Aceitamos criptomoeda — para que as moedas que detém possam comprar material informático real.Pague em cripto, gaste-a em material informático real. Envio seguro gratuito para todo o mundo acima de 399 $. Caixas neutras, sem marca, expedidas a partir da Alemanha.Envio seguro gratuito acima de 399 $. A sua taxa fica bloqueada durante 30 minutos assim que abre a finalização da compra.Taxa bloqueada 30 minutos. Cada artigo é selado, o seu número de série é verificado, e está coberto por uma garantia de 24 meses.Selado, garantido 24 meses.

Artigos guardados Iniciar sessão
NVIDIA Tesla V100 GPU accelerator SXM2 module passive cooler 32 GB HBM2 memory

NVIDIA

NVIDIA V100 32 GB HBM2 graphics card

SKU CH-D6Q4-HXQJB MPN V100

Data-centre GPU with 32 GB HBM2 memory and Volta architecture for AI, HPC and graphics workloads

  • Chipset ManufacturerNVIDIA
  • GPUV100
  • Memory TypeHBM2
  • OpenGLOpenCL
  • Form FactorSXM2
  • Massive 32 GB HBM2 memory feeds large AI models without swapping
  • Volta architecture delivers up to 100 CPU-equivalent performance in one GPU
  • SXM2 form factor enables dense server deployment for HPC clusters
  • Passive cooler runs silently in rack environments with directed airflow
  • OpenCL, OpenACC and DirectCompute support covers major compute APIs

Data-center AI accelerator

The NVIDIA Tesla V100 32 GB HBM2 is a data-center GPU built on the Volta architecture for AI training, high-performance computing and professional graphics workloads. Its SXM2 form factor targets server platforms that support this mounting standard. The card delivers 32 GB of high-bandwidth memory to handle large models and datasets.

Passive SXM2 module

A passive cooler means the card relies on the host system's airflow, so deployment requires a compatible server chassis with directed cooling. The SXM2 connector provides a high-speed link to the motherboard without traditional PCIe brackets or external power cables. This design suits dense multi-GPU configurations in qualified OEM systems.

Upgrade from smaller HBM2 cards

The 32 GB capacity lets you train models or run simulations that exceed the memory limits of lower-capacity HBM2 variants. Moving to this accelerator is worthwhile when workload size demands the larger frame buffer and the platform already supports SXM2 power and cooling infrastructure.

Pontos fortes

  • Massive 32 GB HBM2 memory feeds large AI models without swapping
  • Volta architecture delivers up to 100 CPU-equivalent performance in one GPU
  • SXM2 form factor enables dense server deployment for HPC clusters
  • Passive cooler runs silently in rack environments with directed airflow
  • OpenCL, OpenACC and DirectCompute support covers major compute APIs

Especificações

BrandNVIDIA
SeriesTESLA V100 32GB
ModelTESLA V100 32GB
Chipset ManufacturerNVIDIA
GPUV100
Memory TypeHBM2
OpenGLOpenCL
Form FactorSXM2

Perguntas sobre este artigo

Will this card fit in a standard PCIe desktop slot?

The card uses the SXM2 form factor which requires a compatible server motherboard socket and cannot be installed in a standard PCIe slot.

What cooling solution do I need for this GPU?

The card ships with a passive cooler so it relies on high‑volume forced airflow from the server chassis; no fan is included.

Which programming APIs are supported for compute workloads?

The accelerator supports CUDA, OpenCL, OpenACC and DirectCompute for AI, HPC and graphics workloads.

Também nesta secção

Os clientes também compararam

Mesma prateleira, mesma finalização da compra — oito moedas e uma taxa bloqueada durante 30 minutos.