NVIDIA PG506-217 vs SPARKLE Arc A310 OmniView

Comparison of NVIDIA PG506-217 with 24 GB HBM2 and 3,584 cores vs SPARKLE Arc A310 OmniView with 4 GB GDDR6 and 768 cores.

Loading...

Performance Rating

NVIDIA PG506-217 outperforms SPARKLE Arc A310 OmniView by 462.73% in the overall GPU ARK performance rating

A100 A100
H200 H200
MI325X MI325X

NVIDIA PG506-217

12.4

NVIDIA PG506-217

12.4
RX 7900 XTX RX 7900 XTX
MI250 MI250
Instinct MI300X Instinct MI300X

SPARKLE Arc A310 OmniView

2.2

SPARKLE Arc A310 OmniView

2.2

Contents:

Memory ML Performance Compute Power Architecture & Compatibility ML Software Support Clocks & Performance Power Consumption Rendering Additional

Memory

Memory Size

🔥 +500% 24 ГБ
4 ГБ

Memory Type

HBM2 GDDR6

Memory Bandwidth

933.1 GB/s 124.0 GB/s

Memory Bus Width

3,072 бит 64 бит

ML Performance

FP16 (Half Precision)

🔥 +68% 10.32 TFLOPS
6.144 TFLOPS

BF16 (Brain Float)

No No

TF32 (TensorFloat)

No No

Compute Power

FP32 (Single Precision)

🔥 +236% 10.32 TFLOPS
3.072 TFLOPS

FP64 (Double Precision)

🔥 +572% 5.161 TFLOPS
0.768 TFLOPS

CUDA Cores

🔥 +367% 3,584
768

RT Cores

No
🔥 6

Architecture & Compatibility

GPU Architecture

Ampere Generation 12.7

SM (Streaming Multiprocessor)

🔥 56
No

PCIe Version

PCIe 4.0 x16 PCIe 4.0 x8

ML Software Support

CUDA Version

8.0 No

Clocks & Performance

Base Clock

930
🔥 +115% 2,000

Boost Clock

1,440
🔥 +39% 2,000

Memory Clock

1,215
🔥 +59% 1,937

Power Consumption

TDP/TGP

165 W
🔥 -55% 75 W

Recommended PSU

450 W
🔥 -44% 250 W

Power Connector

8-pin EPS None

Rendering

Texture Units (TMU)

🔥 +600% 224
32

ROP

No
🔥 6

L2 Cache

24 MB
🔥 4 MB

Additional

Slots

Dual-slot Single-slot

Release Date

April 12, 2021 Nov. 18, 2024

Display Outputs

No outputs
4x mini-DisplayPort 2.0
4x HDMI 2.1

Renting is cheaper than buying