CONNECT WITH US
NEWS TAGGED TPU
Tuesday 30 December 2025
Commentary: Why China chose multi-accelerator strategy in AI chip development
Facing US restrictions on high-end computing products, China is restructuring its AI chip industry by advancing GPU, TPU, and NPU technologies simultaneously. Domestic firms struggle...
Tuesday 30 December 2025
China's TPU upstarts take aim at US AI chip dominance
Amid ongoing US export restrictions, Chinese company Zhonghao Xinying plans to launch its second-generation self-developed Tensor Processing Unit (TPU) chip in 2026. Industry observers...
Monday 29 December 2025
Zhonghao Xinying targets 2026 launch for 2nd-gen TPU as AI inference takes the lead
As global AI compute demand pivots from large-scale model training toward application deployment, Zhonghao Xinying (Hangzhou) Technology founder and CEO Yang Gongyifan said in an...
Sunday 28 December 2025
TorchTPU collaboration makes software compatibility the new AI hardware battleground
Google is advancing its TorchTPU initiative to optimize PyTorch performance on its proprietary TPU chips, as reported by Reuters. This effort, undertaken in collaboration...
Friday 26 December 2025
Nvidia Groq partnership signals shift toward specialized inference chips as AI workloads scale
Nvidia is set to include innovations from Groq, an AI inference chip startup, into its product ecosystem by the end of 2025, responding to an expected surge in AI inference demand...
Thursday 18 December 2025
Google reportedly launches TorchTPU project to boost TPU compatibility with PyTorch
Google has reportedly initiated the TorchTPU project to enhance support for the PyTorch machine learning framework on its tensor processing units (TPUs), aiming to challenge the software...
Tuesday 16 December 2025
Baidu's Kunlunxin spinoff mirrors Google's TPU playbook
Baidu confirmed on December 7, 2025, that it is evaluating the spin-off and listing of its AI chip unit, Kunlunxin. This announcement came just one day before the US permitted H200...
Tuesday 16 December 2025
Samsung steadies 12-Layer HBM3E, eyes larger Google TPU orders
Samsung Electronics has reportedly stabilized the performance of its 12-layer high-bandwidth memory, a development that could allow the company to expand shipments to Google's artificial...
Wednesday 10 December 2025
Google TPU gains test Nvidia's lead, but the GPU ecosystem still rules
Google's latest TPU gains have rattled the AI hardware market, but IC distributors maintain that Nvidia's GPU-led ecosystem still sets the benchmark for modern AI compute. TPUs may...
Wednesday 10 December 2025
Google TPU stays strong, Inventec confident for 2026 server outlook
ASIC server shipment growth momentum in 2026 is expected to surpass that of Nvidia-led GPU servers, with Google's TPU servers accounting for the largest shipment volume. Inventec,...
Tuesday 9 December 2025
Google TPU demand remains strong, but CoWoS and memory cap 2026 production
Google's TPU demand is expected to stay high through 2027, driven by AI applications like Gemini. However, production growth in 2026 may be limited by advanced packaging (CoWoS) and...
Monday 8 December 2025
Weekly news roundup:Intel hires TSMC veteran, Samsung reportedly wins TPU HBM4 order
These are the most-read DIGITIMES Asia stories in the week of December 1 to December 7, 2025.
Friday 5 December 2025
Broadcom CEO visits South Korea as Samsung HBM4 reportedly obtains Google TPU certification
Update: Samsung has since reached out to DIGITIMES and denied this report. The company also declined to provide more details.
Friday 5 December 2025
Google TPU surge drives Taiwan's advanced testing boom
Google DeepMind's latest AI model, Gemini 3, has made a stunning debut, outperforming ChatGPT, Claude, and other competitors, signaling a new phase in the generative AI landscape...
Thursday 4 December 2025
Tech Forum 2026: Quanta races ahead to lead high-end AI servers in 2026
Google's TPU has drawn massive attention to the competitive landscape between ASIC and GPU servers. During the DIGITIMES Tech Forum 2026, DIGITIMES Research senior analyst Jim Hsiao...