Home GPU Comparison Intel Arctic Sound-M vs NVIDIA L40 CNX

Intel Arctic Sound-M vs NVIDIA L40 CNX

We compared two Professional market GPUs: 16GB VRAM Arctic Sound M and 24GB VRAM L40 CNX to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.

Main Differences

Intel Arctic Sound-M 's Advantages
Larger VRAM bandwidth (1229GB/s vs 864.0GB/s)
NVIDIA L40 CNX 's Advantages
Released 9 months late
Boost Clock2475MHz
More VRAM (24GB vs 16GB)
9984 additional rendering cores
Lower TDP (300W vs 500W)

Score

Benchmark

FP32 (float)
Arctic Sound M
14.75 TFLOPS
L40 CNX +509%
89.97 TFLOPS
VS

Graphics Card

Jan 2022
Release Date
Oct 2022
Data Center GPU
Generation
Tesla Ada
Professional
Type
Professional
PCIe 4.0 x16
Bus Interface
PCIe 4.0 x16

Clock Speeds

-
Base Clock
1005 MHz
-
Boost Clock
2475 MHz
1200 MHz
Memory Clock
2250 MHz

Memory

16GB
Memory Size
24GB
HBM2e
Memory Type
GDDR6
4096bit
Memory Bus
384bit
1229GB/s
Bandwidth
864.0GB/s

Render Config

-
SM Count
142
-
Compute Units
-
8192
Shading Units
18176
256
TMUs
568
128
ROPs
192
-
Tensor Cores
568
-
RT Cores
142
-
L1 Cache
128 KB (per SM)
8 MB
L2 Cache
48 MB

Theoretical Performance

115.2 GPixel/s
Pixel Rate
475.2 GPixel/s
230.4 GTexel/s
Texture Rate
1406 GTexel/s
29.49 TFLOPS
FP16 (half)
89.97 TFLOPS
14.75 TFLOPS
FP32 (float)
89.97 TFLOPS
3.686 TFLOPS
FP64 (double)
1406 GFLOPS

Graphics Processor

Arctic Sound
GPU Name
AD102
-
GPU Variant
AD102-???-A1
Generation 12.5
Architecture
Ada Lovelace
Intel
Foundry
TSMC
10 nm
Process Size
4 nm
8 billion
Transistors
76.3 billion
190 mm²
Die Size
608 mm²

Board Design

500W
TDP
300W
900 W
Suggested PSU
700 W
No outputs
Outputs
1x HDMI 2.1 3x DisplayPort 1.4a
8-pin EPS
Power Connectors
1x 16-pin

Graphics Features

12 (12_1)
DirectX
12 Ultimate (12_2)
4.6
OpenGL
4.6
3.0
OpenCL
3.0
N/A
Vulkan
1.3
-
CUDA
8.9
6.6
Shader Model
6.6

Related GPU Comparisons

Related News

© 2024 - TopCPU.net   Contact Us Privacy Policy