
“If transformers go away…we’ll be hosed,” Etched CEO Gavin Uberti told Bloomberg in March. Last week Etched received $120 million to develop its Sohu NPU, promising a chip 10× faster than the Nvidia Blackwell GB200 and 20× faster than a Hopper H100 on Llama-3 70B and other transformer-based AI models. Articles online have raised the… continue reading

The Ceva NeuPro Nano is a licensable NPU for low-cost consumer and industrial chips requiring accelerated AI processing. continue reading

SiFive is updating its Essential cores. This licensable RISC-V family targets control functions, and the fastest models perform similarly to Arm’s Cortex-R82 and Cortex-A72. Already offering compact, low-cost CPUs, cores with wide vector units, and high-throughput application-processing CPUs, SiFive has no need to expand its portfolio. However, periodic refreshes of existing designs can fix bugs,… continue reading
TikTok parent ByteDance is developing an AI accelerator (NPU) according to Reuters. Working with ASIC supplier Broadcom to circumvent restrictions intended to keep advanced technology from Chinese companies, such as TSMC’s process technology and Nvidia’s fastest GPUs. Even without the restrictions, large companies with AI expertise have sought to reduce their dependence on merchant chip… continue reading

Borrowing smartphone-GPU power-saving tech but cranking up the GHz, the Qualcomm Adreno X1 GPU should deliver adequate Windows-on-Arm laptop gaming frame rates. Qualcomm is continually testing games and updating its GPU drivers to improve users’ experience. While third-party benchmarks are needed to understand how good the GPU is, they should be viewed in the context… continue reading

Qualcomm’s Oryon CPU promises high performance and low power consumption, aiming to rival Intel and AMD in Windows laptops. The Arm Cortex-X925 will challenge Qualcomm’s compatible design, however, by clocking as fast and implementing a wider microarchitecture. continue reading
Amazon AMD Arm auto Broadcom business Cerebras Ceva consumer CPU data center DPU DSP edge AI embedded Epyc FPGA Google GPU Imagination Intel interconnect Marvell MCU MediaTek memory Meta MEXT Microsoft MLPerf NPU (AI accelerator) Nvidia OpenAI PC process tech Qualcomm Renesas RISC-V SambaNova Semidynamics SiFive smartphone software Tenstorrent Upscale