7 articles
Huawei and Shanghai Jiao Tong University introduced HyperOffload, a compiler-assisted framework that reduces LLM peak device memory usage by up to 26%. The solution rethinks how data moves across AI hardware - maintaining performance while solving one of the field's core infrastructure challenges.
Researchers from Huawei and partner institutes introduced CLI-Gym, a framework designed to automatically generate command-line troubleshooting tasks for AI agents. The system created a dataset of 1,655 tasks and helped the LiberCoder model reach 46.1% on Terminal-Bench.
Huawei Korea plans to introduce its Ascend AI chip lineup domestically next year as an alternative in the AI computing market, while expanding its Harmony OS ecosystem and delivering complete AI infrastructure solutions.
Huawei just dropped EMMA, a unified multimodal AI system that handles understanding, generation, and image editing all in one place. Early tests show it's hitting state-of-the-art performance while staying surprisingly efficient.
New Bernstein projections reveal Huawei's path to dominating China's AI accelerator market next year, while domestic chip production is on track to surpass national demand by 2028.
Fresh benchmark data for Huawei's Kirin 9030 chipset reveals a redesigned CPU structure and new GPU powering the Mate80 Pro Max, with results showing solid performance gains despite not running at full clock speeds.
Huawei has released Flex:ai, an open-source software system designed to increase AI chip utilization by roughly 30%. The launch supports China's efforts to strengthen its domestic AI hardware ecosystem as US export limits continue.
We use cookies to improve your experience on our site and to show you relevant advertising. To find our more, read our privacy policy and cookie policy
Eseandre Mordi
Saad Ullah
Alex Dudov
Sergey Diakov
Peter Smith
