NVIDIA Confidential 1
Mathematics and Parallel Computing: Powering the 21st Century Industrial Revolution Shanker Trivedi, Vice President (Business Development) - NVIDIA Corporation
NVIDIA Confidential 2
Super Phones to Super Computers
NVIDIA Confidential 3
Tesla in 3 of Top 5 Super Computers
0
500
1000
1500
2000
2500
Tianhe-1A Jaguar Nebulae Tsubame Hopper II
Gig
aflo
ps
NVIDIA Confidential 4
0
1
2
3
4
5
6
7
8
0
500
1000
1500
2000
2500
Tianhe-1A Jaguar Nebulae Tsubame Hopper II
Meg
awat
ts
Gig
aflo
ps
Tesla “Green” Supercomputing
NVIDIA Confidential 5
Super Computing Across Industries
Seismic Imaging Reverse Time Migration
Automotive Design Computational Fluid Dynamics
Product Development Finite Difference Time Domain
Options Pricing Monte Carlo
Weather Forecasting Atmospheric Physics
Drug Design Molecular Dynamics
Medical Imaging Computed Tomography
Astrophysics n-body
NVIDIA Confidential 6
Gene Sequencing
NVIDIA Confidential 7
Diagnostic Imaging
NVIDIA Confidential 8
Seismic Processing
NVIDIA Confidential 9
Defense and Intelligence
NVIDIA Confidential 10
Financial Risk Management
NVIDIA Confidential 11
Automobile Design – CFD
NVIDIA Confidential 12
Languages & APIs
Developer Tools
Libraries
Fortran
Cloud Computing
CUDA Zone
EC2
CUDA Ecosystem Applications
Partners Microsoft, Dell
HP, IBM, ASUS, SuperMicro
Research & Education
CUDA C/C++ OpenCL DirectX11 Accelereyes, Acceleware, EM
Photonics, MAGMA, PGI, Rogue Wave, StoneRidge pyCUDA GPU.NET
NVIDIA Confidential 13
[Industrial Light and Magic video]
NVIDIA Confidential 14
NVIDIA Confidential 15
Better Science @ Lower Cost Using Industry Standard Parallel Computing
NVIDIA Confidential 16
[Three Pillars of Science video] http://www.nvidia.com/object/race-for-better-science.html
NVIDIA Confidential 17
Additional Materials
! “Illumination Effect in Reverse Time Migration”, Bruno Kaelin et al ! http://www.fusiongeo.com/publications/k3d/3DGeo-
IlluminationEffects_EAGE2007.pdf ! “Implementing 3D Finite Difference Codes on the GPU” by P.
Micikevicius ! Slides: http://www.nvidia.com/content/GTC/documents/1006_GTC09.pdf ! Video :
http://developer.download.nvidia.com/compute/cuda/docs/GTC09Materials.htm (talk 1006P
CUDA Resources
! Introduction (basics, assume no GPU knowledge): ! Video: http://developer.download.nvidia.com/CUDA/training/cudaintrowebinar.mp4 ! Slides: http://www.nvidia.com/content/PDF/sc_2010/CUDA_Tutorial/SC10_CUDA_C_Basics.pdf
! Fundamental optimization: ! The two presentations are largely the same, the second one has some updates but no audio/video ! Slides/video: http://www.nvidia.com/object/gtc2010-presentation-archive.html#session2011 ! http://www.nvidia.com/content/PDF/sc_2010/CUDA_Tutorial/SC10_Fundamental_Optimizations.pdf
! Advanced optimization: ! The two presentations are largely the same, the second one has some updates but no audio/video ! Slides/video: http://www.nvidia.com/object/gtc2010-presentation-archive.html#session2012 ! http://www.nvidia.com/content/PDF/sc_2010/CUDA_Tutorial/SC10_Analysis_Driven_Optimization.pdf
! Miscellaneous: ! all NVIDIA webinars: http://developer.nvidia.com/object/gpu_computing_online.html#Previously ! all NVIDIA tutorials from Supercomputing 2010: http://www.nvidia.com/object/sc10_cuda_tutorial.html
ICME Stanford May 7, 2011
Shanker Trivedi VP, Business Development