Nick Winkworth

After The Moon-Shot: GPUs In The Enterprise

Blog Post created by Nick Winkworth Employee on Oct 31, 2018

space-pan-crop4.jpg

Perhaps some of you have watched the recent movie “First Man” – about the US space program’s “Apollo” mission to land a man on the moon in 1969. The movie focuses on the human cost, but equally on the technological advances required to make that mission a success – an amazing amount of new technology had to be developed to make the dream of landing a man on the Moon a reality.

 

As is well known, much of the science and technology developed for the Apollo program was later transferred to the civilian world and used to solve problems their inventors could never have dreamed of – often far outside the realms of space exploration.

 

Most people will think of the non-stick coating used on cookware (although technically this was invented prior to Apollo), but there are many examples. Imaging technologies used to select lunar landing sites went on to be used by the cosmetics industry to measure skin surface structure and improve cosmetics. Other technologies found applications on fields as diverse as athletic footwear, cordless power tools, alternative energy and even hog farming!

 

“How does this relate to IT?”, you ask. IT is transforming to address new challenges like artificial intelligence (AI) and machine learning, and faces the daunting task of analyzing the huge quantity of data generated by web based transactions, social media and the internet-of-things (IoT). New challenges demand new solutions – and what better than to re-purpose a solution that was developed to address a completely different problem – in this case, computer gaming!

 

Gamers want realistic graphics, and that requires every pixel on a screen to be changed as fast as possible. If each pixel were to be addressed in turn, even by a powerful CPU, processing a whole screen would take far too long and the aliens would have won before you saw them coming!

 

game-over-screen-2.jpg

Graphic Processing Units (GPUs) were developed to solve this. By using many tiny processor cores, hundreds of pixels can be processed simultaneously, so the screen refreshes faster, the video looks smooth …and you beat the aliens!

 

Fast forward to today and look at all the data coming from, for example, a transportation system or a manufacturing line – thousands of tiny packets of information flooding into the data center needing to be processed at the same time. If that reminds you of processing data for thousands of pixels on a screen, you would not be alone!

 

It shouldn’t surprise you then to learn that IT has been quick to exploit GPU technology developed for the gaming industry for data analysis, and in support of those trends Hitachi Vantara is introducing a new platform to enable those solutions. The Hitachi Advanced Server DS225.

 

On the surface the DS225 looks like a pretty standard 2U dual socket rack mount server, but under the covers is some exceptional capability - it can support up to 4 NVIDIA® Tesla® GPUs including the M10, M60, P40 and V100 models optimized for different applications and use cases.

Reeder_170608_0136.jpgDS225 top copy2.jpgHitachi Advanced Server DS225

 

According to NVIDIA®, data scientists and researchers can use their NVIDIA® Tesla® GPUs to parse petabytes of data, orders of magnitude faster than they could by using traditional CPUs, in applications ranging from energy exploration to deep learning, enabling them to tackle challenges that were once thought impractical to take on, due to the sheer amount of data that needed to be analyzed. Thanks to the addition of GPU technology, an astounding amount of mathematical processing power is now available to the data center.

 

The Hitachi Advanced Server DS225 with NVIDIA® Tesla® GPUs will be available with the Hitachi Unified Compute Platform CI (UCP CI) converged systems and Hitachi Unified Compute Platform HC (UCP HC) hyperconverged systems to address different environments, use cases and applications.

 

To learn more, come see us in booth P311 at VMworld Barcelona!

 

Related: Hu Yoshida’s blog, Monique Lucey’s blog

Outcomes