|
Chinese artificial intelligence developer DeepSeek partnered with Huawei Technologies to create programming tools tailored for Huaweis Ascend processors, marking an effort by domestic technology companies to reduce reliance on the Nvidia ecosystem.
According to a Reuters news report, the tie-up surfaced as Chinese technology companies pushed to construct domestic alternatives to foreign semiconductor hardware and software. Through a statement published on its official WeChat account, DeepSeek confirmed the release of open-source programming infrastructure for Huaweis Ascend platform, delivering specialised compute and communication libraries to developers. The development followed Huaweis recent unveiling of its next-generation artificial intelligence processors and supernode computing systems, where the hardware maker projected wide adoption of its computing systems for model training in the coming year. As per Reuters, DeepSeek noted that Huawei delivered full technical backing throughout the development of the shared programming infrastructure. The two enterprises jointly progressed a supernode configuration built on 128 Ascend 950 chips, refining computational performance and communication links across the clustered system. Software efficiency formed a central pillar of the joint effort, with DeepSeek pointing to TileLang, an open-source high-level programming language intended for artificial intelligence chips, as a driver of developer efficiency and streamlined code logic. "To build a new generation of independent, self-controlled GPU software ecosystems, the first priority is establishing a high-level language that is universal, easy to program, and still capable of reaching the hardware's full performance potential," the Reuters news report quoted DeepSeek. As per the news report, the company stated that existing market alternatives required simpler operational alternatives to maintain competitive software standards. "TileLang was created precisely to meet this need," DeepSeek said. The group further stated that the language offered "a simpler programming model" than Nvidias proprietary CUDA computing platform. (ANI)
|
