To The Editor,
Recently, the Military & Aerospace Electronics Web site reported that the “first NVIDIA GPGPU OpenVPX engine…” had just been introduced. We were surprised to see this story coming from an authoritative publication like yours, especially since the article “High-performance computing benefits signal and data processing in aerospace and defense applications,” which details Mercury’s NVIDIA general-purpose graphics processing unit (GPGPU) OpenVPX solution and our customer’s use of the technology, ran three days earlier.
This mistake mandates that we correct some key facts for your readers. Mercury Computer Systems first implemented GPGPUs in 2004.
In June 2010, Mercury announced we were shipping the GSC6200, which is actually the industry’s first fielded 6U OpenVPX GPGPU solution. Hailed as ground-breaking by the industry and receiving an Editor’s Choice Award in July 2010, this rugged conduction-cooled 6U OpenVPX GPGPU subsystem is currently deployed in an airborne defense platform.
On May 25, 2011, Mercury announced significant advances in radar subsystem performance enhanced by upgrading the GSC6200 with support for NVIDIA’s latest Fermi-class architecture. The GSC6200 now delivers over a teraflop of processing power in a single 6U slot.
We appreciate the opportunity to set the record straight and to continue working closely with Military & Aerospace Electronics to bring critical and accurate information to the industry.
Director, Marketing and Corporate Communications, Mercury Computer Systems, Chelmsford, Mass.