IBM Cloud First To Offer NVIDIA’s New “Tesla” Accelerator, Speeding Big Data, Analytics and Visualization for AI and Cognitive Workloads
IBM Cloud delivers groundbreaking performance running GPU accelerators for tackling the toughest compute-intensive workloads
Las Vegas – May 19, 2016 – IBM has become the first company to offer the latest NVIDIA GPU accelerator—the NVIDIA® Tesla® M60—in the cloud, giving companies of all sizes easier and more affordable access to the latest virtual desktop and supercomputing resources along with helping enable the latest AI and cognitive applications.
By provisioning IBM Cloud bare metal servers with Tesla M60 GPU accelerators, customers can deploy fewer, more powerful cloud servers while churning through complex jobs faster, speeding a range of compute-intensive workloads, including data analytics, graphics, energy exploration and deep learning/artificial intelligence. IBM continues to be in lock step with NVIDIA in bringing its latest GPU technology to the cloud, which is driving the next wave of innovation and real-time collaboration.
The Tesla M60 with NVIDIA GRID™ virtualization technology is designed to help accelerate virtualized desktop applications especially in the area of CAD/CAM (computer-aided design and computer-aided manufacturing) including AutoCAD. Companies can spin up these GPU resources on IBM Cloud on an on-demand basis to help cut their processing time from days (or weeks) to down to hours, compared to using CPU-only based servers.
The Tesla M60 joins the other NVIDIA GPU offerings on IBM Cloud, including the Tesla K80, GRID K2 and Tesla K10.
“With NVIDIA GPU technology on IBM Cloud, we are one step closer to offering supercomputing performance on a pay-as-you-go basis, which makes this new approach to tackling big data problems accessible to customers of all sizes,” says Jerry Gutierrez, HPC leader for SoftLayer, an IBM Company. “We are at an interesting inflection point in our industry, where GPU technology is opening the door for the next wave of breakthroughs across multiple industries.”
In a cloud environment, GPUs work in conjunction with a server’s CPU to accelerate application performance. The CPU offloads compute-intensive portions of the application to the GPU, which processes large blocks of data at one time rather than sequentially boosting the overall performance in a server environment. GPUs are better for high performance computing than CPUs alone because of the thousands of efficient, high-performance cores designed to process information faster.
This performance boost has allowed IBM Cloud clients like M+W Group and MapD to achieve groundbreaking results. By using IBM Cloud bare metal servers infused with NVIDIA GPU technology, M+W Group has centralized their infrastructure while providing global access to their management system, and MapD has created a super-high-speed database and visualization platform that filters and correlates multiple dimensions of multi-billion row datasets in milliseconds, without lag.
M+W Group Fosters Global Collaboration and Increases Productivity
Engineering consultation company M+W Group offers concept, design and turnkey solutions across a range of engineering disciplines, from architecture to process systems. The company relies on numerous applications, including Autodesk, which have a high volume of users working simultaneously on large, complex and technically challenging projects.
The company recently decided to start virtualizing its applications and move to the cloud. It could no longer justify the continual investments required in upgrading expensive workstations and other hardware components, so M+W Group turned to the IBM Cloud to develop a centralized, virtual management system that would enable them to monitor each project, provide resources on demand, and forecast future scalability.
MapD Enables Lightning-Fast Data Exploration
With a mission to make big data exploration visually interactive and insightful, MapD uses NVIDIA Tesla K80 GPU accelerators running on SoftLayer bare metal servers to build its super-high-speed database and visualization platform. The massive bandwidth and parallelism delivered by NVIDIA GPUs enable the MapD database to filter and correlate multiple dimensions of multi-billion row datasets in milliseconds, without lag.
“Today, everyone is being inundated and overwhelmed by data and today’s database systems are simply to slow to handle it,” says Todd Mostak, CEO and founder of MapD. “Imagine harnessing the parallel processing power of GPU acceleration for a big data platform that can deliver up to a 100x performance increase—all via the global IBM Cloud. This is what we’ve done with MapD, a platform that can perform massive calculations in real-time and provide visualization tools for decision makers to gain instant insights and context from their data, regardless of a user’s location.”
MapD analyzes multi-billion row datasets in just milliseconds, but lightning speed is only part of the equation. By harnessing the power of GPU acceleration in the scalable cloud, MapD is also able offer this supercomputing performance on a pay-as-you-go basis, which makes this new approach to tackling big data problems accessible to customers of all sizes. Furthermore, the MapD platform can be deployed in public or private cloud, in virtualized environments, on-premise and in combinations of any of these scenarios.
About IBM Cloud:
For more information, visit: http://www.ibm.com/cloud-computing.