Search results

Demo programs skip error checks (and the wrapped calls) when built with -DNDEBUG

Status: Open.

#50 In NVIDIA/cuda-checkpoint;

· rocker-zhang opened on May 29, 2026

Feature request: GDS (cuFile) or general RDMA transport for checkpoint/restore

Status: Open.

#49 In NVIDIA/cuda-checkpoint;

· 4D0R opened on May 14, 2026

cuda-checkpoint default restore preserves stale cuDeviceGetPCIBusId cache across cross-node gpu migration; NVML lookups fail as NVML look ups include cudaDeviceGetPciBusId() which returns old id and causes failure post restore

Status: Open.

#48 In NVIDIA/cuda-checkpoint;

· shubh3794 opened on Apr 22, 2026

Could not restore on process ID 3406143: "invalid argument"

Status: Open.

#47 In NVIDIA/cuda-checkpoint;

· wangzhen2271 opened on Apr 16, 2026

Is it not possible to restore from GPU 0 to GPU 0?

Status: Open.

#46 In NVIDIA/cuda-checkpoint;

· wangzhen2271 opened on Apr 9, 2026

cuda-checkpoint --toggle hangs when checkpointing one process of torch.distributed NCCL job

Status: Open.

#45 In NVIDIA/cuda-checkpoint;

· Taya666 opened on Feb 10, 2026

Restore permanently fails after initial OOM — process stuck in CHECKPOINTED state

Status: Open.

#44 In NVIDIA/cuda-checkpoint;

· asayyah opened on Jan 28, 2026

590 Features

Status: Open.

#43 In NVIDIA/cuda-checkpoint;

· rst0git opened on Dec 18, 2025

Fails with "out of memory" on MIG-enabled GPUs

Status: Open.

#42 In NVIDIA/cuda-checkpoint;

· ayushr2 opened on Dec 17, 2025

cuda-checkpoint fails to throw error when performing CRIU on non-CUDA processes

Status: Open.

#41 In NVIDIA/cuda-checkpoint;

· xuhongqiang-kernel opened on Dec 16, 2025

criu failed

Status: Open.

#40 In NVIDIA/cuda-checkpoint;

· gaogaoSpark opened on Oct 24, 2025

[vllm dump] need to toggle all the subprocesses before criu dump, it's too slow

Status: Open.

#39 In NVIDIA/cuda-checkpoint;

· Lu-yq opened on Oct 14, 2025