AMD Instinct MI100 vs NVIDIA RTX PRO 4000 Blackwell SFF Full Specs

Summary

7,680 Shaders 1.5GHz	8,960 Shaders 1.34GHz
32GB HBM21.23TB/s	24GB GDDR7432GB/s
Peak·FP32·FP16 23.07 TFLOPS	Peak·FP32·FP16 23.96 TFLOPS
Form Factor PCIe Card	Form Factor PCIe Card
TDP 300W	TDP 70W
Power Connectors 2x 8-Pin	Power Connectors -

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 Tensor FP32 FP64 BF16 Tensor BF16 BF16 Tensor Sparse Tensor TF32 Tensor INT8 Tensor INT4 Memory Bandwidth TDP

Instinct MI100184.6 TFLOPSFP16-16 Tensor

x1

RTX PRO 4000 Blackwell SFF766.7 TFLOPSFP4 Tensor Sparse

x4.15

Show Top Score

Theoretical Performance

Clock Speed1502MHz UC·Base·Boost·OC	Clock Speed1337MHz UC·Base·Boost·OC
Peak OPS 184.6 TFLOPSFP16-16 Tensor	Peak OPS 766.7 TFLOPSFP4 Tensor Sparse
-	Tensor FP4 383.3 TFLOPS FP4 Tensor Sparse 766.7 TFLOPS
-	Tensor FP8-16 191.7 TFLOPS FP8-16 Tensor Sparse 383.3 TFLOPS Tensor FP8-32 95.84 TFLOPS FP8-32 Tensor Sparse 191.7 TFLOPS
Tensor FP16-16 184.6 TFLOPS FP16-16 Tensor Sparse - Tensor FP16-32 184.6 TFLOPS FP16-32 Tensor Sparse -	Tensor FP16-16 95.84 TFLOPS FP16-16 Tensor Sparse 191.7 TFLOPS Tensor FP16-32 47.92 TFLOPS FP16-32 Tensor Sparse 95.84 TFLOPS
BF16 - Tensor BF16 92.28 TFLOPS BF16 Tensor Sparse -	BF16 23.96 TFLOPS Tensor BF16 47.92 TFLOPS BF16 Tensor Sparse 95.84 TFLOPS
Tensor TF32 -	Tensor TF32 23.96 TFLOPS
FP32 23.07 TFLOPS Tensor FP32 46.14 TFLOPS	FP32 23.96 TFLOPS Tensor FP32 -
FP64 11.54 TFLOPS	FP64 374.4 GFLOPS
Tensor INT4 92.28 TOPS	Tensor INT4 -
Tensor INT8 92.28 TOPS	Tensor INT8 191.7 TOPS
Ray -	Ray 72.66 TOPS
Pixel Rate -	Pixel Rate 128.4 GPixel/s
Texture Rate 721 GTexel/s	Texture Rate 374.4 GTexel/s

Core Config

Shaders 7,680 Shaders	Shaders 8,960 Shaders
TMUs 480 TMUs	TMUs 280 TMUs
ROPs -	ROPs 96 ROPs
Tensor Cores 480 T-Cores	Tensor Cores 280 T-Cores
RT Cores -	RT Cores 70 RT-Cores
CUs 120 CUs	SMs 70 SMs

Clocks

Base Clock 1GHz	Base Clock 790MHz
Boost Clock 1.5GHz	Boost Clock 1.34GHz

Cache

L2 Cache 8.2MB shared

L2 Cache 49.1MB shared

Memory

32GB HBM2	24GB GDDR7
Memory Bus 4096-bit	Memory Bus 192-bit
Memory Speed 2.4GT/s	Memory Speed 18GT/s
Memory Bandwidth 1.23TB/s	Memory Bandwidth 432GB/s
ECC Yes	ECC Yes

Power & Thermals

TDP 300W

TDP 70W

Video Output

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

Ports

-	4x DisplayPort 2.1

Encoder

Encoder Model VCN 2.5

Encoder Model 2x NVENC 9

Decoder

Decoder Model VCN 2.5

Decoder Model 2x NVDEC 6

Card

Form Factor PCIe Card	Form Factor PCIe Card
PCIe 2-Slots	PCIe 1-Slots
Height 111 mm (4.37") Width 267 mm (10.51") Depth 37 mm (1.46")	Height 69 mm (2.72") Width 167 mm (6.58") Depth 20 mm (0.79")
Cooling Passive -	Cooling Blower 1x Fan
Power Connectors 2x 8-Pin	Power Connectors -

Chip

Manufacturer AMD	Manufacturer NVIDIA
Chip Designer AMD	Chip Designer NVIDIA
Architecture CDNA 1	Architecture Blackwell
Family Instinct	Family RTX PRO Blackwell
Branding	Branding
Codename Arcturus	Codename NV193
Chip Variant Arcturus XL	Chip Variant GB203-200-A1
Market Segment Server	Market Segment Workstation
Release Date Nov 16, 2020	Release Date Aug 11, 2025

Fabrication

Foundry TSMC	Foundry TSMC
Fabrication Node N7	Fabrication Node 4NP
Die Size 750mm²	Die Size 378mm²
Transistor Count 25.6 Billion	Transistor Count 45.6 Billion
Transistor Density 34.13 MTr/mm²	Transistor Density 120.6 MTr/mm²

Images

No images available

No images available

Competitors

AMDInstinct MI100 vs GPU AMDInstinct MI300P AMDInstinct MI100 vs GPU NVIDIAGeForce RTX 5070 AMDInstinct MI100 vs GPU NVIDIAGeForce RTX 5070 Ti Show 12 more

NVIDIARTX PRO 4000 Blackwell SFF vs GPU NVIDIARTX PRO 4000 Blackwell NVIDIARTX PRO 4000 Blackwell SFF vs GPU NVIDIARTX 4500 Ada NVIDIARTX PRO 4000 Blackwell SFF vs GPU NVIDIARTX PRO 2000 Blackwell Show 12 more