GPUs - HMC

GPUs

NVIDIA

GeForce RTX 5070 Laptop 50W

Shaders

4,608

Boost

1.43GHz

Memory

8GB

TDP

50W

FP32

13.13 TFLOPS—

Peak

665 TFLOPSFP4 Tensor Sparse—

—

NVIDIA

GeForce RTX 5090

Shaders

21,760

Boost

2.41GHz

Memory

32GB

TDP

575W

FP32

104.8 TFLOPS—

Peak

3.35 PFLOPSFP4 Tensor Sparse—

$2,000

NVIDIA

GeForce RTX 5080

Shaders

10,752

Boost

2.62GHz

Memory

16GB

TDP

360W

FP32

56.28 TFLOPS—

Peak

1.8 PFLOPSFP4 Tensor Sparse—

$1,000

NVIDIA

GeForce RTX 5070 Ti

Shaders

8,960

Boost

2.45GHz

Memory

16GB

TDP

300W

FP32

43.94 TFLOPS—

Peak

1.41 PFLOPSFP4 Tensor Sparse—

$750

NVIDIA

GeForce RTX 5070

Shaders

6,144

Boost

2.51GHz

Memory

12GB

TDP

250W

FP32

30.87 TFLOPS—

Peak

987.8 TFLOPSFP4 Tensor Sparse—

$550

Intel

UHD 770 1660MHz

Shaders

256

Boost

1.66GHz

Memory

Shared

TDP

Shared

FP32

849.9 GFLOPS—

Peak

1.7 TFLOPSFP16—

—

ARM

Mali-G720 MP7 1300MHz

Shaders

896

Boost

1.3GHz

Memory

Shared

TDP

Shared

FP32

2.33 TFLOPS—

Peak

4.66 TFLOPSFP16—

—

Intel

Arc B580

Shaders

2,560

Boost

2.67GHz

Memory

12GB

TDP

190W

FP32

13.67 TFLOPS—

Peak

437.5 TOPSINT4 Tensor—

$250

Intel

Arc B570

Shaders

2,304

Boost

2.5GHz

Memory

10GB

TDP

150W

FP32

11.52 TFLOPS—

Peak

368.6 TOPSINT4 Tensor—

$220

ARM

Mali-G615 MP6 1400MHz

Shaders

768

Boost

1.4GHz

Memory

Shared

TDP

Shared

FP32

2.15 TFLOPS—

Peak

4.3 TFLOPSFP16—

—

AMD

Radeon PRO W7800 48GB

Shaders

4,480

Boost

2.52GHz

Memory

48GB

TDP

260W

FP32

45.19 TFLOPS—

Peak

180.8 TOPSINT4 Tensor—

—

Apple

40-Core M4 Max

Shaders

5,120

Boost

1.58GHz

Memory

Shared

TDP

Shared

FP32

16.16 TFLOPS—

Peak

16.16 TFLOPSFP16—

—

Apple

32-Core M4 Max

Shaders

4,096

Boost

1.58GHz

Memory

Shared

TDP

Shared

FP32

12.93 TFLOPS—

Peak

12.93 TFLOPSFP16—

—

Apple

20-Core M4 Pro

Shaders

2,560

Boost

1.58GHz

Memory

Shared

TDP

Shared

FP32

8.08 TFLOPS—

Peak

8.08 TFLOPSFP16—

—

Apple

16-Core M4 Pro

Shaders

2,048

Boost

1.58GHz

Memory

Shared

TDP

Shared

FP32

6.46 TFLOPS—

Peak

6.46 TFLOPSFP16—

—

Apple

10-Core M4

Shaders

1,280

Boost

1.58GHz

Memory

Shared

TDP

Shared

FP32

4.04 TFLOPS—

Peak

4.04 TFLOPSFP16—

—

Apple

8-Core M4

Shaders

1,024

Boost

1.47GHz

Memory

Shared

TDP

Shared

FP32

3.01 TFLOPS—

Peak

3.01 TFLOPSFP16—

—

Samsung

Xclipse 540

Shaders

256

Boost

1.3GHz

Memory

Shared

TDP

Shared

FP32

1.33 TFLOPS—

Peak

5.33 TOPSINT4 Tensor—

—

Apple

5-Core A17 Pro

Shaders

640

Boost

1.34GHz

Memory

Shared

TDP

Shared

FP32

1.71 TFLOPS—

Peak

1.71 TFLOPSFP16—

—

AMD

Instinct MI325X

Shaders

19,456

Boost

2.1GHz

Memory

256GB

TDP

1000W

FP32

163.4 TFLOPS—

Peak

5.23 POPSINT8 Tensor Sparse—

—

AMD

Radeon PRO V710

Shaders

3,456

Boost

2GHz

Memory

28GB

TDP

158W

FP32

27.65 TFLOPS—

Peak

110.6 TOPSINT4 Tensor—

—

Intel

Graphics 96EU 1550MHz

Shaders

768

Boost

1.55GHz

Memory

Shared

TDP

Shared

FP32

2.38 TFLOPS—

Peak

4.76 TFLOPSFP16—

—

Intel

Graphics 64EU 1550MHz

Shaders

512

Boost

1.55GHz

Memory

Shared

TDP

Shared

FP32

1.59 TFLOPS—

Peak

3.17 TFLOPSFP16—

—

Intel

Graphics 48EU 1400MHz

Shaders

384

Boost

1.4GHz

Memory

Shared

TDP

Shared

FP32

1.07 TFLOPS—

Peak

2.15 TFLOPSFP16—

—

ARM

Immortalis-G925 MP12 1612MHz

Shaders

1,536

Boost

1.61GHz

Memory

Shared

TDP

Shared

FP32

4.95 TFLOPS—

Peak

9.9 TFLOPSFP16—

—

Page 8