GPU

NVIDIA Tesla K80

Edit@1 months ago

Intergrated Memory(VRAM)
Capacity

12 GB

(GDDR5 384-bit)

Bandwidth

240 GB/s

34 Token/s

Vector Compute
FP64
1.37 T
FP32
4.11 T
FP16
X
BF16
X
INT32
NI
INT8
X

NVIDIA Tesla K80 General-Purpose Float-Point performance (Vector Performance / Scalar Performance)

FP64: 1.37 TFLOPS

FP32: 4.11 TFLOPS

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

NVIDIA Tesla K80 AI performance (Tensor Performance / Matrix Performance)

Hardware Specs
NVIDIA Tesla K80 is a 28nm chip, has 7100 million transistors, launched by NVIDIA at 2014. It has 12 GB built-in(On-Board/On-Chip) memory with bandwidth up to 240 GB/s. It has 2496 general-purpose ALUs(CUDA cores/Shader cores).
Process Node
28 nm
Launch Year
2014
Vector(CUDA) Cores
2496
Matrix(Tensor) Cores
Core Frequency
562 ~ 824 MHz
Cache
1.5MB

Comment without registration

Share your experience with NVIDIA Tesla K80 / Found an Error? Help Us Improve!