IBM and NVIDIA are partnering on a new deep learning tool to help train computers to think and learn in more human-like ways, at a faster pace. Deep learning is an emerging machine learning method that crunches through high volumes of data to detect and rank the most important aspects of that data.
Already supported by leading consumer web and mobile application companies, deep learning is being adopted by more traditional enterprises. Deep learning and other AI capabilities are being used in banking to advance fraud detection through facial recognition; in the automotive sector for self-driving automobiles and in retail for fully automated call centres that can better support the use of natural language.
The deep learning software toolkit, IBM PowerAI, is a set of binary distributions of popular deep learning frameworks including Caffe, Torch and Theano. Additional distributions include the IBM and NVIDIA versions of the Caffe deep learning frameworks, IBM-Caffe and NVCaffe.
IBM has optimised the distributions to take advantage of the recently announced IBM POWER8 chip with the NVIDIA NVLink interface that runs IBM’s highest performing server in its OpenPOWER LC lineup, the IBM Power S822LC for high performance computing (HPC). The server also features NVIDIA’s latest GPU technology.
The POWER8 with NVIDIA NVLink chip is the result of collaboration between OpenPOWER Foundation members IBM and NVIDIA. The new chip enables tight integration between IBM’s POWER8 CPU server architecture and the new Pascal architecture NVIDIA’s Tesla P100 GPU accelerators. The CPUs and GPUs integrated into the Power S822LC for HPC are connected to each other via the high-speed NVIDIA NVLink interconnect, removing potential bottlenecks created by the PCIe interface found in most Intel x86-based servers.
The hardware-software solution provides more than 2X performance over comparable servers with four GPUs running AlexNet with Berkeley Vision and Learning Center's (BVLC's) Caffe*. The four-GPU Power-based configuration running Alexnet with Caffe outperforms eight M40 GPU-based x86 configurations1, making it the world's fastest commercially available enterprise systems platform on two versions of a key deep learning framework.
IBM PowerAI also provides a continued path for Watson, IBM’s cognitive solutions platform, to extend its AI expertise in the enterprise by using several deep learning methods to train Watson.
"PowerAI democratises deep learning and other advanced analytic technologies by giving enterprise data scientists and research scientists alike an easy to deploy platform to rapidly advance their journey on AI,” said Ken King, GM, OpenPOWER Foundation. "Coupled with our high performance computing servers built for AI, IBM provides what we believe is the best platform for enterprises building AI-based software, whether it’s chatbots for customer engagement, or real-time analysis of social media data."
“Our innovation with IBM on NVIDIA NVLink has created new opportunities for POWER in the deep learning and analytics market,” said Ian Buck, VP and GM of Accelerated Computing Group at NVIDIA.
“NVIDIA’s GPUDL libraries in PowerAI will provide world class high-performance tools to power GPU-accelerated deep learning applications.”
Initial clients for the new IBM Power S822LC for HPC, launched in September, have included SC3 Electronics, a cloud supercomputing centre in Turkey. The organisation announced last month that it is creating the largest HPC cluster in the Middle East and North Africa region based on Power S822LC for HPC servers.
Interested?
IBM PowerAI is available at no charge to customers of IBM’s Power S822LC for HPC servers.
Interested?
Download IBM PowerAI at www.ibm.biz/powerai
*Based on AlexNet Training for Top-1 50% Accuracy. IBM Power S822LC for HPC configuration: 16 cores (eight cores/socket) at 4.025 GHz with 4xNvidia Pascal P100 GPUs; 512 GB memory; Ubuntu 16.04.1 running NVCaffe 0.14.5 compared to IBM Power S822L configuration: 20 cores (10 cores/socket) at 3.694 GHz with 4xNvidia M40 GPUs; 512 GB memory; Ubuntu 16.04 running BVLC-Caffe f28f5ae2f2453f42b5824723efc326a04dd16d85. Software stack details for both configurations: G++ - 5.3.1, Gfortran –5.3.1, OpenBlas - 0.2.18, Boost –1.58.0, CUDA 8.0 Toolkit, Lapack –3.6.0, Hdf5 –1.8.16, Opencv –2.4.9.
1 IBM Power S822LC for HPC configuration: 20 cores (10 cores/socket) at 3.95 GHz with 4xNvidia Pascal P100 GPUs; 512 GB memory; Ubuntu 16.04 LE running IBM version BVLC 1.0.0-rc3 compared to Intel E5-2640v4 (Broadwell): 20 cores (10 cores/socket) at 3.6 GHz with 8xNvidia M40 GPUs; 512 GB memory; Ubuntu 16.04 LE running BVLC-Caffe 985493e9ce3e8b61e06c072a16478e6a74e3aa5a. Software stack details for both configurations: G++ - 5.4, Gfortran .4, OpenBlas - 0.2.19, Boost .58.0, CUDA 8.0 Toolkit, Lapack .6.0, Hdf5 .8.16, Opencv .4.9
Already supported by leading consumer web and mobile application companies, deep learning is being adopted by more traditional enterprises. Deep learning and other AI capabilities are being used in banking to advance fraud detection through facial recognition; in the automotive sector for self-driving automobiles and in retail for fully automated call centres that can better support the use of natural language.
The deep learning software toolkit, IBM PowerAI, is a set of binary distributions of popular deep learning frameworks including Caffe, Torch and Theano. Additional distributions include the IBM and NVIDIA versions of the Caffe deep learning frameworks, IBM-Caffe and NVCaffe.
IBM has optimised the distributions to take advantage of the recently announced IBM POWER8 chip with the NVIDIA NVLink interface that runs IBM’s highest performing server in its OpenPOWER LC lineup, the IBM Power S822LC for high performance computing (HPC). The server also features NVIDIA’s latest GPU technology.
The POWER8 with NVIDIA NVLink chip is the result of collaboration between OpenPOWER Foundation members IBM and NVIDIA. The new chip enables tight integration between IBM’s POWER8 CPU server architecture and the new Pascal architecture NVIDIA’s Tesla P100 GPU accelerators. The CPUs and GPUs integrated into the Power S822LC for HPC are connected to each other via the high-speed NVIDIA NVLink interconnect, removing potential bottlenecks created by the PCIe interface found in most Intel x86-based servers.
The hardware-software solution provides more than 2X performance over comparable servers with four GPUs running AlexNet with Berkeley Vision and Learning Center's (BVLC's) Caffe*. The four-GPU Power-based configuration running Alexnet with Caffe outperforms eight M40 GPU-based x86 configurations1, making it the world's fastest commercially available enterprise systems platform on two versions of a key deep learning framework.
IBM PowerAI also provides a continued path for Watson, IBM’s cognitive solutions platform, to extend its AI expertise in the enterprise by using several deep learning methods to train Watson.
"PowerAI democratises deep learning and other advanced analytic technologies by giving enterprise data scientists and research scientists alike an easy to deploy platform to rapidly advance their journey on AI,” said Ken King, GM, OpenPOWER Foundation. "Coupled with our high performance computing servers built for AI, IBM provides what we believe is the best platform for enterprises building AI-based software, whether it’s chatbots for customer engagement, or real-time analysis of social media data."
“Our innovation with IBM on NVIDIA NVLink has created new opportunities for POWER in the deep learning and analytics market,” said Ian Buck, VP and GM of Accelerated Computing Group at NVIDIA.
“NVIDIA’s GPUDL libraries in PowerAI will provide world class high-performance tools to power GPU-accelerated deep learning applications.”
Initial clients for the new IBM Power S822LC for HPC, launched in September, have included SC3 Electronics, a cloud supercomputing centre in Turkey. The organisation announced last month that it is creating the largest HPC cluster in the Middle East and North Africa region based on Power S822LC for HPC servers.
Interested?
IBM PowerAI is available at no charge to customers of IBM’s Power S822LC for HPC servers.
Interested?
Download IBM PowerAI at www.ibm.biz/powerai
*Based on AlexNet Training for Top-1 50% Accuracy. IBM Power S822LC for HPC configuration: 16 cores (eight cores/socket) at 4.025 GHz with 4xNvidia Pascal P100 GPUs; 512 GB memory; Ubuntu 16.04.1 running NVCaffe 0.14.5 compared to IBM Power S822L configuration: 20 cores (10 cores/socket) at 3.694 GHz with 4xNvidia M40 GPUs; 512 GB memory; Ubuntu 16.04 running BVLC-Caffe f28f5ae2f2453f42b5824723efc326a04dd16d85. Software stack details for both configurations: G++ - 5.3.1, Gfortran –5.3.1, OpenBlas - 0.2.18, Boost –1.58.0, CUDA 8.0 Toolkit, Lapack –3.6.0, Hdf5 –1.8.16, Opencv –2.4.9.
1 IBM Power S822LC for HPC configuration: 20 cores (10 cores/socket) at 3.95 GHz with 4xNvidia Pascal P100 GPUs; 512 GB memory; Ubuntu 16.04 LE running IBM version BVLC 1.0.0-rc3 compared to Intel E5-2640v4 (Broadwell): 20 cores (10 cores/socket) at 3.6 GHz with 8xNvidia M40 GPUs; 512 GB memory; Ubuntu 16.04 LE running BVLC-Caffe 985493e9ce3e8b61e06c072a16478e6a74e3aa5a. Software stack details for both configurations: G++ - 5.4, Gfortran .4, OpenBlas - 0.2.19, Boost .58.0, CUDA 8.0 Toolkit, Lapack .6.0, Hdf5 .8.16, Opencv .4.9
posted from Bloggeroid
No comments:
Post a Comment