Python Data Science and Machine Learning at Scale with Intel and Anaconda
Download as PPTX, PDF0 likes1,479 views
Intel and Anaconda have partnered to enhance Python's performance and scalability for data science and machine learning through optimizations and new tools. This collaboration aims to accelerate the data science pipeline utilizing Intel's oneAPI libraries and simplified installations via Anaconda. The document outlines the availability of various Intel optimizations on Anaconda's defaults channel, along with future developments like Data Parallel-Python for unified programming in Python.
![14 2021
Intel® Data Parallel C++ (DPC++) Compiler is now available on Anaconda!
• Contribute to the oneAPI Python ecosystem using DPC++ to expand
the Python ecosystem and increase Python usage across industry
hardware platforms through open-source!
• Install the DPC++ compiler using Anaconda today!
– conda install dpcpp_<your_platform>
‐ <your_platform> = linux-64 OR win-64
– Now available on the Anaconda defaults channel and the Anaconda IntelChannel
• Use the DPC++ compiler in your conda-build
recipes by configuring your meta.yaml file:
Expanding the Python ecosystem:
requirements:
build:
- {{ compiler('dpcpp') }} # [ linux or win ]](https://image.slidesharecdn.com/pythondatascienceandmachinelearningatscale-211130191047/85/Python-Data-Science-and-Machine-Learning-at-Scale-with-Intel-and-Anaconda-13-320.jpg)
![19 2021
Workloads and Configurations
See all benchmarks and configurations: https://software.intel.com/content/www/us/en/develop/articles/blazing-fast-python-data-science-ai-performance.html. Each performance claim and
configuration data is available in the body of the article listed under sections 1, 2, 3, 4, and 5. Please also visit this page for more details on all scores, and measurements derived.
Testing Date: Performance results are based on testing by Intel as of October 16, 2020 and may not reflect all publicly available updates. Configurations details and Workload Setup: 2 x Intel®
Xeon® Platinum 8280 @ 28 cores, OS: Ubuntu 19.10.5.3.0-64-generic Mitigated 384GB RAM (192 GB RAM (12x 32GB 2933). SW: Modin 0.81. Scikit-learn 0.22.2. Pandas 1.01, Python 3.8.5,
DAL(DAAL4Py) 2020.2, Census Data, (21721922.45) Dataset is from IPUMS USA, University of Minnesota, www.ipums.org [Steven Ruggles, Sarah Flood, Ronald Goeken, Josiah Grover, Erin
Meyer, Jose Pacas and Matthew Sobek. IPUMS USA: Version 10.0 [dataset], Minneapolis, MN. IPUMS, 2020. https//doc.org/10.18128/D010.V10.0]
Testing Date: Performance results are based on testing by Intel® as of October 23, 2020 and may not reflect all publicly available updates. Configuration Details and Workload Setup: Intel®
oneAPI Data Analytics Library 2021.1 (oneDAL). Scikit-learn 0.23.1, Intel® Distribution for Python 3.8; Intel® Xeon® Platinum 8280LCPU @ 270GHz, 2 sockets, 28 cores per socket, 10M samples,
10 features, 100 clusters, 100 iterations, float32.
Testing Date: Performance results are based on testing by Intel® as of October 23, 2020 and may not reflect all publicly available updates. Configuration Details and Workload Setup: Intel®
oneAPI AI Analytics Toolkit v2021.1; Intel® oneAPI Data Analytics Library (oneDAL) beta10, Scikit-learn 0.23.1, Intel® Distribution for Python 3.7, Intel® Xeon® Platinum 8280 CPU @ 2.70GHz, 2
sockets, 28 cores per socket, microcode: 0x4003003, total available memory 376 GB, 12X32GB modules, DDR4. AMD Configuration: AMD Rome 7742 @2.25 GHz, 2 sockets, 64 cores per socket,
microcode: 0x8301038, total available memory 512 GB, 16X32GB modules, DDR4, oneDAL beta10, Scikit-learn 0.23.1, Intel® Distribution for Python 3.7. NVIDIA Configuration: NVIDIA Tesla V100
– 16 Gb, total available memory 376 GB, 12X32GB modules, DDR4, Intel® Xeon Platinum 8280 CPU @ 2.70GHz, 2 sockets, 28 cores per socket, microcode: 0x5003003, cuDF 0.15, cuML 0.15,
CUDA 10.2.89, driver 440.33.01, Operation System: CentOS Linux 7 (Core), Linux 4.19.36 kernel.
Testing Date: Performance results are based on testing by Intel® as of October 13, 2020 and may not reflect all publicly available updates. Configurations details and Workload Setup: CPU:
c5.18xlarge AWS Instance (2 x Intel® Xeon® Platinum 8124M @ 18 cores. OS: Ubuntu 20.04.2 LTS, 193 GB RAM. GPU: p3.2xlarge AWS Instance (GPU: NVIDIA Tesla V100 16GB, 8 vCPUs, OS:
Ubuntu 18.04.2LTS, 61 GB RAM. SW: XGBoost 1.1: build from sources compiler – G++ 7.4, nvcc 9.1 Intel® DAAL: 2019.4 version: Python env: Python 3.6, Numpy 1.16.4, Pandas 0.25 Scikit-learn
0.21.2.](https://image.slidesharecdn.com/pythondatascienceandmachinelearningatscale-211130191047/85/Python-Data-Science-and-Machine-Learning-at-Scale-with-Intel-and-Anaconda-18-320.jpg)
![20 2021
Workloads and Configurations
Testing Date: Performance results are based on testing by Intel® as of October 26, 2020 and may not reflect all publicly available updates. Configuration Details and Workload Setup: Intel®
Optimization for Tensorflow v2.2.0; oneDNN v1.2.0; Intel® Low Precision Optimization Tool v1.0; Platform; Intel® Xeon® Platinum 8280 CPU; #Nodes 1; #Sockets: 2; Cores/socket: 28;
Threads/socket: 56; HT: On; Turbo: On; BIOS version:SE5C620.86B.02.01.0010.010620200716; System DDR Mem Config: 12 slots/16GB/2933; OS: CentOS Linux 7.8; Kernel: 4.4.240-1.el7.elrepo
x86_64.
Testing Date: Performance results are based on testing by Intel® as of February 3, 2021 and may not reflect all publicly available updates. Configuration Details and Workload Setup: Intel®
Optimization for PyTorch v1.5.0; Intel® Extension for PyTorch (IPEX) 1.1.0; oneDNN version: v1.5; DLRM: Training batch size (FP32/BF16): 2K/instance, 1 instance; DLRM dataset (FP32/BF16):
Criteo Terabyte Dataset; BERT-Large: Training batch size (FP32/BF16): 24/Instance. 1 Instance on a CPU socket. Dataset (FP32/BF16): WikiText-2 [https://www.salesforce.com/products/einstein/ai-
research/the-wiktext-dependency-language-modeling-dataset/]: ResNext101-32x4d: Training batch size (FP32/BF16): 128/Instance, 1 instance on a CPU socket, Dataset (FP32/BF16): ILSVRC2012;
DLRM: Inference batch size (INT8): 16/instance, 28 instances, dummy data. Intel® Xeon® Platinum 8380H Processor, 4 socket, 28 cores HT On Turbo ON Total memory 768 GB (24
slots/32GB/3200 MHz), BIOS; WLYDCRBLSYS.0015.P96.2005070242 (ucode: OX 700001b), Ubuntu 20.04 LTS, kernel 5.4.0-29-genen: ResNet50: [https://github.com/Intel/optimized-
models/tree/master/pytorch/ResNet50]: ResNext101 32x4d: [https://github.com/intel/optimized-models/tree/master/pytorch/ResNext101_32x4ct: DLRM: https//github.com/intel/optimized-
models/tree/master/pytorch/dlrm].
Testing Date: Performance results are based on testing by Intel® as of October 4, 2021 and may not reflect all publicly available updates. Configuration Details and Workload Setup: Hardware
(same for all configurations): 1-node, 2x 2nd Gen Intel® Xeon® Gold 6258R on Lenovo 30BC003DUS with 768GB (12 slots/ 64GB/ 2666) total DDR4 memory and 2TB (4 slots/ 512GB/ 2666)
DCPMM memory, microcode 0x5003102, HT on, Turbo on, Ubuntu 20.04.3 LTS, 5.10.0-1049-oem, 1x Samsung 1TB SSD OS Drive, 4x Samsung 2TB SSD in RAID0 data drive, 3x NVIDIA Quadro RTX
8000. 3 months of NYCTaxi Data on Stock Software Configuration: Python 3.9.7, Pandas 1.3.3, Scikit-Learn 1.0, XGBoost 0.81, IPython 7.28.0, IPKernel 6.4.1. Full 30 months of NYCTaxi Data on
Nvidia RAPIDS Software Configuration: Python 3.7.10, Pandas 1.2.5, XGBoost 1.4.2, cuDF 21.08.03, cudatoolkit 11.2.72, dask-cudf 21.08.03, dask-cuda 21.08.00, IPython 7.28.0, IPKernel 6.4.1. Full
30 months of NYCTaxi Data on Intel Optimized Software Configuration: Python 3.9.7, Pandas 1.3.3, Modin 0.11.0, OmniSci 5.7.0, Scikit-:earn 1.0, Intel® Extension for Scikit-Learn*
2021.3.0, XGBoost 1.4.2, IPython 7.28.0, IPKernel 6.4.1. NYCTaxi Dataset from New York City (nyc.gov): [https://www1.nyc.gov/site/tlc/about/tlc-trip-record-data.page]](https://image.slidesharecdn.com/pythondatascienceandmachinelearningatscale-211130191047/85/Python-Data-Science-and-Machine-Learning-at-Scale-with-Intel-and-Anaconda-19-320.jpg)
Python Data Science and Machine Learning at Scale with Intel and Anaconda
- 2. Python Data Science and Machine Learning at Scale Rachel Oberman, IntelAITechnical Consulting Engineer ToddTomashek, Intel Machine Learning Engineer Albert De Fusco, Anaconda Data Scientist October 28th, 2021
- 3. A partnership grounded in powerful speed and scale: Intel and Anaconda: BetterTogether
- 4. 2021 5 Intel and Anaconda have partnered to bring high performance Python optimizations with simple installations! Let’s take a step back and look at what this means. +
- 5. What is Intel achieving today with Anaconda? Python Scalability and Fast Performance with Intel
- 6. 7 2021 Intel’s Commitment to a More Powerful Python AI Ecosystem: • Accelerate the end-to-end Data Science pipeline at all steps for ultra fast performance and large data scalability • Utilize drop-in acceleration tools for popular AI and data frameworks (Pandas, Scikit-Learn, etc.) built using oneAPI libraries (i.e., oneMKL, oneDNN, oneCCL, oneDAL, and more) • Empower developers to build and contribute to the oneAPI Python ecosystem using DPC++ to extend Python to ever increasing industry hardware platforms in an open-performant method Taking data science to the next level:
- 7. 8 2021 A Brief Overview of Intel’s AI Python Offerings For larger scale and increased performance in data science workloads:
- 8. 9 2021 *Performance improvements shown here are based off hardware running on Intel Cascade Lake processors. This chart will be updated once data from Ice Lake is available. See backup for workloads and configurations. Results may vary.
- 9. And now, a technical preview of the power of Intel and Anaconda: Demo
- 10. Bringing Intel’s Python optimizations to life: Simple Installation and Distribution with Anaconda
- 11. 12 2021 Anaconda and its Historic Partnership with Intel • Anaconda is one of the most popular open-source package distribution and management platforms – Available for Linux, Windows, and MacOS – Try it out on your favorite Cloud Provider: Anaconda is available on Amazon Web Services, Google Cloud, Microsoft Azure, and others • Our history with Intel: Intel® oneAPI Math Kernel Library (oneMKL) has been the default BLAS library for Anaconda since 2016 – oneMKL optimizations already part of NumPy and SciPy on the defaults channel – mkl-fft, mkl-random, mkl-dnn packages also available on Anaconda provide Python interfaces with more MKL functionality Making Intel® optimizations more accessible:
- 12. 13 2021 New IntelOptimizations Available on the Anaconda defaults Channel • Intel® Extension for Scikit-Learn is now available on the Anaconda defaults channel! – Provides drop-in accelerations for Scikit-Learn’s many regression, classification, clustering, dimensionality, and utility functions with a single-line code change using Intel® oneAPI Data Analytics Library (oneDAL) – Install now: conda install scikit-learn-intelex • Other awesome Intel Data Science optimizations are also now available on the Anaconda defaults channel (with more coming soon)! – Intel® Distribution of Modin: Performant, parallel, and distributed dataframe system with infinite scalability for the Pandas API through a single line code-change, powered by OmniSci in the backend – daal4py: Simplified Python API for Intel oneAPI Data Analytics Library with machine learning optimizations that also power Intel Extension for Scikit-Learn’s drop-in accelerations – mkl_umath: Optimized loops for NumPy universal functions (ufuncs) that is utilized in NumPy and SciPy Taking Intel® optimizations to the next level with Anaconda:
- 13. 14 2021 Intel® Data Parallel C++ (DPC++) Compiler is now available on Anaconda! • Contribute to the oneAPI Python ecosystem using DPC++ to expand the Python ecosystem and increase Python usage across industry hardware platforms through open-source! • Install the DPC++ compiler using Anaconda today! – conda install dpcpp_<your_platform> ‐ <your_platform> = linux-64 OR win-64 – Now available on the Anaconda defaults channel and the Anaconda IntelChannel • Use the DPC++ compiler in your conda-build recipes by configuring your meta.yaml file: Expanding the Python ecosystem: requirements: build: - {{ compiler('dpcpp') }} # [ linux or win ]
- 14. What can we expect from this partnership in the future?
- 15. 16 2021 Data Parallel-Python (DPPY): An XPU experience for Python • Simple, unified offload programming model • Standards-based: Python Data API Standards + Khronos SYCL + extensions • Interoperates with vast Python ecosystem on host Looking to the future with Intel and Anaconda:
- 16. 17 2021 Call to Action For more details on Intel and Anaconda’s partnership, visit Intel and Anaconda Collaboration Announcement Intel and Anaconda Intel® Extension for Scikit-Learn Technical Blog Intel Partner Page on Anaconda Intel Optimized Packages Information on the Anaconda Defaults Channel For more details on specific Intel Python Data Science software options, visit Install Intel® oneAPI AI Analytics Toolkit with Anaconda Intel® oneAPI AI Analytics Toolkit Code Samples Intel® Distribution for Python Support Forum Machine Learning and Data Analytics Support Forum Install Intel Python software from Anaconda for easy, fast, and scalable data science tools!
- 17. 18 2021 Notices and Disclaimers • Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex. • Performance results are based on testing as of dates shown in configurations and may not reflect all publicly available updates. See backup for configuration details. No product or component can be absolutely secure. • Intel does not control or audit third-party data. You should consult other sources to evaluate accuracy. • Your costs and results may vary. • Intel technologies may require enabled hardware, software or service activation. • © Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.
- 18. 19 2021 Workloads and Configurations See all benchmarks and configurations: https://software.intel.com/content/www/us/en/develop/articles/blazing-fast-python-data-science-ai-performance.html. Each performance claim and configuration data is available in the body of the article listed under sections 1, 2, 3, 4, and 5. Please also visit this page for more details on all scores, and measurements derived. Testing Date: Performance results are based on testing by Intel as of October 16, 2020 and may not reflect all publicly available updates. Configurations details and Workload Setup: 2 x Intel® Xeon® Platinum 8280 @ 28 cores, OS: Ubuntu 19.10.5.3.0-64-generic Mitigated 384GB RAM (192 GB RAM (12x 32GB 2933). SW: Modin 0.81. Scikit-learn 0.22.2. Pandas 1.01, Python 3.8.5, DAL(DAAL4Py) 2020.2, Census Data, (21721922.45) Dataset is from IPUMS USA, University of Minnesota, www.ipums.org [Steven Ruggles, Sarah Flood, Ronald Goeken, Josiah Grover, Erin Meyer, Jose Pacas and Matthew Sobek. IPUMS USA: Version 10.0 [dataset], Minneapolis, MN. IPUMS, 2020. https//doc.org/10.18128/D010.V10.0] Testing Date: Performance results are based on testing by Intel® as of October 23, 2020 and may not reflect all publicly available updates. Configuration Details and Workload Setup: Intel® oneAPI Data Analytics Library 2021.1 (oneDAL). Scikit-learn 0.23.1, Intel® Distribution for Python 3.8; Intel® Xeon® Platinum 8280LCPU @ 270GHz, 2 sockets, 28 cores per socket, 10M samples, 10 features, 100 clusters, 100 iterations, float32. Testing Date: Performance results are based on testing by Intel® as of October 23, 2020 and may not reflect all publicly available updates. Configuration Details and Workload Setup: Intel® oneAPI AI Analytics Toolkit v2021.1; Intel® oneAPI Data Analytics Library (oneDAL) beta10, Scikit-learn 0.23.1, Intel® Distribution for Python 3.7, Intel® Xeon® Platinum 8280 CPU @ 2.70GHz, 2 sockets, 28 cores per socket, microcode: 0x4003003, total available memory 376 GB, 12X32GB modules, DDR4. AMD Configuration: AMD Rome 7742 @2.25 GHz, 2 sockets, 64 cores per socket, microcode: 0x8301038, total available memory 512 GB, 16X32GB modules, DDR4, oneDAL beta10, Scikit-learn 0.23.1, Intel® Distribution for Python 3.7. NVIDIA Configuration: NVIDIA Tesla V100 – 16 Gb, total available memory 376 GB, 12X32GB modules, DDR4, Intel® Xeon Platinum 8280 CPU @ 2.70GHz, 2 sockets, 28 cores per socket, microcode: 0x5003003, cuDF 0.15, cuML 0.15, CUDA 10.2.89, driver 440.33.01, Operation System: CentOS Linux 7 (Core), Linux 4.19.36 kernel. Testing Date: Performance results are based on testing by Intel® as of October 13, 2020 and may not reflect all publicly available updates. Configurations details and Workload Setup: CPU: c5.18xlarge AWS Instance (2 x Intel® Xeon® Platinum 8124M @ 18 cores. OS: Ubuntu 20.04.2 LTS, 193 GB RAM. GPU: p3.2xlarge AWS Instance (GPU: NVIDIA Tesla V100 16GB, 8 vCPUs, OS: Ubuntu 18.04.2LTS, 61 GB RAM. SW: XGBoost 1.1: build from sources compiler – G++ 7.4, nvcc 9.1 Intel® DAAL: 2019.4 version: Python env: Python 3.6, Numpy 1.16.4, Pandas 0.25 Scikit-learn 0.21.2.
- 19. 20 2021 Workloads and Configurations Testing Date: Performance results are based on testing by Intel® as of October 26, 2020 and may not reflect all publicly available updates. Configuration Details and Workload Setup: Intel® Optimization for Tensorflow v2.2.0; oneDNN v1.2.0; Intel® Low Precision Optimization Tool v1.0; Platform; Intel® Xeon® Platinum 8280 CPU; #Nodes 1; #Sockets: 2; Cores/socket: 28; Threads/socket: 56; HT: On; Turbo: On; BIOS version:SE5C620.86B.02.01.0010.010620200716; System DDR Mem Config: 12 slots/16GB/2933; OS: CentOS Linux 7.8; Kernel: 4.4.240-1.el7.elrepo x86_64. Testing Date: Performance results are based on testing by Intel® as of February 3, 2021 and may not reflect all publicly available updates. Configuration Details and Workload Setup: Intel® Optimization for PyTorch v1.5.0; Intel® Extension for PyTorch (IPEX) 1.1.0; oneDNN version: v1.5; DLRM: Training batch size (FP32/BF16): 2K/instance, 1 instance; DLRM dataset (FP32/BF16): Criteo Terabyte Dataset; BERT-Large: Training batch size (FP32/BF16): 24/Instance. 1 Instance on a CPU socket. Dataset (FP32/BF16): WikiText-2 [https://www.salesforce.com/products/einstein/ai- research/the-wiktext-dependency-language-modeling-dataset/]: ResNext101-32x4d: Training batch size (FP32/BF16): 128/Instance, 1 instance on a CPU socket, Dataset (FP32/BF16): ILSVRC2012; DLRM: Inference batch size (INT8): 16/instance, 28 instances, dummy data. Intel® Xeon® Platinum 8380H Processor, 4 socket, 28 cores HT On Turbo ON Total memory 768 GB (24 slots/32GB/3200 MHz), BIOS; WLYDCRBLSYS.0015.P96.2005070242 (ucode: OX 700001b), Ubuntu 20.04 LTS, kernel 5.4.0-29-genen: ResNet50: [https://github.com/Intel/optimized- models/tree/master/pytorch/ResNet50]: ResNext101 32x4d: [https://github.com/intel/optimized-models/tree/master/pytorch/ResNext101_32x4ct: DLRM: https//github.com/intel/optimized- models/tree/master/pytorch/dlrm]. Testing Date: Performance results are based on testing by Intel® as of October 4, 2021 and may not reflect all publicly available updates. Configuration Details and Workload Setup: Hardware (same for all configurations): 1-node, 2x 2nd Gen Intel® Xeon® Gold 6258R on Lenovo 30BC003DUS with 768GB (12 slots/ 64GB/ 2666) total DDR4 memory and 2TB (4 slots/ 512GB/ 2666) DCPMM memory, microcode 0x5003102, HT on, Turbo on, Ubuntu 20.04.3 LTS, 5.10.0-1049-oem, 1x Samsung 1TB SSD OS Drive, 4x Samsung 2TB SSD in RAID0 data drive, 3x NVIDIA Quadro RTX 8000. 3 months of NYCTaxi Data on Stock Software Configuration: Python 3.9.7, Pandas 1.3.3, Scikit-Learn 1.0, XGBoost 0.81, IPython 7.28.0, IPKernel 6.4.1. Full 30 months of NYCTaxi Data on Nvidia RAPIDS Software Configuration: Python 3.7.10, Pandas 1.2.5, XGBoost 1.4.2, cuDF 21.08.03, cudatoolkit 11.2.72, dask-cudf 21.08.03, dask-cuda 21.08.00, IPython 7.28.0, IPKernel 6.4.1. Full 30 months of NYCTaxi Data on Intel Optimized Software Configuration: Python 3.9.7, Pandas 1.3.3, Modin 0.11.0, OmniSci 5.7.0, Scikit-:earn 1.0, Intel® Extension for Scikit-Learn* 2021.3.0, XGBoost 1.4.2, IPython 7.28.0, IPKernel 6.4.1. NYCTaxi Dataset from New York City (nyc.gov): [https://www1.nyc.gov/site/tlc/about/tlc-trip-record-data.page]