When working with big data or complex algorithms, we often look to parallelize our code to optimize runtime. By taking advantage of a GPUs 1000+ cores, a data scientist can quickly scale out solutions inexpensively and sometime more quickly than using traditional CPU cluster computing. In this webinar, we will present ways to incorporate GPU computing to complete computationally intensive tasks in both Python and R.
See the full presentation here: 👉 https://vimeo.com/153290051
Learn more about the Domino data science platform: https://www.dominodatalab.com
Predicting Salary Using Data Science: A Comprehensive Analysis.pdf
GPU Computing for Data Science
1. GPU Computing for Data Science
John Joo
john.joo@dominodatalab.com
Data Science Evangelist @ Domino Data Lab
2. Outline
• Why use GPUs?
• Example applications in data science
• Programming your GPU
3. Case Study:
Monte Carlo Simulations
• Simulate behavior when randomness
is a key component
• Average the results of many
simulations
• Make predictions
4. Little Information in One “Noisy Simulation”
Price(t+1) = Price(t) e InterestRate•dt + noise
5. Many “Noisy Simulations” ➡ Actionable Information
Price(t+1) = Price(t) e InterestRate•dt + noise
6. Monte Carlo Simulations Are Often Slow
• Lots of simulation data is required to
create valid models
• Generating lots of data takes time
• CPU works sequentially
7. CPUs designed for sequential, complex tasks
Source: Mythbusters https://youtu.be/-P28LKWTzrI
8. GPUs designed for parallel, low level tasks
Source: Mythbusters https://youtu.be/-P28LKWTzrI
9. GPUs designed for parallel, low level tasks
Source: Mythbusters https://youtu.be/-P28LKWTzrI
10. Applications of GPU Computing in Data Science
• Matrix Manipulation
• Numerical Analysis
• Sorting
• FFT
• String matching
• Monte Carlo simulations
• Machine learning
• Search
Algorithms for GPU Acceleration
• Inherently parallel
• Matrix operations
• High FLoat-point Operations Per Sec
(FLOPS)
11. GPUs Make Deep Learning Accessible
Google
Datacenter
Stanford AI Lab
# of machines 1,000 3
# of CPUs or
GPUs
2,000 CPUs 12 GPUs
Cores 16,000 18,432
Power used 600 kW 4 kW
Cost $5,000,000 $33,000
Adam Coates, Brody Huval,Tao Wang, David Wu, Bryan Catanzaro, Ng Andrew ; JMLR W&CP 28 (3) : 1337–1345, 2013
12. CPU vs GPU Architecture:
Structured for Different Purposes
CPU
4-8 High Performance Cores
GPU
100s-1000s of bare bones cores
13. Both CPU and GPU are required
CPU GPU
Compute intensive
functions
Everything else
General Purpose GPU Computing (GPGPU)
Heterogeneous Computing
14. Getting Started: Hardware
• Need a computer with GPU
• GPU should not be operating your
display
Spin up a GPU/CPU computer with 1 click.
8 CPU cores, 15 GB RAM
1,536 GPU cores, 4GB RAM
16. Programming CPU
• Sequential
• Write code top to bottom
• Can do complex tasks
• Independent
Programming GPU
• Parallel
• Multi-threaded - race conditions
• Low level tasks
• Dependent on CPU
Getting Started: Software
17. Talking to your GPU
CUDA and OpenCL are GPU computing frameworks
18. Choosing How to Interface with GPU:
Simplicity vs Flexibility
Application
specific
libraries
General
purpose GPU
libraries
Custom
CUDA/
OpenCL code
Flexibility
Simplicity
Low
Low
High
High
19. Application Specific Libraries
Python
• Theano - Symbolic math
• TensorFlow - ML
• Lasagne - NN
• Pylearn2 - ML
• mxnet - NN
• ABSsysbio - Systems Bio
R
• cudaBayesreg - fMRI
• mxnet - NN
• rpud -SVM
• rgpu - bioinformatics
Tutorial on using Theano, Lasagne, and no-learn:
http://blog.dominodatalab.com/gpu-computing-and-deep-learning/
20. General Purpose GPU Libraries
• Python and R wrappers for basic matrix
and linear algebra operations
• scikit-cuda
• cudamat
• gputools
• HiPLARM
• Drop-in library
22. Custom CUDA/OpenCL Code
1. Allocate memory on the GPU
2. Transfer data from CPU to GPU
3. Launch the kernel to operate on the CPU
cores
4. Transfer results back to CPU
23. Example of using Python and CUDA:
Monte Carlo Simulations
• Using PyCuda to interface Python and
CUDA
• Simulating 3 million paths, 100 time steps
each
24. Python Code for CPU
Python/PyCUDA Code for GPU
8 more lines of code
25. Python Code for CPU
Python/PyCUDA Code for CPU
1. Allocate memory on the GPU
26. Python Code for CPU
Python/PyCUDA Code for CPU
2. Transfer data from CPU to GPU
27. Python Code for CPU
Python/PyCUDA Code for CPU
3. Launch the kernel to operate on the CPU cores
28. Python Code for CPU
Python/PyCUDA Code for CPU
4. Transfer results back to CPU
29. Python Code for CPU
26 sec
Python/PyCUDA Code for CPU
8 more lines of code
1.5 sec
17x speed up
30. Some sample Jupyter notebooks
• https://app.dominodatalab.com/johnjoo/gpu_examples
• Monte Carlo example using PyCUDA
• PyCUDA example compiling CUDA C for kernel
instructions
• Scikit-cuda example of matrix multiplication
• Calculating a distance matrix using rpud
31. More resources
• NVIDIA
• https://developer.nvidia.com/how-to-cuda-python
• Berkeley GPU workshop
• http://www.stat.berkeley.edu/scf/paciorek-
gpuWorkshop.html
• Duke Statistics on GPU (Python)
• http://people.duke.edu/~ccc14/sta-663/
CUDAPython.html
• Andreas Klockner’s webpage (Python)
• http://mathema.tician.de/
• Summary of GPU libraries
• http://fastml.com/running-things-on-a-gpu/
32. More resources
• Walk through of CUDA programming in R
• http://blog.revolutionanalytics.com/2015/01/parallel-
programming-with-gpus-and-r.html
• List of libraries for GPU computing in R
• https://cran.r-project.org/web/views/
HighPerformanceComputing.html
• Matrix computations in Machine Learning
• http://numml.kyb.tuebingen.mpg.de/numl09/
talk_dhillon.pdf