NVIDIA/cudf

cuDF - GPU DataFrame Library

Stars: 9.8k
Forks: 1.1k
Open beginner issues: 11
Indexed issues: 1.2k
Avg merge: 2d 16h
Merged PRs (30d): 251

Dominant language: C++
License: Apache-2.0
Last GitHub push: Sep 19, 2026
Latest indexed: Sep 17, 2026
Contributing guide: Contributing guide
Code of conduct: No code of conduct
Beginner labels: good first issue

Beginner friendly Everything

Open All

[FEA] Support `polars.scan_delta/iceberg` in cudf-polars Open

cudf-polars

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

NVIDIA/cudf#24247 · Opened Sep 21, 2026
[BUG] CUDA compiler warning for CUDF_HOST_DEVICE on constexpr variables Open

? - Needs Triage bug

Difficulty 3/5 1-2 days Newbie friendliness 30/100

NVIDIA/cudf#24238 · 4 comments · Opened Sep 20, 2026
[FEA] Support outer null checks for Array columns in cudf-polars Open

? - Needs Triage feature request

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

NVIDIA/cudf#24235 · Opened Sep 19, 2026
[BUG] Ambiguous union_member::get overloads with GCC 16 break ORC/Parquet writer builds Open

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

NVIDIA/cudf#24233 · Opened Sep 18, 2026
[BUG] Missing #include <variant> in cudf/ast/expressions.hpp breaks build with GCC 16 Open

Difficulty 1/5 Under an hour Newbie friendliness 94/100

NVIDIA/cudf#24232 · Opened Sep 18, 2026
[BUG] HashCSR increases retained hash_join memory by 82% for large unique builds Open

bug libcudf

NVIDIA/cudf#24222 · 1 reaction · 1 assignee · Opened Sep 18, 2026
[FEA] Explore lto fragments in cudf-polars Open

cudf-polars feature request

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

NVIDIA/cudf#24200 · 2 comments · Opened Sep 16, 2026
[FEA] Decode Parquet decimals directly into prescribed per-column types Open

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

NVIDIA/cudf#24194 · Opened Sep 16, 2026
[FEA] Improving read parquet + udf + partitioned parquet write at scale Open

? - Needs Triage feature request

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

NVIDIA/cudf#24166 · 1 comment · Opened Sep 14, 2026
[BUG] FULL join filtering emits spurious unmatched rows for duplicate equality keys Open

? - Needs Triage bug

Difficulty 4/5 3-5 days Newbie friendliness 62/100

NVIDIA/cudf#24145 · Opened Sep 13, 2026
[FEA] Reusable filtered joins in Java for semi/anti joins Open

feature request improvement Java

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

NVIDIA/cudf#24144 · Opened Sep 13, 2026
[FEA] Split primitive row equality and hashing into separate headers Open

? - Needs Triage feature request

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

NVIDIA/cudf#24140 · Opened Sep 12, 2026
`Select(len())` fast path doesn't cover `DataFrameScan` Open

bug cudf-polars

Difficulty 4/5 3-5 days Newbie friendliness 68/100

NVIDIA/cudf#24112 · Opened Sep 10, 2026
[Story][FEA] Support as-of joins in libcudf, pylibcudf, cuDF Python, and cudf-polars Open

cudf-polars feature request libcudf

NVIDIA/cudf#24110 · 3 comments · 2 assignees · Opened Sep 10, 2026
UInt128 handling in cudf-polars only covers one specific polars rewrite shape Open

cudf-polars Needs Triage

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

NVIDIA/cudf#24108 · 1 comment · Opened Sep 10, 2026
[FEA] Expose filter_join_indices in Java Open

? - Needs Triage feature request Java

Difficulty 4/5 3-5 days Newbie friendliness 68/100

NVIDIA/cudf#24096 · 1 comment · 1 reaction · Opened Sep 9, 2026
[FEA] Add DECIMAL128 support to streaming groupby Open

feature request libcudf

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

NVIDIA/cudf#24065 · Opened Sep 9, 2026
[BUG] percentile_approx returns non-empty null list rows for mixed empty and non-empty TDigests Open

? - Needs Triage bug libcudf Spark

NVIDIA/cudf#24056 · 1 assignee · Opened Sep 9, 2026
[FEA] Add LZO decompression support to the ORC reader Open

? - Needs Triage cuIO feature request libcudf Spark

Difficulty 4/5 3-5 days Newbie friendliness 58/100

NVIDIA/cudf#24051 · Opened Sep 9, 2026
[FEA] Improve diagnostics for CUDA initialization failures Open

? - Needs Triage feature request

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

NVIDIA/cudf#24048 · Opened Sep 8, 2026
[FEA] Deduplicate CUDA UDF sources in the URL log transforms example Open

? - Needs Triage feature request

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

NVIDIA/cudf#24046 · Opened Sep 8, 2026
[BUG] groupby().sum() changes uint64 input to int64 Open

? - Needs Triage bug

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

NVIDIA/cudf#24040 · 3 comments · Opened Sep 8, 2026
[BUG] Canonicalize noncanonical BOOL8 values in murmurhash3_x86_32 Open

bug libcudf

Difficulty 2/5 1-3 hours Newbie friendliness 78/100

NVIDIA/cudf#24039 · Opened Sep 8, 2026
[BUG] DECIMAL128 groupby has a performance cliff when per-block cardinality crosses 128 Open

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

NVIDIA/cudf#24026 · 1 comment · 1 reaction · Opened Sep 8, 2026
[BUG] chunked_parquet_reader hangs or hits cudaErrorIllegalAddress on zstd DataPageV2 files: nvCOMP temp-size query is given the uncompressed level bytes Open

2 - In Progress

NVIDIA/cudf#24002 · 1 assignee · Opened Sep 6, 2026
[BUG] Parquet reader reads nearly the entire file when the page index has offset indexes but no column indexes Open

Difficulty 4/5 3-5 days Newbie friendliness 68/100

NVIDIA/cudf#24000 · Opened Sep 6, 2026
[BUG] Java: narrowing Table.writeParquetChunked to ParquetTableWriter breaks binary compatibility and downstream test mocks Open

? - Needs Triage bug

Difficulty 4/5 3-5 days Newbie friendliness 58/100

NVIDIA/cudf#23972 · 1 comment · Opened Sep 3, 2026
[Story] Rewrite parquet decode kernels around a staged, page-global preprocessing pass for validity construction and output position calculations Open

cuIO improvement libcudf Performance

NVIDIA/cudf#23965 · 1 assignee · Opened Sep 3, 2026
[Story]: Trim cudf-polars's own test suite down to what upstream polars can't cover Open

cudf-polars improvement

Difficulty 5/5 Over a week Newbie friendliness 38/100

NVIDIA/cudf#23963 · 2 comments · Opened Sep 2, 2026
[FEA] Generate docs description of cudf-polars engine options from code Open

cudf-polars feature request

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

NVIDIA/cudf#23949 · 1 reaction · Opened Sep 2, 2026
[FEA] Harden Java native-resource cleanup and close safety Open

feature request improvement Java

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

NVIDIA/cudf#23939 · Opened Sep 2, 2026
[FEA] Align HashJoin and DistinctHashJoin Java/JNI conventions Open

feature request improvement Java

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

NVIDIA/cudf#23938 · 3 comments · Opened Sep 2, 2026
[BUG]: Bug in cudf-polars to_string for datetime column with timezone Open

? - Needs Triage bug

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

NVIDIA/cudf#23935 · Opened Sep 2, 2026
[DOC] Replace references to RAPIDS ecosystem to NVIDIA CUDA-X for data science where applicable Open

? - Needs Triage doc

NVIDIA/cudf#23917 · 1 assignee · Opened Sep 1, 2026
[FEA] Evaluate fusing AST evaluation into the HashCSR retrieve kernel Open

feature request libcudf Performance

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

NVIDIA/cudf#23908 · Opened Sep 1, 2026
[BUG] computeColumn in 26.08.1 fails without NVRTC Open

? - Needs Triage bug

Difficulty 4/5 3-5 days Newbie friendliness 55/100

NVIDIA/cudf#23903 · 1 comment · Opened Aug 31, 2026
Support storage_options in Parquet metadata reads (cudf.io.parquet.read_parquet_metadata) Open

Difficulty 4/5 3-5 days Newbie friendliness 58/100

NVIDIA/cudf#23899 · 1 comment · Opened Aug 31, 2026
[FEA] Add LTO transform example for per-row-exponent decimal arithmetic Open

? - Needs Triage feature request

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

NVIDIA/cudf#23891 · 1 comment · Opened Aug 31, 2026
[FEA] Provide a unified config API for all I/O controls currently managed with environment variables Open

? - Needs Triage feature request

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

NVIDIA/cudf#23881 · Opened Aug 28, 2026
[FEA] Explore adding Vortex file format support to cuDF Open

? - Needs Triage feature request

Difficulty 5/5 Over a week Newbie friendliness 28/100

NVIDIA/cudf#23877 · 3 comments · Opened Aug 28, 2026
[FEA] libcudf APIs accepting `std::vector<T> const&` should instead accept a span Open

? - Needs Triage feature request libcudf

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

NVIDIA/cudf#23875 · 2 comments · 2 reactions · Opened Aug 28, 2026
[PERF]: Establish a default for max_concurrent_io_tasks on local scans Open

cudf-polars Performance

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

NVIDIA/cudf#23864 · Opened Aug 27, 2026
[PERF] Use per-file sizes for greedy partition binning Open

cudf-polars

NVIDIA/cudf#23859 · 1 assignee · Opened Aug 27, 2026
[FEA] Support compression="infer" in to_csv Open

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

NVIDIA/cudf#23848 · Opened Aug 26, 2026
[DOC] Document cloud/cluster engine recommendations for cudf-polars Open

cudf-polars doc

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

NVIDIA/cudf#23837 · 2 comments · Opened Aug 26, 2026
Shared-memory groupby hangs when multi-column packing exceeds the single-column compatibility guard Open

Difficulty 4/5 3-5 days Newbie friendliness 68/100

NVIDIA/cudf#23829 · Opened Aug 26, 2026
filter_join_indices corrupts results and reads out of bounds past 2^31 candidate pairs Open

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

NVIDIA/cudf#23827 · Opened Aug 26, 2026
[FEA][Java][JNI] Expose Parquet VARIANT DECIMAL decoding Open

? - Needs Triage feature request Java

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

NVIDIA/cudf#23819 · Opened Aug 25, 2026
List scatter misindexes child rows for sliced sources and targets Open

Difficulty 3/5 1-2 days Newbie friendliness 78/100

NVIDIA/cudf#23818 · Opened Aug 25, 2026
`num_py_executors` thread pool visibility Open

Difficulty 4/5 3-5 days Newbie friendliness 68/100

NVIDIA/cudf#23814 · Opened Aug 25, 2026

Showing the newest 100

This page lists what was indexed most recently. Advanced filter has the whole inventory, narrowed by language, difficulty and how long a task takes.

Open advanced filter

NVIDIA/cudf

Showing the newest 100

Get new issues in your inbox