Hopp til hovedinnhold
Fredag 24. april 2026AI-nyheter, ferdig filtrert for ledere
SISTE:
DeepSeek åpner V4 Preview med 1M kontekst og API-kompatibilitetOpenAI lanserer GPT-5.5 for ChatGPT og CodexAnthropic og Amazon utvider AI-alliansen med 5 GW kapasitet og ny investeringDeepSeek åpner V4 Preview med 1M kontekst og API-kompatibilitetOpenAI lanserer GPT-5.5 for ChatGPT og CodexAnthropic og Amazon utvider AI-alliansen med 5 GW kapasitet og ny investering
DeepSeek V4 Will Run Entirely on Huawei Chips — China's AI Independence Push Goes Live
DeepSeekChinaCIO

DeepSeek V4 Will Run Entirely on Huawei Chips — China's AI Independence Push Goes Live

JH
Joachim Høgby
5. april 20265. april 20263 min lesingKilde:

DeepSeek's upcoming V4 model will run exclusively on Huawei's new Ascend 950PR chips, according to The Information. It marks a concrete milestone in China's push to break free from Western semiconductor dependence.

Nvidia was not given early access to V4 — a deliberate signal that DeepSeek, in collaboration with Huawei and chip designer Cambricon, has spent months porting the model to domestic hardware.

Huawei claims the Ascend 950PR delivers roughly 2.8 times the compute of Nvidia's H20, though it still trails the H200. Alibaba, ByteDance, and Tencent have reportedly placed large orders for the new chip to power their cloud services.

DeepSeek V4 is expected to launch within the coming weeks. The company is also developing two additional V4 variants optimized for different capabilities, both built to run on Chinese silicon.

For CIOs evaluating AI vendor strategy, this matters: Chinese AI models are actively decoupling from Western infrastructure. That changes the calculus on geopolitical risk, sanctions exposure, and long-term vendor stability.

📬 Likte du denne?

AI-nyheter for ledere. Kuratert av en CIO som bygger det selv. Daglig i innboksen.