Skip to content

Commit a2e0e43

Browse files
NvTimLiuNvTimLiu
authored
[DOC] update the download doc for 2506 release [skip ci] (#12846) (#13009)
This PR updates download docs for 25.06.0 release.
2 parents 95a5928 + a0e7e16 commit a2e0e43

File tree

2 files changed

+115
-19
lines changed

2 files changed

+115
-19
lines changed

‎docs/archive.md‎

Lines changed: 93 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -5,6 +5,99 @@ nav_order: 15
55
---
66
Below are archived releases for RAPIDS Accelerator for Apache Spark.
77

8+
## Release v25.04.0
9+
### Hardware Requirements:
10+
11+
The plugin is designed to work on NVIDIA Volta, Turing, Ampere, Ada Lovelace, Hopper and Blackwell generation datacenter GPUs. The plugin jar is tested on the following GPUs:
12+
13+
GPU Models: NVIDIA V100, T4, A10, A100, L4, H100 and B100 GPUs
14+
15+
### Software Requirements:
16+
17+
OS: Spark RAPIDS is compatible with any Linux distribution with glibc >= 2.28 (Please check ldd --version output). glibc 2.28 was released August 1, 2018.
18+
Tested on Ubuntu 20.04, Ubuntu 22.04, Rocky Linux 8 and Rocky Linux 9
19+
20+
NVIDIA Driver*: R470+
21+
22+
Runtime:
23+
Scala 2.12, 2.13
24+
Python, Java Virtual Machine (JVM) compatible with your spark-version.
25+
26+
* Check the Spark documentation for Python and Java version compatibility with your specific
27+
Spark version. For instance, visit `https://spark.apache.org/docs/3.4.1` for Spark 3.4.1.
28+
29+
Supported Spark versions:
30+
Apache Spark 3.2.0, 3.2.1, 3.2.2, 3.2.3, 3.2.4
31+
Apache Spark 3.3.0, 3.3.1, 3.3.2, 3.3.3, 3.3.4
32+
Apache Spark 3.4.0, 3.4.1, 3.4.2, 3.4.3, 3.4.4
33+
Apache Spark 3.5.0, 3.5.1, 3.5.2, 3.5.3, 3.5.4, 3.5.5, 3.5.6
34+
35+
Supported Databricks runtime versions for Azure and AWS:
36+
Databricks 11.3 ML LTS (GPU, Scala 2.12, Spark 3.3.0)
37+
Databricks 12.2 ML LTS (GPU, Scala 2.12, Spark 3.3.2)
38+
Databricks 13.3 ML LTS (GPU, Scala 2.12, Spark 3.4.1)
39+
Databricks 14.3 ML LTS (GPU, Scala 2.12, Spark 3.5.0)
40+
41+
Supported Dataproc versions (Debian/Ubuntu/Rocky):
42+
GCP Dataproc 2.1
43+
GCP Dataproc 2.2
44+
45+
Supported Dataproc Serverless versions:
46+
Spark runtime 1.1 LTS
47+
Spark runtime 1.2
48+
Spark runtime 2.0
49+
Spark runtime 2.1
50+
Spark runtime 2.2
51+
52+
*Some hardware may have a minimum driver version greater than R470. Check the GPU spec sheet
53+
for your hardware's minimum driver version.
54+
55+
*For Cloudera and EMR support, please refer to the
56+
[Distributions](https://docs.nvidia.com/spark-rapids/user-guide/latest/faq.html#which-distributions-are-supported) section of the FAQ.
57+
58+
### RAPIDS Accelerator's Support Policy for Apache Spark
59+
The RAPIDS Accelerator maintains support for Apache Spark versions available for download from [Apache Spark](https://spark.apache.org/downloads.html)
60+
61+
### Download RAPIDS Accelerator for Apache Spark v25.04.0
62+
63+
| Processor | Scala Version | Download Jar | Download Signature |
64+
|-----------|---------------|--------------|--------------------|
65+
| x86_64 | Scala 2.12 | [RAPIDS Accelerator v25.04.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.04.0/rapids-4-spark_2.12-25.04.0.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.04.0/rapids-4-spark_2.12-25.04.0.jar.asc) |
66+
| x86_64 | Scala 2.13 | [RAPIDS Accelerator v25.04.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.04.0/rapids-4-spark_2.13-25.04.0.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.04.0/rapids-4-spark_2.13-25.04.0.jar.asc) |
67+
| arm64 | Scala 2.12 | [RAPIDS Accelerator v25.04.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.04.0/rapids-4-spark_2.12-25.04.0-cuda11-arm64.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.04.0/rapids-4-spark_2.12-25.04.0-cuda11-arm64.jar.asc) |
68+
| arm64 | Scala 2.13 | [RAPIDS Accelerator v25.04.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.04.0/rapids-4-spark_2.13-25.04.0-cuda11-arm64.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.04.0/rapids-4-spark_2.13-25.04.0-cuda11-arm64.jar.asc) |
69+
70+
This package is built against CUDA 11.8. It is tested on V100, T4, A10, A100, L4, H100 and GB100 GPUs with
71+
CUDA 11.8 and CUDA 12.8.
72+
73+
### Verify signature
74+
* Download the [PUB_KEY](https://keys.openpgp.org/search?q=sw-spark@nvidia.com).
75+
* Import the public key: `gpg --import PUB_KEY`
76+
* Verify the signature for Scala 2.12 jar:
77+
`gpg --verify rapids-4-spark_2.12-25.04.0.jar.asc rapids-4-spark_2.12-25.04.0.jar`
78+
* Verify the signature for Scala 2.13 jar:
79+
`gpg --verify rapids-4-spark_2.13-25.04.0.jar.asc rapids-4-spark_2.13-25.04.0.jar`
80+
81+
The output of signature verify:
82+
83+
gpg: Good signature from "NVIDIA Spark (For the signature of spark-rapids release jars) <sw-spark@nvidia.com>"
84+
85+
### Release Notes
86+
* Support approx_count_distinct
87+
* Support group by on binary type
88+
* Support ArrayPosition function
89+
* Support Databricks 14.3 ML LTS (without support for Deletion Vector reads in Delta Lake)
90+
* Support Slice
91+
* Enable Hive text writer
92+
* Refine split-retry logs when out of memory happens to expose the real reason
93+
* Allow BigSizedJoinIterator#buildPartitioner to produce more sub-partitions to avoid CudfColumnSizeOverflowException
94+
95+
Note: There is a known issue in the 25.04.0 release when decompressing gzip files on H100 GPUs.
96+
Please find more details in [issue-16661](https://github.com/rapidsai/cudf/issues/16661).
97+
98+
For a detailed list of changes, please refer to the
99+
[CHANGELOG](https://github.com/NVIDIA/spark-rapids/blob/main/CHANGELOG.md).
100+
8101
## Release v25.02.1
9102
### Hardware Requirements:
10103

‎docs/download.md‎

Lines changed: 22 additions & 19 deletions
Original file line numberDiff line numberDiff line change
@@ -18,7 +18,7 @@ cuDF jar, that is either preinstalled in the Spark classpath on all nodes or sub
1818
that uses the RAPIDS Accelerator For Apache Spark. See the [getting-started
1919
guide](https://docs.nvidia.com/spark-rapids/user-guide/latest/getting-started/overview.html) for more details.
2020

21-
## Release v25.04.0
21+
## Release v25.06.0
2222
### Hardware Requirements:
2323

2424
The plugin is designed to work on NVIDIA Volta, Turing, Ampere, Ada Lovelace, Hopper and Blackwell generation datacenter GPUs. The plugin jar is tested on the following GPUs:
@@ -46,14 +46,14 @@ The plugin is designed to work on NVIDIA Volta, Turing, Ampere, Ada Lovelace, Ho
4646
Apache Spark 3.5.0, 3.5.1, 3.5.2, 3.5.3, 3.5.4, 3.5.5, 3.5.6
4747

4848
Supported Databricks runtime versions for Azure and AWS:
49-
Databricks 11.3 ML LTS (GPU, Scala 2.12, Spark 3.3.0)
5049
Databricks 12.2 ML LTS (GPU, Scala 2.12, Spark 3.3.2)
5150
Databricks 13.3 ML LTS (GPU, Scala 2.12, Spark 3.4.1)
5251
Databricks 14.3 ML LTS (GPU, Scala 2.12, Spark 3.5.0)
5352

5453
Supported Dataproc versions (Debian/Ubuntu/Rocky):
5554
GCP Dataproc 2.1
5655
GCP Dataproc 2.2
56+
GCP Dataproc 2.3
5757

5858
Supported Dataproc Serverless versions:
5959
Spark runtime 1.1 LTS
@@ -71,14 +71,14 @@ for your hardware's minimum driver version.
7171
### RAPIDS Accelerator's Support Policy for Apache Spark
7272
The RAPIDS Accelerator maintains support for Apache Spark versions available for download from [Apache Spark](https://spark.apache.org/downloads.html)
7373

74-
### Download RAPIDS Accelerator for Apache Spark v25.04.0
74+
### Download RAPIDS Accelerator for Apache Spark v25.06.0
7575

7676
| Processor | Scala Version | Download Jar | Download Signature |
7777
|-----------|---------------|--------------|--------------------|
78-
| x86_64 | Scala 2.12 | [RAPIDS Accelerator v25.04.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.04.0/rapids-4-spark_2.12-25.04.0.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.04.0/rapids-4-spark_2.12-25.04.0.jar.asc) |
79-
| x86_64 | Scala 2.13 | [RAPIDS Accelerator v25.04.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.04.0/rapids-4-spark_2.13-25.04.0.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.04.0/rapids-4-spark_2.13-25.04.0.jar.asc) |
80-
| arm64 | Scala 2.12 | [RAPIDS Accelerator v25.04.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.04.0/rapids-4-spark_2.12-25.04.0-cuda11-arm64.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.04.0/rapids-4-spark_2.12-25.04.0-cuda11-arm64.jar.asc) |
81-
| arm64 | Scala 2.13 | [RAPIDS Accelerator v25.04.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.04.0/rapids-4-spark_2.13-25.04.0-cuda11-arm64.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.04.0/rapids-4-spark_2.13-25.04.0-cuda11-arm64.jar.asc) |
78+
| x86_64 | Scala 2.12 | [RAPIDS Accelerator v25.06.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.06.0/rapids-4-spark_2.12-25.06.0.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.06.0/rapids-4-spark_2.12-25.06.0.jar.asc) |
79+
| x86_64 | Scala 2.13 | [RAPIDS Accelerator v25.06.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.06.0/rapids-4-spark_2.13-25.06.0.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.06.0/rapids-4-spark_2.13-25.06.0.jar.asc) |
80+
| arm64 | Scala 2.12 | [RAPIDS Accelerator v25.06.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.06.0/rapids-4-spark_2.12-25.06.0-cuda11-arm64.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.12/25.06.0/rapids-4-spark_2.12-25.06.0-cuda11-arm64.jar.asc) |
81+
| arm64 | Scala 2.13 | [RAPIDS Accelerator v25.06.0](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.06.0/rapids-4-spark_2.13-25.06.0-cuda11-arm64.jar) | [Signature](https://repo1.maven.org/maven2/com/nvidia/rapids-4-spark_2.13/25.06.0/rapids-4-spark_2.13-25.06.0-cuda11-arm64.jar.asc) |
8282

8383
This package is built against CUDA 11.8. It is tested on V100, T4, A10, A100, L4, H100 and GB100 GPUs with
8484
CUDA 11.8 and CUDA 12.8.
@@ -87,25 +87,28 @@ CUDA 11.8 and CUDA 12.8.
8787
* Download the [PUB_KEY](https://keys.openpgp.org/search?q=sw-spark@nvidia.com).
8888
* Import the public key: `gpg --import PUB_KEY`
8989
* Verify the signature for Scala 2.12 jar:
90-
`gpg --verify rapids-4-spark_2.12-25.04.0.jar.asc rapids-4-spark_2.12-25.04.0.jar`
90+
`gpg --verify rapids-4-spark_2.12-25.06.0.jar.asc rapids-4-spark_2.12-25.06.0.jar`
9191
* Verify the signature for Scala 2.13 jar:
92-
`gpg --verify rapids-4-spark_2.13-25.04.0.jar.asc rapids-4-spark_2.13-25.04.0.jar`
92+
`gpg --verify rapids-4-spark_2.13-25.06.0.jar.asc rapids-4-spark_2.13-25.06.0.jar`
9393

9494
The output of signature verify:
9595

9696
gpg: Good signature from "NVIDIA Spark (For the signature of spark-rapids release jars) <sw-spark@nvidia.com>"
9797

9898
### Release Notes
99-
* Support approx_count_distinct
100-
* Support group by on binary type
101-
* Support ArrayPosition function
102-
* Support Databricks 14.3 ML LTS (without support for Deletion Vector reads in Delta Lake)
103-
* Support Slice
104-
* Enable Hive text writer
105-
* Refine split-retry logs when out of memory happens to expose the real reason
106-
* Allow BigSizedJoinIterator#buildPartitioner to produce more sub-partitions to avoid CudfColumnSizeOverflowException
107-
108-
Note: There is a known issue in the 25.04.0 release when decompressing gzip files on H100 GPUs.
99+
* Support functions that have time zones with daylight savings transitions
100+
* Support Spark Connect for Spark 3.5.6+ (Spark Connect is supported for Spark 3.4-3.5.5 if the plugin jar is built with a single shim)
101+
* Support for Iceberg 1.6.1 on Spark 3.5.x for read with deletions
102+
* Support array_distinct
103+
* Support bit_count
104+
* Support bitwise aggregate functions (bit_and, bit_or and bit_xor) in groupby and reduction
105+
* Support conv
106+
* Support sha1
107+
* Introduce more metrics for troubleshooting (max writers number, memory bookkeepings)
108+
* Fix a GPU Out-of-memory bug when spark.speculation is on
109+
* Fix a get_json_object bug when encounter some special pattern path
110+
111+
Note: There is a known issue in the 25.06.0 release when decompressing gzip files on H100 GPUs.
109112
Please find more details in [issue-16661](https://github.com/rapidsai/cudf/issues/16661).
110113

111114
For a detailed list of changes, please refer to the

0 commit comments

Comments
 (0)