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
Dual-slot full-height PCIe card with passive heatsink

NVIDIA

NVIDIA Tesla P100 16GB HBM2 PCIe graphics card

SKU CH-C828-ABR1N

Dual-slot PCIe 3.0 x16 accelerator with 3584 CUDA cores, 16 GB HBM2 memory and 250 W passive cooling for FP64, FP32 and FP16 compute workloads

  • InterfacePCI Express 3.0 x16
  • Chipset ManufacturerNVIDIA
  • CUDA Cores3584
  • Memory Size16GB
  • CoolerFanless
  • Form FactorPlug-in CardFull-height
  • 16 GB HBM2 CoWoS memory feeds 3584 CUDA cores
  • Passive heat sink needs only chassis airflow for 250 W TDP
  • PCIe 3.0 x16 interface slots into standard servers
  • 7 TFLOPS FP64 plus 9.3 TFLOPS FP32 for mixed-precision workloads
  • GPU Boost clocks dynamically within the 250 W power cap

16 GB HBM2 accelerator for compute workloads

The NVIDIA Tesla P100 is a PCI Express 3.0 x16 add-in card built around a Pascal GP100 GPU with 3584 CUDA cores. It delivers 16 GB of HBM2 memory on a 4096-bit interface and targets double-precision, single-precision and half-precision tasks in datacentre and workstation environments.

Passive dual-slot design requires chassis airflow

The full-height card occupies a 10.5 inch dual-slot footprint and uses a fanless heatsink that depends on system-level air movement to stay within its 250 W thermal limit. No onboard fan means silent operation, but the host enclosure must supply sufficient directed airflow across the cooler.

GPU Boost and unified memory simplify deployment

NVIDIA GPU Boost dynamically adjusts clocks to maximise throughput inside the power envelope while CUDA and DirectCompute APIs provide broad software compatibility. Unified virtual memory with page migration engine lets applications treat GPU and system memory as a single address space, reducing data-copy overhead.

Pontos fortes

  • 16 GB HBM2 CoWoS memory feeds 3584 CUDA cores
  • Passive heat sink needs only chassis airflow for 250 W TDP
  • PCIe 3.0 x16 interface slots into standard servers
  • 7 TFLOPS FP64 plus 9.3 TFLOPS FP32 for mixed-precision workloads
  • GPU Boost clocks dynamically within the 250 W power cap

Especificações

BrandNVIDIA
ModelTesla P100
InterfacePCI Express 3.0 x16
Chipset ManufacturerNVIDIA
CUDA Cores3584
Memory Size16GB
CoolerFanless
Form FactorPlug-in CardFull-height

Perguntas sobre este artigo

Does this card fit in a standard full-height PCIe 3.0 x16 slot?

Yes, it is a full-height, dual-slot plug-in card measuring 10.5 inches and uses a PCI Express 3.0 x16 interface.

What power supply capacity and cooling are required?

The board draws up to 250 W and uses a passive heat sink, so the chassis must deliver strong, directed airflow across the card.

Which compute APIs and precision modes are supported?

It supports CUDA and DirectCompute with FP64, FP32 and FP16 precision, delivering 4.7 TFLOPS double-precision and 9.3 TFLOPS single-precision peak performance.

Também nesta secção

Os clientes também compararam

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