DeepSelect
An open-source efficient data selection component released by DeepSeek that provides a TopK operator implementation for sparse attention and samplers. It achieves speeds two to twenty times faster than the native PyTorch TopK implementation.
Core Facts
Timeline (last 90 days)
DeepEP corresponds to NCCL, handling communication between multiple GPUs; FlashMLA corresponds to FlashAttention, used for ultra-long text processing; TileKernels corresponds to CUTLASS; DeepSelect corresponds to data pipelines such as DALI.
The open-source list includes TileLang, DeepGEMM Ascend, DeepEP Ascend, TileKernels, FlashMLA, and DeepSelect, all hosted on GitHub under the MIT license.
The six components have distinct roles: DeepGEMM accelerates general matrix operations, DeepEP handles large-scale cross-device communication, TileKernels provides common vector computation and memory access operators, FlashMLA provides sparse attention operators, and DeepSelect performs efficient data selection.
DeepSelect released an Ascend version of the TopK operator, which DeepSeek officially claims is 2 to 20 times faster than the native Torch TopK.
On September 30, 2025, DeepSeek open-sourced a batch of underlying infrastructure components, including operators, a programming language, compute libraries, and communication libraries
All Facts (5)
On September 30, 2025, DeepSeek open-sourced a batch of underlying infrastructure components, including operators, a programming language, compute libraries, and communication libraries
65%UnverifiedDeepEP corresponds to NCCL, handling communication between multiple GPUs; FlashMLA corresponds to FlashAttention, used for ultra-long text processing; TileKernels corresponds to CUTLASS; DeepSelect corresponds to data pipelines such as DALI.
50%UnverifiedThe six components have distinct roles: DeepGEMM accelerates general matrix operations, DeepEP handles large-scale cross-device communication, TileKernels provides common vector computation and memory access operators, FlashMLA provides sparse attention operators, and DeepSelect performs efficient data selection.
50%UnverifiedDeepSelect released an Ascend version of the TopK operator, which DeepSeek officially claims is 2 to 20 times faster than the native Torch TopK.
50%UnverifiedThe open-source list includes TileLang, DeepGEMM Ascend, DeepEP Ascend, TileKernels, FlashMLA, and DeepSelect, all hosted on GitHub under the MIT license.
50%