Unverified50% confidenceDecision RuleExact time
DeepGEMM-Ascend was developed and validated based on the Ascend 950 series, requiring CANN 9.20 and Torch NPU.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/3/2026
First Seen
Valid until: 1/1/2027
Sources
Related Entities
Related Claims
UnverifiedDeepGEMM Ascend was first released on 2026.09.30, supports Ascend 950 devices, is fully compatible with DeepGEMM's API, supports BF16, FP8, and FP4 matrix multiplication, and uses the same package name as the original version.66% similarUnverifiedvLLM provided native support for DeepSeek-V4.1-Flash from the day of its release and completed validation on both NVIDIA and AMD GPU platforms.64% similarUnverifiedDeepSeek-V4.1-Flash adopts a MoE (Mixture of Experts) backbone architecture with 552B parameters.63% similarUnverifiedLlama 3.2 is the first version in Meta's open-source model series released in 2024 to introduce visual capabilities, offering two multimodal configurations of 11B and 90B.62% similarUnverifiedDeepSeek V3和R1采用MoE（混合专家）架构，使Mac Studio或DGX Spark这类设备能以可用速度运行大模型60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/970599API
curl https://kongchang.com/api/v1/knowledge/claims/970599MCP
get_claim(id=970599)