NVIDIA A800 - graphics card - A800 - 40 GB

SKU 8D6C0AA
Product Description:NVIDIA A800 - graphics card - A800 - 40 GB
Device Type:Graphics card
Bus Type:PCI Express 4.0 x16
Graphics Engine:NVIDIA A800
Memory:40 GB HBM2
CUDA Cores:6912
Memory Interface:5120-bit
07-01-2025
  • Specifications
  • Description
General
Device Type:Graphics card
Bus Type:PCI Express 4.0 x16
Graphics Engine:NVIDIA A800
CUDA Cores:6912
Features:432 NVIDIA Tensor Cores, dual slot width, NVIDIA NVLink technology, 19.5 Tflops single-precision floating point performance, 9.7 Tflops double-precision floating point performance, 623.8 Tflops peak tensor performance
Memory
Size:40 GB
Technology:HBM2
Bus Width:5120-bit
Bandwidth:1555 GBps
Miscellaneous
Power Consumption Operational:240 Watt
Depth:26.7 cm
Height:11.2 cm
vis mere
The NVIDIA A800 graphics card by HP Inc. sets a new standard for video cards, offering performance and efficiency for professionals and enthusiasts alike. With its impressive 19.5 Tflops of single-precision and 9.7 Tflops of double-precision floating point performance, it's designed to handle the most demanding computing tasks with ease. The card's 40 GB of HBM2 memory and a 5120-bit bus width provide a bandwidth of 1555 GBps, ensuring smooth and efficient data transfer for high-resolution projects.
In addition to its powerful computing capabilities, the NVIDIA A800 boasts advanced features for AI and machine learning, including 432 NVIDIA Tensor Cores and 623.8 Tflops peak tensor performance. This makes it an ideal choice for professionals working on AI applications. The card's support for PCI Express 4.0 x16 interface and NVIDIA NVLink technology allows for high-speed connectivity and the ability to link multiple GPUs, further enhancing its performance. With a dual slot width and a power consumption of 240 W, the NVIDIA A800 combines high performance with energy efficiency, making it a choice for those seeking to push the boundaries of what's possible.
  • 19.5 Tflops single-precision floating point performance for demanding computing tasks
  • 40 GB of HBM2 memory for high-speed data processing
  • 5120-bit bus width and 1555 GBps bandwidth for superior performance
  • 432 NVIDIA Tensor Cores for advanced AI and machine learning capabilities
  • PCI Express 4.0 x16 interface and NVIDIA NVLink technology for robust connectivity




The technical details come from 3rd party. edgemo is not responsible for any errors.

Loading...