AMD Radeon PRO W7800 vs NVIDIA RTX PRO 4500 Blackwell Full Specs

Summary

4,480 Shaders 2.52GHz	10,496 Shaders 2.62GHz
32GB GDDR6576GB/s	32GB GDDR7896GB/s
Peak·FP32·FP16 45.25 TFLOPS	Peak·FP32·FP16 54.94 TFLOPS
Form Factor PCIe Card	Form Factor PCIe Card
TDP 260W	TDP 200W
Power Connectors - 2x 8-Pin	Power Connectors 1x 16-Pin 12VHPWR -

Highlights

Peak OPS

Peak OPS FP32 FP16 Tensor FP4 FP4 Tensor Sparse Tensor FP8-16 FP8-16 Tensor Sparse Tensor FP8-32 FP8-32 Tensor Sparse Tensor FP16-16 FP16-16 Tensor Sparse Tensor FP16-32 FP16-32 Tensor Sparse FP64 BF16 Tensor BF16 BF16 Tensor Sparse Tensor TF32 Tensor INT8 Tensor INT4 Memory Bandwidth TDP

Radeon PRO W7800181 TOPSINT4 Tensor

x1

RTX PRO 4500 Blackwell1.76 PFLOPSFP4 Tensor Sparse

x9.71

Show Top Score

Theoretical Performance

Clock Speed2525MHz UC·Boost·OC	Clock Speed2617MHz UC·Boost·OC
Peak OPS 181 TOPSINT4 Tensor	Peak OPS 1.76 PFLOPSFP4 Tensor Sparse
-	Tensor FP4 879 TFLOPS FP4 Tensor Sparse 1.76 PFLOPS
-	Tensor FP8-16 439.5 TFLOPS FP8-16 Tensor Sparse 879 TFLOPS Tensor FP8-32 219.7 TFLOPS FP8-32 Tensor Sparse 439.5 TFLOPS
Tensor FP16-16 45.25 TFLOPS FP16-16 Tensor Sparse - Tensor FP16-32 45.25 TFLOPS FP16-32 Tensor Sparse -	Tensor FP16-16 219.7 TFLOPS FP16-16 Tensor Sparse 439.5 TFLOPS Tensor FP16-32 109.9 TFLOPS FP16-32 Tensor Sparse 219.7 TFLOPS
BF16 90.5 TFLOPS Tensor BF16 45.25 TFLOPS BF16 Tensor Sparse -	BF16 54.94 TFLOPS Tensor BF16 109.9 TFLOPS BF16 Tensor Sparse 219.7 TFLOPS
Tensor TF32 -	Tensor TF32 54.94 TFLOPS
FP32 45.25 TFLOPS	FP32 54.94 TFLOPS
FP64 1.41 TFLOPS	FP64 858.4 GFLOPS
Tensor INT4 181 TOPS	Tensor INT4 -
Tensor INT8 45.25 TOPS	Tensor INT8 439.5 TOPS
Ray -	Ray 166.6 TOPS
Pixel Rate 323.2 GPixel/s	Pixel Rate 293.1 GPixel/s
Texture Rate 707 GTexel/s	Texture Rate 858.4 GTexel/s

Core Config

Shaders 4,480 Shaders	Shaders 10,496 Shaders
TMUs 280 TMUs	TMUs 328 TMUs
ROPs 128 ROPs	ROPs 112 ROPs
Tensor Cores 140 T-Cores	Tensor Cores 328 T-Cores
RT Cores 70 RT-Cores	RT Cores 82 RT-Cores
CUs 70 CUs	SMs 82 SMs

Clocks

Boost Clock 2.52GHz

Boost Clock 2.62GHz

Cache

L2 Cache 6.1MB shared	L2 Cache 65.5MB shared
L3 Cache 64MB shared	L3 Cache -
L3 Bandwidth 2.33TB/s	L3 Bandwidth -

Memory

32GB GDDR6	32GB GDDR7
Memory Bus 256-bit	Memory Bus 256-bit
Memory Speed 18GT/s	Memory Speed 28GT/s
Memory Bandwidth 576GB/s	Memory Bandwidth 896GB/s
ECC No	ECC Yes

Power & Thermals

TDP 260W

TDP 200W

Video Output

Multi-Monitor 3	Multi-Monitor 4
HDCP HDCP 2.3	HDCP HDCP 2.3

Ports

3x DisplayPort 2.1	4x DisplayPort 2.1
1x Mini DisplayPort 2.1	-

Encoder

Encoder Model VCN 4.0

Encoder Model 2x NVENC 9

Decoder

Decoder Model VCN 4.0

Decoder Model 2x NVDEC 6

Card

Form Factor PCIe Card	Form Factor PCIe Card
PCIe 2-Slots	PCIe 2-Slots
Height 110 mm (4.33") Width 280 mm (11.02") Depth 40 mm (1.57")	Height 111.8 mm (4.4") Width 266.7 mm (10.5") Depth 40 mm (1.57")
Cooling Blower 1x Fan	Cooling Blower 1x Fan
Power Connectors 2x 8-Pin	Power Connectors 1x 16-Pin 12VHPWR

Chip

Manufacturer AMD	Manufacturer NVIDIA
Chip Designer AMD	Chip Designer NVIDIA
Architecture RDNA 3	Architecture Blackwell
Family Radeon PRO W	Family RTX PRO Blackwell
Branding	Branding
Codename Plum Bonito	Codename NV193
Chip Variant Navi 31 XL	Chip Variant GB203-200-A1
Market Segment Workstation	Market Segment Workstation
Release Date Apr 13, 2023	Release Date Mar 18, 2025

Fabrication

Foundry TSMC	Foundry TSMC
Other Foundries TSMCMCD	Other Foundries -
Fabrication Node N5 N6MCD	Fabrication Node 4NP -
Die Size 304mm² 225mm²MCD	Die Size 378mm² -
Transistor Count 45.4 Billion 12.3 BillionMCD	Transistor Count 45.6 Billion -
Transistor Density 149.2 MTr/mm² 54.67 MTr/mm²MCD	Transistor Density 120.6 MTr/mm² -

Images

No images available

No images available

Competitors

AMDRadeon PRO W7800 vs GPU NVIDIARTX 5000 Ada AMDRadeon PRO W7800 vs GPU NVIDIARTX 4500 Ada AMDRadeon PRO W7800 vs GPU AMDRadeon PRO W7900 Show 12 more

NVIDIARTX PRO 4500 Blackwell vs GPU NVIDIARTX PRO 4000 Blackwell NVIDIARTX PRO 4500 Blackwell vs GPU NVIDIARTX PRO 5000 Blackwell NVIDIARTX PRO 4500 Blackwell vs GPU NVIDIARTX 5000 Ada Show 12 more