Home GPU Comparison NVIDIA Tesla K8 vs H200 SXM

NVIDIA Tesla K8 vs H200 SXM

AI GPU We compared a Professional market GPU: 8GB VRAM Tesla K8 and a GPU: 141GB VRAM H200 SXM to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.

Main Differences

NVIDIA Tesla K8 's Advantages
Lower TDP (100W vs 700W)
H200 SXM 's Advantages
Released 9 years and 2 months late
Boost Clock has increased by 144% (1980MHz vs 811MHz)
More VRAM (141GB vs 8GB)
Larger VRAM bandwidth (4800GB/s vs 160.0GB/s)
15360 additional rendering cores

Score

Benchmark

FP32 (float)
Tesla K8
2.491 TFLOPS
H200 SXM +2589%
67 TFLOPS
VS

Graphics Card

Sep 2014
Release Date
Nov 2023
Tesla
Generation
Hopper
Professional
Type
AI GPU
PCIe 2.0 x16
Bus Interface
PCIe 5.0 x16

Clock Speeds

693 MHz
Base Clock
1590 MHz
811 MHz
Boost Clock
1980 MHz
1250 MHz
Memory Clock
1313 MHz

Memory

8GB
Memory Size
141GB
GDDR5
Memory Type
HBM3e
256bit
Memory Bus
5120bit
160.0GB/s
Bandwidth
4800GB/s

Render Config

-
SM Count
132
-
Compute Units
-
1536
Shading Units
16896
128
TMUs
528
32
ROPs
24
-
Tensor Cores
528
-
RT Cores
-
16 KB (per SMX)
L1 Cache
256 KB (per SM)
512 KB
L2 Cache
50 MB

Theoretical Performance

25.95 GPixel/s
Pixel Rate
47.52 GPixel/s
103.8 GTexel/s
Texture Rate
1045 GTexel/s
-
FP16 (half)
1979 TFLOPS
2.491 TFLOPS
FP32 (float)
67 TFLOPS
103.8 GFLOPS
FP64 (double)
34 TFLOPS

Graphics Processor

GK104
GPU Name
GH200
-
GPU Variant
-
Kepler
Architecture
Hopper
TSMC
Foundry
TSMC
28 nm
Process Size
4 nm
3.54 billion
Transistors
80 billion
294 mm²
Die Size
814 mm²

Board Design

100W
TDP
700W
300 W
Suggested PSU
1100 W
No outputs
Outputs
No outputs
1x 6-pin
Power Connectors
8-pin EPS

Graphics Features

12 (11_0)
DirectX
N/A
4.6
OpenGL
N/A
3.0
OpenCL
3.0
1.1
Vulkan
N/A
3.0
CUDA
9.0
5.1
Shader Model
N/A

Related GPU Comparisons

© 2024 - TopCPU.net   Contact Us Privacy Policy