Scientific Computing Card

Tesla A40

Ampere GPU 10,752 NVIDIA? CUDA? Cores 336 NVIDIA? Tensor Cores 84 NVIDIA? RT Cores 48GB GDDR6 Memory with ECC

GPU Architecture	Ampere
CUDA Parallel Processing cores	10,752
NVIDIA Tensor Cores	336
NVIDIA RT Cores	84
Frame Buffer Memory	48 GB GDDR6 with ECC
Memory Interface	384-bit
Memory Bandwidth	768 GB/s
Power Consumption	300 W
Graphics Bus	PCI Express 4.0 x16
Display Connectors	DP 1.4 (3) **
Form Factor	4.4” H x 10.5” L Dual Slot
Product Weight	1.179 kg
Thermal Solution	Passive
vGPU Software Support	NVIDIA^® GRID^® , NVIDIA Quadro^® Virtual Data Center Workstation, NVIDIA Virtual Compute Server*
vGPU Profiles Supported	1 GB, 2 GB, 3 GB, 4 GB, 6 GB, 8 GB, 12 GB, 16 GB, 24 GB, 48 GB
Power Connector	1x 8-pin
Frame lock	Compatible (with Quadro Sync II)
NVLink Interconnect	112.5 GB/s (bidirectional)
NVLink	2-way low profile (2-slot)

* A40 does not support MIG.

* Display ports are not active by default for A40. NVIDIA vGPU software is only supported when display ports are not active.