GPU Comparison

A

AMD

RADEON

AMD Radeon Instinct MI250X

CORE STATE Aldebaran

VRAM 128 GB

CLOCK SPEED 1700 MHz

TDP 500 W

BUS WIDTH 8192 bit

ARCHITECTURE CDNA 2.0

nm

PROCESS 6 nm

LAUNCH DATE 2021

VS

NVIDIA

GEFORCE

GeForce RTX 3070 Ti

CORE STATE GA104

VRAM 8 GB

CLOCK SPEED 1770 MHz

TDP 290 W

BUS WIDTH 256 bit

ARCHITECTURE Ampere

nm

PROCESS 8 nm

LAUNCH DATE 2021

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12

N/A

3,478

geekbench_opencl

N/A

137,518

geekbench_vulkan

N/A

130,778

passmark_directx_10

N/A

155

passmark_directx_11

N/A

192

passmark_directx_12

N/A

91

passmark_directx_9

N/A

261

passmark_g2d

N/A

1,055

passmark_g3d

N/A

23,356

passmark_gpu_compute

N/A

11,601

DETAILED SPECIFICATIONS

SPECIFICATION

Instinct MI250X

RTX 3070 Ti

Core Specs

Shading Units

14,080

6,144 -56.4%

Shaders

14,080

6,144 -56.4%

TMUs

880

192 -78.2%

ROPs

0

96 +∞%

Compute Units

220

—

SM Count

—

48

Clocks

Base Clock

1000 MHz

1575 MHz

Boost Clock

1700 MHz

1770 MHz

Memory Clock

1600 MHz 3.2 Gbps effective

1188 MHz 19 Gbps effective

Memory

Memory Size

128 GB

8 GB

VRAM (MB)

131,072

8,192 -93.8%

Memory Type

HBM2e

GDDR6X

Memory Bus

8192 bit

256 bit

Bandwidth

3.28 TB/s

608.3 GB/s

Cache

L1 Cache

16 KB (per CU)

128 KB (per SM)

L2 Cache

16 MB

4 MB

Performance

Pixel Rate

0 MPixel/s

169.9 GPixel/s

Texture Rate

1,496.0 GTexel/s

339.8 GTexel/s

FP32 (TFLOPS)

47.87 TFLOPS

21.75 TFLOPS

FP64 (TFLOPS)

47.87 TFLOPS (1:1)

339.8 GFLOPS (1:64)

FP16 (TFLOPS)

383.0 TFLOPS (8:1)

21.75 TFLOPS (1:1)

AI/RT

RT Cores

—

48

Tensor Cores

—

192

Power

TDP

500 W

290 W

TDP (W)

500

290 -42.0%

Suggested PSU

900 W

600 W

Power Connectors

2x 8-pin

1x 12-pin

Architecture

Architecture

CDNA 2.0

Ampere

GPU Name

Aldebaran

GA104

Generation

Radeon Instinct (MIx)

GeForce 30

Process Size

6 nm

8 nm

Transistors

58,200 million

17,400 million

Die Size

724 mm²

392 mm²

Foundry

TSMC

Samsung

Density

80.4M / mm²

44.4M / mm²

AMD MCM

MCM

2

—

API Support

DirectX

—

12 Ultimate (12_2)

OpenGL

—

4.6

Vulkan

—

1.4

OpenCL

3.0

3.0

CUDA

—

8.6

Shader Model

—

6.8

Physical

Slot Width

Dual-slot

Dual-slot

Length

267 mm 10.5 inches

267 mm 10.5 inches

Height

111 mm 4.4 inches

112 mm 4.4 inches

Outputs

No outputs

1x HDMI 2.13x DisplayPort 1.4a

Bus Interface

PCIe 4.0 x16

PCIe 4.0 x16

Other

Launch Price

—

599 USD

Production

—

End-of-life

Predecessor

FirePro Data Center

GeForce 20

Successor

—

GeForce 40

View Radeon Instinct MI250X Details View GeForce RTX 3070 Ti Details