Unverified50% confidenceFactExact time
According to Bloomberg, DeepSeek plans to deploy at least 160,000 Huawei Ascend 950 chips.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
11/4/2026
First Seen
Valid until: 11/18/2026
Sources
Is DeepSeek Really the OpenAI of Yesteryear? A Five-Point Comparison
bilibiliFred_Talk11/3/2026
Related Entities
Related Claims
UnverifiedOn the Huawei 950DT chip, DeepGEMM Ascend edition achieves a hardware utilization rate of 99.8% for BF16 multiplied by BF16.71% similarUnverifiedHuawei's Ascend 950 super-node physical machine has officially made its debut, achieving 1,024-card interconnection with computing power reaching the exaflops scale.64% similarUnverifiedThe Huawei Ascend 950PR successfully ran a pruned version of the DeepSeek V4-Flash model in a single-card configuration, achieving a peak decode speed of 208 TPS.61% similarUnverified21GB of VRAM plus fifty to sixty GB or more of CPU memory is a viable configuration for running DeepSeek V4.1-Flash locally.61% similarUnverifiedMuyefengyuan's hardware configuration is a 3060 with 12G of VRAM plus a 12400F processor, which runs these plugins without pressure; if the configuration is insufficient, one can install fewer models and prioritize the cloud, and if the configuration is sufficient, one can deploy more locally to save tokens.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1133473API
curl https://kongchang.com/api/v1/knowledge/claims/1133473MCP
get_claim(id=1133473)