TOTAL: £0.00


Search Right
Need help
Need help?
Contact us on
+44 1672 511166
News Bytes - Latest Edition


Featured product
Agilex 9 Direct RF FPGA SFF Module

NVIDIA GPGPUs – Not Just for Gamers?


Posted by Laura B
May 21, 2021
In a previous blog post we looked at the fundamentals of GPGPUs. In this article we take a deeper dive into why NVIDIA GPGPUs are a great fit for defence and military applications, as well as projects requiring image processing or cutting-edge technology such as real-time ray tracing.

When GPUs were introduced to the wider market in 1999 they were targeted for use in modern films and games, specifically to process and render images and video. Following a period of implementation, engineers discovered they could be used successfully to convert vast amounts of raw data into intelligence or information. These use cases birthed the GPGPU – General Purpose Graphics Processing Unit.

Despite the extraordinary capability of GPUs, they have been overlooked in defence programs. Not not because of their performance, but because traditionally GPGPUs used in gaming are upgraded every 18-24 months, making continuity of supply a real concern.  To answer this problem NVIDIA has made significant investment in their embedded sector and now offers availability up to 10 years in some cases.

Many of the technical benefits of NVIDIA GPGPUs that are so widely recognised in gaming platforms, such as low latency, advanced rendering capability and support for cutting edge GDDR6 memory makes them an excellent choice for aerospace and defence applications, which require significant video data processing ability including video stabilisation, image processing, terrain analysis, object tracking, and even 3D visualisation of data.

In addition to the hardware itself, the CUDA® framework provides the industry’s most advanced and efficient GPGPU framework, opening new doors to the embedded military environment for highly efficient, near real-time, processing of data on platforms. 

Turing’ is the codename for the latest GPGPU microarchitecture from NVIDIA and is named after British mathematician and cryptanalyst Alan Turing. The Turing architecture introduces the first consumer products capable of real-time ray tracing, which has been a goal for the computer graphics industry for a long time. According to the NVIDIA website, ray tracing ‘calculates the colour of pixels by tracing the path that light would take if it were to travel from the eye of the viewer through the virtual 3D scene’. This is a key feature in next gen gaming consoles like the PS5.

Powered by NVIDIA Turing™ architecture, the Condor GR5-RTX5000 delivers exceptional power and performance to a defence market eager to expose the vast potential of artificial intelligence (AI). This brand new board from EIZO Rugged Solutions has been designed for latency-sensitive applications supporting ISR, Degraded Visual Environments (DVE), Digital Signal Processing (DSP), Electronic Warfare (EW), Signals Intelligence (SIGINT) and Data Science projects. 

The Condor GR5-RTX5000 offers up to 9.4 TFLOPs of FP32 floating point performance, 16 GB GDDR6 memory, 48 RT Ray-Tracing Cores for real-time rendering of photorealistic objects and 384 Tensor Cores for deep learning training and AI inferencing. With 3072 CUDA cores for parallel processing, this GPGPU card also supports advanced shading technologies such as Mesh, Texture, and Variable Rate Shading.

For more information about the Condor GR5-RTX5000 or about a new GPGPU application you are working on please get in touch via email – or give us a call: +44 1672 511166.


May 2024 - Sarsen Technology Newsletter

Check out all the latest embedded technology news and product releases from Sarsen Technology and our partners.