Huawei Ascend
Huawei's AI chip and computing platform brand, targeting large-model training and inference scenarios. It provides a full-stack AI solution spanning chip hardware and software toolchains.
Core Facts
Cube Studio explicitly supports Huawei Ascend NPUs and domestic CPU/GPU adaptation.
DeepSeek has open-sourced infrastructure components for the Ascend platform, covering the TileLang compiler, compute library, and communication library.
On September 30, 2025, DeepSeek open-sourced a batch of underlying infrastructure components, including operators, a programming language, compute libraries, and communication libraries
Timeline (last 90 days)
主流NPU被广泛应用于手机、自动驾驶和边缘服务器中，如华为昇腾、苹果神经引擎和高通的Hexagon处理器。
An NPU (neural processing unit) is a processor designed specifically for neural network inference and training tasks, and its architecture typically includes large multiply-accumulate arrays (MAC arrays), on-chip caches, and dedicated data scheduling units.
Huawei Ascend introduced the AscendC programming framework in its early years.
DeepSeek open-sourced DeepGEMM Ascend edition, a BLAS-level GEMM operator library specifically designed to run on Huawei Ascend hardware.
Where Huawei Ascend hardware truly lags behind NVIDIA is in software ecosystem; Huawei has built its own CANN ecosystem but still has gaps when it comes to supporting mainstream large models.
The original DeepGEMM is a CUDA kernel library, and the Ascend version provides a relatively familiar interface for Ascend while handling layout and data movement.
DeepGEMM-Ascend's target is explicitly and exclusively Huawei Ascend, and it does not establish a universal software layer applicable to all chip vendors.
DeepSeek released the Ascend version of DeepGEMM (DeepGEMM-Ascend) on September 30, 2026.
The interconnect bandwidth of DeepSeek operators on Huawei Ascend supernodes has approached the hardware limit.
DeepSeek open-sourced operators and communication components for Huawei Ascend, corresponding one-to-one with the previously released GPU versions.
17 more timeline events
All Facts (20)
Cube Studio explicitly supports Huawei Ascend NPUs and domestic CPU/GPU adaptation.
80%VerifiedDeepSeek has open-sourced infrastructure components for the Ascend platform, covering the TileLang compiler, compute library, and communication library.
70%VerifiedOn September 30, 2025, DeepSeek open-sourced a batch of underlying infrastructure components, including operators, a programming language, compute libraries, and communication libraries
65%Unverified美国对华芯片管制客观上刺激了中国自主芯片研发的加速推进，华为昇腾系列芯片、寒武纪等国产AI芯片研发明显提速
80%Partially VerifiedAn NPU (neural processing unit) is a processor designed specifically for neural network inference and training tasks, and its architecture typically includes large multiply-accumulate arrays (MAC arrays), on-chip caches, and dedicated data scheduling units.
65%Unverified主流NPU被广泛应用于手机、自动驾驶和边缘服务器中，如华为昇腾、苹果神经引擎和高通的Hexagon处理器。
50%UnverifiedHuawei Ascend introduced the AscendC programming framework in its early years.
50%UnverifiedWhere Huawei Ascend hardware truly lags behind NVIDIA is in software ecosystem; Huawei has built its own CANN ecosystem but still has gaps when it comes to supporting mainstream large models.
50%UnverifiedDeepGEMM-Ascend's target is explicitly and exclusively Huawei Ascend, and it does not establish a universal software layer applicable to all chip vendors.
50%UnverifiedThe original DeepGEMM is a CUDA kernel library, and the Ascend version provides a relatively familiar interface for Ascend while handling layout and data movement.
50%UnverifiedThe interconnect bandwidth of DeepSeek operators on Huawei Ascend supernodes has approached the hardware limit.
50%UnverifiedMoE places extremely stringent demands on multi-card communication, memory bandwidth, and operator scheduling, and general-purpose CUDA operator libraries often cannot directly bring out the best performance of Ascend hardware.
50%UnverifiedTileLang was open-sourced and adapted for Huawei Ascend chips at the first opportunity.
50%UnverifiedHuawei Ascend has its own software stack, CANN, but its ecosystem richness is far inferior to that of CUDA.
50%UnverifiedTileLang allows developers to write a single codebase that can be compiled and run on both NVIDIA servers and Ascend servers, without complex Ascend adaptation.
50%UnverifiedWhat has been open-sourced this time are operators and libraries, not models, which does not mean Ascend can directly run DeepSeek services.
50%UnverifiedThe open-sourced Ascend components cover operator programming languages, matrix computation, multi-card communication, and long-context processing.
50%UnverifiedThis open-sourcing represents a critical step in addressing software shortcomings, but does not mean a comprehensive overtaking of or immediate replacement of the NVIDIA ecosystem.
50%UnverifiedDevelopers can migrate code from NVIDIA servers to the Ascend platform simply by switching the backend, without large-scale rewrites.
50%UnverifiedThe gap between Huawei Ascend and NVIDIA is mainly reflected in two dimensions: interconnect bandwidth and software ecosystem, with Ascend's corresponding CANN framework having started relatively late.
50%