-7nm-77-billion-transistors-up-to-2-8x-faster-than-nvidia-ampere-at-550w\/"},"modified":"2022-08-23T16:33:48","modified_gmt":"2022-08-23T16:33:48","slug":"1074mm2-on-7nm-77-billion-transistors-up-to-2-8x-faster-than-nvidia-ampere-at-550w","status":"publish","type":"post","link":"https:\/\/harchi90.com\/1074mm2-on-7nm-77-billion-transistors-up-to-2-8x-faster-than-nvidia-ampere-at-550w\/","title":{"rendered":"1074mm2 on 7nm, 77 Billion Transistors, Up To 2.8x Faster Than NVIDIA Ampere at 550W"},"content":{"rendered":"
Earlier this month, we reported that Birentech, a company hailing from China, was working on its fastest GPU to date, the Biren BR100. Based on what the company has publicly revealed, the Biren BR100 aims to be a General-Purpose GPU that would offer faster performance than NVIDIA’s A100 GPUs in AI processing. Now at Hot Chips 34, the company is presenting us with more details on the specs and architecture within its Biren GPGPU lineup.<\/p>\n
The Birentech BR100 is the flagship General-Purpose GPU that China has to offer, featuring an in-house GPU architecture that utilizes a 7nm process node and houses 77 Billion transistors within its die. The GPU has been fabricated on TSMC’s 2.5D CoWoS design and also comes packed with 300 MB of on-chip cache, 64 GB of HBM2e with a memory bandwidth of 2.3 TB\/s, and support for PCIe Gen 5.0 (CXL interconnect protocol). The whole chip measures 1074mm2 which is beyond the reticle limit of the process node.<\/p>\n Some of the fundamentals that went into designing the BR100 GPU included:<\/p>\n\n