kanshan/TensorRT-LLMs
1
0
Fork 0
mirror of https://github.com/NVIDIA/TensorRT-LLM.git synced 2026-01-14 06:27:45 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity
a030a898d1
TensorRT-LLMs/cpp/tensorrt_llm/kernels/communicationKernels
History
kanghui0204 6f3922f318
feat: Low Precision Allreduce for PCIe based GPU (#4344) 
This PR adds a customized allreduce to TensorRT-LLM. The new allreduce is used for communication on PCIe-based GPUs via low-precision quantization, which can accelerate the PCIe allreduce process.

Signed-off-by: Hui Kang <hkang@nvidia.com>
Co-authored-by: Hui Kang <hkang@nvidia.com>
2025-05-20 06:53:46 +08:00
..
allReduceFusionKernels.cu fix potential issues in allreduce fusion kernel and ut (#4226) 2025-05-19 17:38:29 +08:00
allReduceFusionKernels.h Unify two versions of AllReduce custom op (#3032) 2025-04-22 21:58:42 +08:00
allReduceWorkspace.cu chore: bump version to 0.19.0 (#3598) (#3841) 2025-04-29 16:57:22 +08:00
allReduceWorkspace.h feat: fix and improve allreduce and fusion kernels (#3064) 2025-04-08 19:33:52 +08:00
customLowPrecisionAllReduceKernels.cu feat: Low Precision Allreduce for PCIe based GPU (#4344) 2025-05-20 06:53:46 +08:00
customLowPrecisionAllReduceKernels.h feat: Low Precision Allreduce for PCIe based GPU (#4344) 2025-05-20 06:53:46 +08:00
moeAllReduceFusionKernels.cu fix potential issues in allreduce fusion kernel and ut (#4226) 2025-05-19 17:38:29 +08:00
moeAllReduceFusionKernels.h feat: fix and improve allreduce and fusion kernels (#3064) 2025-04-08 19:33:52 +08:00