PulseAugur
EN
LIVE 20:31:21
中文(ZH) DeepSeek秘密造芯!专攻推理,一年前已启动,招聘全程不公开

DeepSeek developing custom AI inference chip to cut Nvidia reliance · 1 source tracked

Chinese AI firm DeepSeek is reportedly developing its own AI chip, aiming to reduce reliance on Nvidia and other hardware providers. The chip, designed for inference rather than training, is in its early stages, with DeepSeek engaging with chip design firms, foundries, and memory suppliers. This move follows a significant funding round of approximately 51 billion RMB (around $7.4 billion USD) and aligns with a broader industry trend of AI companies pursuing custom hardware solutions. AI

IMPACT DeepSeek's move to custom inference chips could signal a broader trend of AI labs diversifying hardware supply chains and optimizing for specific workloads, potentially impacting the GPU market.

RANK_REASON Significant funding round and strategic shift by a major AI lab to develop custom hardware. [lever_c_demoted from significant: ic=1 ai=0.7]

Read on 量子位 (QbitAI) →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

DeepSeek developing custom AI inference chip to cut Nvidia reliance · 1 source tracked

COVERAGE [1]

  1. 量子位 (QbitAI) TIER_1 中文(ZH) · 梦晨 ·

    DeepSeek Secretly Building Chips! Focusing on Inference, Started a Year Ago, Recruitment Entirely Unannounced

    已与芯片设计公司、晶圆代工厂和存储器供应商展开接洽