GPUs - HMC

GPUs

Qualcomm

Adreno X1 2.1T

Shaders

1,024

Boost

1.03GHz

Memory

Shared

TDP

15W

FP32

2.12 TFLOPS—

Peak

4.23 TFLOPSFP16—

—

Qualcomm

Adreno X1 1.7T

Shaders

1,024

Boost

833MHz

Memory

Shared

TDP

15W

FP32

1.71 TFLOPS—

Peak

3.41 TFLOPSFP16—

—

Qualcomm

Adreno X1 3.8T

Shaders

1,536

Boost

1.25GHz

Memory

Shared

TDP

15W

FP32

3.84 TFLOPS—

Peak

7.68 TFLOPSFP16—

—

Qualcomm

Adreno X1 4.6T

Shaders

1,536

Boost

1.5GHz

Memory

Shared

TDP

15W

FP32

4.61 TFLOPS—

Peak

9.22 TFLOPSFP16—

—

Qualcomm

Adreno 750

Shaders

1,536

Boost

900MHz

Memory

Shared

TDP

FP32

2.77 TFLOPS—

Peak

5.53 TFLOPSFP16—

—

AMD

Radeon RX 7800M

Shaders

3,840

Boost

2.33GHz

Memory

12GB

TDP

180W

FP32

35.87 TFLOPS—

Peak

143.5 TOPSINT4 Tensor—

—

AMD

Radeon RX 7900M

Shaders

4,608

Boost

2.09GHz

Memory

16GB

TDP

180W

FP32

38.52 TFLOPS—

Peak

154.1 TOPSINT4 Tensor—

—

AMD

Radeon RX 6750 GRE

Shaders

2,560

Boost

2.58GHz

Memory

12GB

TDP

250W

FP32

13.21 TFLOPS—

Peak

26.43 TFLOPSFP16—

$334

Intel

Arc A580

Shaders

3,072

Boost

2GHz

Memory

6GB

TDP

185W

FP32

12.29 TFLOPS—

Peak

393.2 TOPSINT4 Tensor—

$180

ARM

Mali-G715 MP7 940MHz

Shaders

448

Boost

940MHz

Memory

Shared

TDP

Shared

FP32

842.2 GFLOPS—

Peak

1.68 TFLOPSFP16—

—

ARM

Mali-G715 MP7 890MHz

Shaders

448

Boost

890MHz

Memory

Shared

TDP

Shared

FP32

797.4 GFLOPS—

Peak

1.59 TFLOPSFP16—

—

ARM

Immortalis-G715 MP10 890MHz

Shaders

640

Boost

890MHz

Memory

Shared

TDP

Shared

FP32

1.14 TFLOPS—

Peak

2.28 TFLOPSFP16—

—

AMD

Radeon 660M 4CU 1800MHz

Shaders

256

Boost

1.8GHz

Memory

Shared

TDP

Shared

FP32

921.6 GFLOPS—

Peak

1.84 TFLOPSFP16—

—

Apple

S9

Shaders

128

Boost

440MHz

Memory

Shared

TDP

Shared

FP32

112.6 GFLOPS—

Peak

112.6 GFLOPSFP16—

—

Apple

6-Core A17 Pro

Shaders

768

Boost

1.34GHz

Memory

Shared

TDP

Shared

FP32

2.06 TFLOPS—

Peak

2.06 TFLOPSFP16—

—

ARM

Mali-G57 MP2 1100MHz

Shaders

Boost

1.1GHz

Memory

Shared

TDP

Shared

FP32

140.8 GFLOPS—

Peak

281.6 GFLOPSFP16—

—

Qualcomm

Adreno 710

Shaders

256

Boost

940MHz

Memory

Shared

TDP

FP32

481.3 GFLOPS—

Peak

962.6 GFLOPSFP16—

—

AMD

Radeon RX 7800 XT

Shaders

3,840

Boost

2.43GHz

Memory

16GB

TDP

263W

FP32

37.32 TFLOPS—

Peak

149.3 TOPSINT4 Tensor—

$500

AMD

Radeon RX 7700 XT

Shaders

3,456

Boost

2.54GHz

Memory

12GB

TDP

245W

FP32

35.17 TFLOPS—

Peak

140.7 TOPSINT4 Tensor—

$450

AMD

Radeon RX 7700

Shaders

3,072

Boost

2.4GHz

Memory

12GB

TDP

200W

FP32

29.49 TFLOPS—

Peak

118 TOPSINT4 Tensor—

$400

NVIDIA

RTX 5000 Ada

Shaders

12,800

Boost

2.55GHz

Memory

32GB

TDP

250W

FP32

65.28 TFLOPS—

Peak

2.09 POPSINT4 Tensor Sparse—

—

NVIDIA

RTX 4500 Ada

Shaders

7,680

Boost

2.58GHz

Memory

24GB

TDP

130W

FP32

39.63 TFLOPS—

Peak

1.27 POPSINT4 Tensor Sparse—

—

NVIDIA

RTX 4000 Ada

Shaders

6,144

Boost

2.17GHz

Memory

20GB

TDP

130W

FP32

26.73 TFLOPS—

Peak

855.2 TOPSINT4 Tensor Sparse—

—

AMD

Radeon PRO W7600

Shaders

2,048

Boost

2.44GHz

Memory

8GB

TDP

130W

FP32

19.99 TFLOPS—

Peak

79.95 TOPSINT4 Tensor—

$600

AMD

Radeon PRO W7500

Shaders

1,792

Boost

1.7GHz

Memory

8GB

TDP

70W

FP32

12.19 TFLOPS—

Peak

48.74 TOPSINT4 Tensor—

$430

Page 10