GPU

NVIDIA Tesla K40d

Edit@2 months ago

Intergrated Memory(VRAM)
Capacity

12 GB

(GDDR5 384-bit)

Bandwidth

288 GB/s

41 Token/s

Vector Compute
FP64
1.68 T
FP32
5.05 T
FP16
X
BF16
X
INT32
NI
INT8
X

NVIDIA Tesla K40d General-Purpose Float-Point performance (Vector Performance / Scalar Performance)

FP64: 1.68 TFLOPS

FP32: 5.05 TFLOPS

Matirx Compute
FP64
X
FP32
X
FP16
X
FP8
X
TF32
X
BF16
X
INT16
X
INT8
X
INT4
X

NVIDIA Tesla K40d AI performance (Tensor Performance / Matrix Performance)

Hardware Specs
NVIDIA Tesla K40d is a 28nm chip, has 7080 million transistors, launched by NVIDIA at 2013. It has 12 GB built-in(On-Board/On-Chip) memory with bandwidth up to 288 GB/s. It has 2880 general-purpose ALUs(CUDA cores/Shader cores).
Process Node
28 nm
Launch Year
2013
Vector(CUDA) Cores
2880
Matrix(Tensor) Cores
Core Frequency
745 ~ 876 MHz
Cache
1.5MB