Industry Highlights

Processing Unit (Intel IPU) E2000 is Intel’s first ASIC IPU device, a 200G product co-designed with Google and in production as of 2022. It features a rich packet processing pipeline, RDMA and storage capability includin

systems, with 900GB/s link between Grace and Hopper, or between two Grace chips. The connection provides a unified, cache-coherent memory address space that combines system and HBM GPU memories for simplified programmabi

Ram Bharti1, Derek Carson1, Anton Lawrendra1, Vineet Mudgal1, Vivek Santhosh1, Sunil Shukla2, Te-Chen Tsai1 Tesla, Palo Alto, CA Tesla, Austin, TX 1 2 D1 is the ML training processor in the DOJO exa-scale computer system

technology, the design of the Radeon RX 5700 series GPUs incorporate a 256b memory interface of GDDR6 memory operating at 14Gb/s for a total of 448GB/s bandwidth, a x16 PCIe® Gen4 link interface, and six 1.4 Display Port

Ketan Sawant3, Anitha Kona1, Rob Harrison3 ARM, Austin, TX ARM, Cambridge, United Kingdom 3 ARM, Sheffield, United Kingdom 4 ARM, Bangalore, India 1 2 The Neoverse family of Arm processors target a combination of high pe

Fu-Hing Ho, Thomas To, Vamsi Nalluri, Mrinal Sarmah, Rajeev Patwari Xilinx, San Jose, CA As benefits from Moore’s Law diminish [1], general-purpose compute platforms like CPUs and GPUs continue to become increasingly pow

Wilfred Gomes1, Sanjeev Khushu2, Doug B. Ingerly1, Patrick N. Stover3, Nasirul I. Chowdhury1, Frank O'Mahony1, Ajay Balankutty1, Noam Dolev3, Martin G. Dixon1, Lei Jiang1, Surya Prekke4, Biswajit Patra4, Pavel V. Rott1,