Quantcast

Tesla GPUs continue to drive ground-breaking advantages for all types of industries with demanding computational requirements. The financial industry is the latest example.

Yesterday, we announced that J.P. Morgan, the world’s largest investment bank, is using Tesla GPUs to speed up its risk calculations across a range of products. How much, you ask? J.P. Morgan saw 40X speedups, thanks to Tesla silicon.  Furthermore, peak performance increases were even higher — as much as 100 times faster than using CPUs alone.  This performance boost gives J.P. Morgan the ability to construct much more complex mathematical models, leading to better insights and more confident decision-making.

This not only allows J.P. Morgan to run these processing-intensive calculations in minutes instead of hours. It also lets them slash the cost of running these calculations by a whopping 80 percent. What’s more, it helps make their data centers greener by providing much higher performance for the same power. (Power consumption is a real concern, as we previously blogged about here).

There is a great article by the Wall Street Journal that provides more detail:  http://on.wsj.com/qRajiX

Check it out and let us know what you think!

 

 

  • http://www.facebook.com/profile.php?id=1662954291 Dolan Murvihill

    Not surprised; finance has a lot of the same math as graphics processing (linear algebra, etc.). Props to Nvidia and Chase! Were the Teslas modified at all?

  • http://www.nvidia.com George Millington

    Thanks for your comment.  No, no modifications.  JP Morgan used standard servers with standard Tesla GPUs.

  • http://www.facebook.com/profile.php?id=100001293060152 David Dresden

    They also have Liquid Nitrogen Overclocking, Inc. to thank for their 5.0 GHz product line. Where was that mentioned in the article? (See http://www.LiquidNitrogenOverclocking.com for more info).