- GPU: NVIDIA GM107
- Video memory
– 2 GB GDDR5 SDRAM
– 128-bit wide memory interface
- Dual independent channels
- RGB output: VESA resolutions up to 1920×1200 @ 60 Hz
- Digital output: DVI 1.0 resolutions up to 1920×1200 @ 60 Hz
- Form factor: 3U OpenVPX
- Fabric Interface
– Interconnection between GPU and CPU
– 16-lane PCI Express interface, Gen 3 capable
• Level 1: 0°C to +55°C air-cooled
• Level 2: -20°C to +65°C air-cooled
• Level 4: -40°C to +75°C conduction-cooled
• Level E: -40°C to +80°C conduction-cooled
• OpenGL 4.1 and DirectX 11 drivers for Microsoft
• Windows & Linux running on Intel host card
- Power Requirements: +5V / 3V3_Aux required
Superior performance, small footprint
For both video and graphics generation and general purpose computing (GPGPU), the GRA113 3U VPX graphics board brings NVIDIA®’s ‚Maxwell™’ graphics processing power to the rugged military and aerospace market. With 640 processing cores, single- and double-precision floating point units, improved shared memory architecture and cache hierarchy, together with faster atomic operations, the GRA113’s GM107 GPU is capable of CUDA Compute Capability v5.0.
Delivers significant gains to radar, video, image processing application
There is ever-greater demand for on-platform capabilities – dealing with the vast quantity of data that sensors are generating, and turning that into actionable intelligence. GPGPU offers unparalleled performance per Watt, programmable in high-level language, allowing developers to reduce their time to market and provide significantly enhanced systems capabilities.
Flexibility and options for the perfect fit
The GRA113 is offered in several variants with differing pin-outs and I/O functionality, and is pin-compatible with previous generations of product. It is available in four of Abaco Systems’ rugged build levels, and in a two-level maintenance (2LM) VITA 48-REDI variant with rear covers. The GRA113 may be deployed in the MAGIC1 rugged display computer to provide a fast-to-market solution for high-performance graphics and GPGPU signal processing applications.