rapidsai/rmm

View on GitHub

NVIDIA RMM is a library for allocating and managing GPU memory in C++ and Python.

Stars: 731
Forks: 260
Open beginner issues: 0
Indexed issues: 72
Avg merge: 14h 4m
Merged PRs (30d): 21

Dominant language: C++
License: Apache-2.0
Last GitHub push: Sep 17, 2026
Latest indexed: Sep 19, 2026
Contributing guide: Contributing guide
Code of conduct: Code of conduct
Beginner labels: good first issue

Beginner friendly Everything

Open All

72 open issues indexed

[FEA] Pilot nightly CI remediation Open

? - Needs Triage feature request

Difficulty 5/5 Over a week Newbie friendliness 35/100

rapidsai/rmm#2556 · Opened Sep 18, 2026
Add scheduled Lychee link checks Open

Difficulty 3/5 1-2 days Newbie friendliness 72/100

rapidsai/rmm#2531 · Opened Aug 26, 2026
[BUG] cudaMemcpyBatchAsync path from #2511 breaks CUDA graph capture on CUDA 13 (silently drops the copy) Open

Difficulty 3/5 1-2 days Newbie friendliness 68/100

rapidsai/rmm#2518 · 1 comment · Opened Aug 19, 2026
[FEA] Use CUDA's current async pool as RMM's default memory resource Open

feature request

rapidsai/rmm#2497 · 5 comments · 1 assignee · Opened Aug 3, 2026
[FEA] Improve pinned memory pool initialization time Open

? - Needs Triage feature request

Difficulty 5/5 Over a week Newbie friendliness 45/100

rapidsai/rmm#2496 · 2 comments · 1 reaction · Opened Jul 29, 2026
Handle cudaStreamGetId failures explicitly in stream_ordered_memory_resource Open

Difficulty 4/5 3-5 days Newbie friendliness 48/100

rapidsai/rmm#2478 · 1 comment · Opened Jul 9, 2026
[FEA] Optionally compile rmm as object library Open

? - Needs Triage feature request

Difficulty 3/5 1-2 days Newbie friendliness 57/100

rapidsai/rmm#2463 · Opened Jul 3, 2026
[FEA] Scope out components of rmm required to migrate all of RAPIDS to a CCCL-only dependency Open

? - Needs Triage feature request

Difficulty 5/5 Over a week Newbie friendliness 35/100

rapidsai/rmm#2455 · 2 comments · Opened Jun 25, 2026
CuPy missing from environment files? Open

Difficulty 2/5 1-3 hours Newbie friendliness 55/100

rapidsai/rmm#2428 · 1 comment · Opened May 29, 2026
Move system memory support check to runtime_capabilities.hpp Open

rapidsai/rmm#2415 · 1 assignee · Opened May 22, 2026
Fix callback memory resource allocation API to match CCCL and support alignment Open

Difficulty 4/5 3-5 days Newbie friendliness 48/100

rapidsai/rmm#2397 · Opened May 17, 2026
Key arena per-stream state by CUDA stream ID Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 72/100

rapidsai/rmm#2394 · Opened May 17, 2026
Hide memory resource impl symbols Open

Difficulty 5/5 Over a week Newbie friendliness 35/100

rapidsai/rmm#2390 · Opened May 17, 2026
[FEA] Expand RMM benchmarking Open

feature request

rapidsai/rmm#2385 · 2 comments · 1 assignee · Opened May 15, 2026
[FEA] Use stream protocol and cudaStream_t for Cython bindings to rmm C++ Open

? - Needs Triage feature request

rapidsai/rmm#2359 · 1 reaction · 1 assignee · Opened Apr 14, 2026
Replace cudf JNI custom tracking adaptor with `statistics_resource_adaptor` Open

Difficulty 4/5 3-5 days Newbie friendliness 52/100

rapidsai/rmm#2358 · Opened Apr 14, 2026
Enable stricter compiler warnings (-Wshadow, -Wnon-virtual-dtor, -Woverloaded-virtual) Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

rapidsai/rmm#2338 · 1 reaction · Opened Mar 25, 2026
Enforce alignment in each memory resource Open

Difficulty 4/5 3-5 days Newbie friendliness 48/100

rapidsai/rmm#2336 · Opened Mar 25, 2026
[BUG] Managed memory allocated on gpu side doesnt show up in nvtop/nvidia-smi when used with pytorch+unsloth Open

? - Needs Triage bug

Difficulty 4/5 3-5 days Newbie friendliness 30/100

rapidsai/rmm#2259 · 3 comments · Opened Feb 20, 2026
[BUG] cccl_async_resource_ref conversion operator triggers nvcc error 20013 with C++20 constexpr std::vector Open

? - Needs Triage bug

Difficulty 2/5 1-3 hours Newbie friendliness 55/100

rapidsai/rmm#2255 · Opened Feb 17, 2026
[FEA] Migrate RAPIDS to CCCL `cuda::stream_ref` / `cuda::stream` in C++ and `cuda.core` streams in Python Open

? - Needs Triage feature request

Difficulty 5/5 Over a week Newbie friendliness 30/100

rapidsai/rmm#2250 · 1 comment · 1 reaction · Opened Feb 10, 2026
[BUG] Thrust allocator fails with host/device usage Open

feature request

Difficulty 5/5 Over a week Newbie friendliness 32/100

rapidsai/rmm#2193 · 2 comments · Opened Dec 20, 2025
[FEA] Support batch allocation and deallocation Open

? - Needs Triage feature request

Difficulty 5/5 Over a week Newbie friendliness 42/100

rapidsai/rmm#2190 · 4 comments · 1 reaction · Opened Dec 19, 2025
[FEA] Enable releasing ownership of rmm device buffers Open

? - Needs Triage feature request

Difficulty 4/5 3-5 days Newbie friendliness 48/100

rapidsai/rmm#2189 · 1 comment · Opened Dec 19, 2025
[FEA] Properly support alignment argument in `do_allocate` / `do_deallocate` Open

feature request

Difficulty 4/5 3-5 days Newbie friendliness 45/100

rapidsai/rmm#2154 · 2 comments · Opened Nov 19, 2025
Investigate `host_accessible` property for managed MRs Open

feature request

Difficulty 5/5 Over a week Newbie friendliness 30/100

rapidsai/rmm#2088 · 1 reaction · Opened Oct 13, 2025
[FEA] Add CMake Option to Disable Exceptions Open

feature request

Difficulty 4/5 3-5 days Newbie friendliness 35/100

rapidsai/rmm#2078 · 7 comments · Opened Oct 6, 2025
[FEA] Add a peer-to-peer memory resource for multi-GPU nodes Open

feature request

Difficulty 5/5 Over a week Newbie friendliness 25/100

rapidsai/rmm#2075 · 1 reaction · Opened Oct 3, 2025
[FEA] Add "fallback" memory resource Open

feature request

rapidsai/rmm#2074 · 3 comments · 1 assignee · Opened Oct 3, 2025
[FEA] Expose CUDA 13 async pools for managed and pinned memory Open

feature request

rapidsai/rmm#2054 · 5 comments · 1 reaction · 1 assignee · Opened Sep 26, 2025
[FEA] Support stream semantics in rmm::mr::pool_memory_resource with rmm::mr::pinned_memory_resource as an upstream Open

feature request

Difficulty 5/5 Over a week Newbie friendliness 35/100

rapidsai/rmm#2053 · 2 comments · Opened Sep 25, 2025
Provide a centralized point for RAPIDS-wide memory configuration Open

Difficulty 5/5 Over a week Newbie friendliness 25/100

rapidsai/rmm#2050 · 1 comment · Opened Sep 24, 2025
[FEA] Implement a PerDeviceResourceManager Open

feature request

Difficulty 5/5 Over a week Newbie friendliness 35/100

rapidsai/rmm#2043 · 7 comments · Opened Sep 22, 2025
RFC: Support CUDA Stream Protocol Open

Difficulty 5/5 Over a week Newbie friendliness 30/100

rapidsai/rmm#2040 · 6 comments · 1 reaction · Opened Sep 19, 2025
[DOC] Recommend async memory resource more strongly as a default Open

doc

Difficulty 2/5 1-3 hours Newbie friendliness 45/100

rapidsai/rmm#2035 · 8 comments · 2 reactions · Opened Sep 13, 2025
[DOC] Explain managed memory behavior and prefetching design Open

doc

Difficulty 4/5 3-5 days Newbie friendliness 45/100

rapidsai/rmm#2033 · 4 comments · 1 reaction · Opened Sep 10, 2025
Mock a MemoryError failure in the failure callback MR test Open

Difficulty 3/5 1-2 days Newbie friendliness 35/100

rapidsai/rmm#2032 · Opened Sep 10, 2025
[DOC] Add documentation page on stream-ordered allocation / deallocation Open

doc

Difficulty 3/5 3-5 days Newbie friendliness 50/100

rapidsai/rmm#2015 · Opened Aug 11, 2025
Option to skip event recording on stream-ordered deallocation Open

Difficulty 5/5 Over a week Newbie friendliness 35/100

rapidsai/rmm#1995 · 4 comments · Opened Jul 21, 2025
[BUG] Repeated runs of the same allocation sequence initially succeed but then fail when using the `pool_memory_resource` Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 25/100

rapidsai/rmm#1969 · 3 comments · Opened Jun 19, 2025
[FEA] Support specifying a host mr for `rmm::device_scalar` Open

feature request

Difficulty 4/5 3-5 days Newbie friendliness 35/100

rapidsai/rmm#1959 · 16 comments · Opened Jun 16, 2025
[DOC]Need Best Practices of BinningMemoryResource Open

doc

Difficulty 4/5 3-5 days Newbie friendliness 32/100

rapidsai/rmm#1958 · 3 comments · Opened Jun 13, 2025
[FEA] Support specifying a host mr for `rmm::device_uvector` Open

feature request

Difficulty 4/5 3-5 days Newbie friendliness 35/100

rapidsai/rmm#1955 · 4 comments · 2 reactions · Opened Jun 11, 2025
[BUG] Pool allocators are incompatible with compute-sanitizer memchecks Open

feature request

Difficulty 4/5 3-5 days Newbie friendliness 35/100

rapidsai/rmm#1899 · 2 comments · Opened Apr 25, 2025
[BUG] librmm conda packages are shipping dependencies in the package Open

bug

rapidsai/rmm#1833 · 2 comments · 1 reaction · 1 assignee · Opened Feb 23, 2025
[QST]Where is cudaDevAttrMemoryPoolSupportedHandleTypes Open

question

Difficulty 3/5 1-2 days Newbie friendliness 25/100

rapidsai/rmm#1823 · 1 comment · Opened Feb 11, 2025
Improve docs for NVTX build requirements Open

doc

Difficulty 2/5 1-3 hours Newbie friendliness 35/100

rapidsai/rmm#1817 · 2 comments · Opened Feb 7, 2025
[BUG]: Strange interaction between cuDF spilling and rmm.statistics Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 45/100

rapidsai/rmm#1792 · 3 comments · Opened Jan 24, 2025
[DOC] Add Python API docs for `rmm.librmm` and `rmm.pylibrmm` Open

doc

Difficulty 3/5 1-2 days Newbie friendliness 35/100

rapidsai/rmm#1785 · Opened Jan 16, 2025
[DOC] Add code examples Open

doc

rapidsai/rmm#1784 · 2 comments · 1 reaction · 1 assignee · Opened Jan 14, 2025

rapidsai/rmm

Get new issues in your inbox