Abaco Systems has developed a rugged 3U VPX card that can be used as a VITA 65 OpenVPX graphics card or as a high-performance computer (HPC) form factor with a graphics output board for the SOSA (Sensor Open Systems Architecture) standard.
The GRA115S is based on Nvidia’s Turing architecture using either the Nvida RTX5000 or RTX3000 platforms. Both versions take full advantage of the Nvidia GPUDirect RDMA to minimize latency for real-time video and data streaming.
Both choices of GPU also deliver real-time performance for encoding applications with dedicated H.265 and H.264 encode and decode engines with up to 9.49TFLOPs at peak performance.
The GRA115S also offers a PCI Gen 4 switch and high-speed, high resolution multi-head capability for GPGPU computing, AI inferencing, deep learning, sensor processing, and data analytics.
The GRA115 HPC card reflects the needs of the rugged military and aerospace market and industrial/data sciences markets in the US and Europe for high performance GPGPU computing and graphics generation applications.
With two form-factor options available, using the VITA65 open standard lowers overall risk for the end user while fitting into the 3U ecosystem. The GRA115S also uses the AXIS ImageFlex Support which provides ease and acceleration of application development.
“The GRA115S adds a powerful and flexible graphics output board, specifically aligned to the SOSA standard, to our family of designed to align products. It demonstrates Abaco’s continual industry leadership in adapting the latest GPU technologies for the rugged defense, aerospace, and industrial market sectors,” said Pete Thompson, vice president of product management for Abaco Systems.
- 3U VPX AI board with Nvidia Turing GPU
- First OpenVPX 3U Ampere graphics card
- Rugged XMC with AMD Radeon GPU and five graphics outputs
Other articles on eeNews Europe
- Ruag technology boosts Webb space telescope
- Embedded trends for 2022
- Clinical-grade vital-signs AFE for disease detection
- US sues Nvidia over ARM deal
- €100m project to develop low power edge AI microcontroller
- Raspberry Pi prepares for IPO