NVIDIA

NVIDIA GB200 Grace Blackwell Superchip

Riferimenti dei componenti

Aggiungi al confronto

Configurazione di sistema

Specifiche dettagliate

Specifiche 1 · 7 campi

Modello grafico

GB200^*

Ambito e condizioni

native_context

Physical scope

1 CPU + 2 GPU superchip^*

Denominazione grafica completa

NVIDIA GB200^*

Ambito e condizioni

native_context

Physical scope

1 CPU + 2 GPU superchip^*

Produttore

NVIDIA^*

Ambito e condizioni

native_context

Physical scope

1 CPU + 2 GPU superchip^*

Formato dichiarato

Superchip module (Grace CPU + 2 Blackwell GPUs)^*

Ambito e condizioni

native_context

Physical scope

1 CPU + 2 GPU superchip^*

Capacità memoria nello specifico ambito

372^* GB

Ambito e condizioni

native_context

Capacity unit base

Not independently established^*

Physical scope

1 CPU + 2 GPU superchip^*

Tecnologia memoria

HBM3E^*

Ambito e condizioni

native_context

Physical scope

1 CPU + 2 GPU superchip^*

Potenza dichiarata del package

2.700^* W

Ambito e condizioni

native_context

Physical scope

1 CPU + 2 GPU superchip^*

Specifiche 2 · 13 campi

INT8 Tensor Core 2

20 POPS^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

TF32 Tensor Core 2

5 PFLOPS^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

FP16/BF16 Tensor Core 2

10 PFLOPS^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

NVFP4 Tensor Core 2

40 | 20 PFLOPS^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

FP8/FP6 Tensor Core 2

20 PFLOPS^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

Configuration

1 Grace CPU | 2 Blackwell GPU^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

FP32

160 TFLOPS^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

CPU Memory | Bandwidth

values

• unit

  GB^*

  value

  amount

  480^*

  memory_type

  LPDDR5X^*

  qualifier

  up_to^*

  type

  memory_quantity^*

• unit

  GB/s^*

  value

  amount

  512^*

  qualifier

  up_to^*

  type

  quantity^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

GPU Memory | Bandwidth

values

• unit

  GB^*

  value

  amount

  372^*

  memory_type

  HBM3E^*

  type

  memory_quantity^*

• unit

  TB/s^*

  value

  amount

  16^*

  type

  quantity^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

Nominal GPU Memory

192^* GB · HBM3e^*

Ambito e condizioni

native_context

Memory scope

Nominal per-GPU memory; not the total CPU + GPU module memory or runtime-available capacity^*

memory role

nominal_release_inventory^*

not dedicated VRAM for shared memory systems

Sì^*

not runtime available or HGX usable capacity

Sì^*

NVLink Bandwidth

3,6^* TB/s

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

CPU Core Count

72 Arm Neoverse V2 cores^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

FP64 / FP64 Tensor Core

80 TFLOPS^*

Ambito e condizioni

native_context

performance is theoretical not benchmark

Sì^*

resource scope

superchip_1_CPU_2_GPU^*

sparse and dense figures not interchangeable

Sì^*

Trova un altro elemento da confrontare

Cerca