The Nvidia Tesla V100 is a data center GPU (graphics processing unit). What is means is it’s very different from the desktop GPU’s you put into a standard desktop. It has much heavier requirements (and a much heavier price tag) than a standard GPU, too. While you can buy a Geforce 10x series card (1050, 1060, 1070 etc). for a thousand or less, you’ll be forking out a lot more for the Nvidia Tesla v100. We have one of these cards listed and available here on this page (lower down). The cost of the card is approximately $9,000 USD. Why so much though – why such a high cost of this card?
Whoah – why does this GPU cost that much?
Well, this isn’t the standard GPU or video card you’d buy just to play video games, you know. While you could do this, it’s not what this card is intended for. The concept behind a datacenter graphics card like this is you get the processing power of 100 CPUs out of this one card. You could call your server a supercomputer when you put this card in it – because in all honesty, it basically does become that. This GPU was created for data scientists, researchers and engineers. Because it holds the power of 100 micro-processors inside of one card, it can therefore do the work of 100 computers. For number crunching, determining the output of new algorithms and working on advanced processing elements like the creation of artificial intelligence – that’s what a card like this is for.
I mean, you could buy it to play video games with – if you are a millionaire and don’t mind spending $20 thousand dollars on a GPU!
Note: this card is also available from Dell for about $2,000 more – $19,500.
NVIDIA Tesla V100 Volta GPU Accelerator 32GB Graphics CardNVIDIA TESLA V100 VOLTA GPU ACCELERATOR 32GB HBM2 640 TENSOR CORES 5120 CUDA CORES MEMORY BANDWIDTH 900GB/S PCI-E 3.0 X16 GENERAL PURPOSE GRAPHICS PROCESSING UNIT GPGPU MANUFACTURER: NVIDIA PART NUMBER: 699-2G500-0202-400 ENGINE SPECS: ENGINE ARCHITECTURE: VOLTA TENSOR CORES: 640 CUDA CORES: 5120 GPU CLOCK: 1230 MHZ BOOST CLOCK: 1380 MHZ DOUBLE PRECISION PERFORMANCE: 7 TFLOPS SINGLE PRECISION PERFORMANCE: 14 TFLOPS DEEP LEARNING: 112 TFLOPS MEMORY: MEMORY CONFIGURATION: 32GB MEMORY INTERFACE: HBM2 MEMORY INTERFACE WIDTH: 4096 BIT MEMORY BANDWIDTH: 900 GB/S ERROR CORRECTION CODE ( ECC ): YES ( ENABLED BY DEFAULT ) MAX MEMORY CLOCK: 876 MHZ SUPPORT: INTERCONNECT BANDWIDTH: 32GB/S BUS SUPPORT: PCI-E 3.0 PHYSICAL BUS INTERFACE: PCI-E 3.0 X16 SUPPORTED TECHNOLOGIES: CUDA TECHNOLOGY, NVIDIA GPU BOOST, DIRECTCOMPUTE, OPENCL, OPENACC, VOLTA ARCHITECTURE, TENSOR CORES OPERATING SYSTEMS: MICROSOFT WINDOWS 7, WINDOWS 8, WINDOWS 8.1, WINDOWS 10, WINDOWS SERVER 2008 R2, WINDOWS SERVER 2012 R2, WINDOWS SERVER 2016, LINUX ( ENGLISH ) THERMAL AND POWER SPECS: MAX POWER CONSUMPTION ( W ): 250W POWER CONNECTORS: ( 1 ) ONE 8 PIN POWER CONNECTOR COOLING SYSTEM: PASSIVE ( BIDIRECTIONAL AIRFLOW ) DUAL SLOT, FULL HEIGHT
|
||||||
