NVIDIA T40 vs AMD Radeon RX 9070 XT Full Specs

Summary

4,608 Shaders 1.56GHz	4,096 Shaders 2.97GHz
24GB GDDR6624GB/s	16GB GDDR6640GB/s
Peak·FP32·FP16 14.38 TFLOPS	Peak·FP32·FP16 48.66 TFLOPS
Form Factor PCIe Card	Form Factor PCIe Card
TDP 150W	TDP 304W
Power Connectors 1x 8-Pin 1x 6-Pin	Power Connectors 2x 8-Pin -

Highlights

Peak OPS

Peak OPS FP32 FP16 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

T40460.1 TOPSINT4 Tensor

x1

Radeon RX 9070 XT1.56 POPSINT4 Tensor Sparse

x3.38

Show Top Score

Theoretical Performance

Clock Speed1560MHz UC·Base·Boost·OC	Clock Speed2970MHz UC·Base·Boost·OC
Peak OPS 460.1 TOPSINT4 Tensor	Peak OPS 1.56 POPSINT4 Tensor Sparse
-	Tensor FP8-16 194.6 TFLOPS FP8-16 Tensor Sparse 389.3 TFLOPS Tensor FP8-32 194.6 TFLOPS FP8-32 Tensor Sparse 389.3 TFLOPS
Tensor FP16-16 115 TFLOPS FP16-16 Tensor Sparse - Tensor FP16-32 57.51 TFLOPS FP16-32 Tensor Sparse -	Tensor FP16-16 97.32 TFLOPS FP16-16 Tensor Sparse 194.6 TFLOPS Tensor FP16-32 97.32 TFLOPS FP16-32 Tensor Sparse 194.6 TFLOPS
BF16 - Tensor BF16 57.51 TFLOPS BF16 Tensor Sparse -	BF16 97.32 TFLOPS Tensor BF16 97.32 TFLOPS BF16 Tensor Sparse 194.6 TFLOPS
Tensor TF32 57.51 TFLOPS	Tensor TF32 -
FP32 14.38 TFLOPS	FP32 48.66 TFLOPS
FP64 449.3 GFLOPS	FP64 1.52 TFLOPS
Tensor INT4 460.1 TOPS	Tensor INT4 778.6 TOPS
Tensor INT8 230 TOPS	Tensor INT8 389.3 TOPS
Ray 43.37 TOPS	Ray -
Pixel Rate 149.8 GPixel/s	Pixel Rate 380.2 GPixel/s
Texture Rate 449.3 GTexel/s	Texture Rate 760.3 GTexel/s

Core Config

Shaders 4,608 Shaders	Shaders 4,096 Shaders
TMUs 288 TMUs	TMUs 256 TMUs
ROPs 96 ROPs	ROPs 128 ROPs
Tensor Cores 576 T-Cores	Tensor Cores 128 T-Cores
RT Cores 72 RT-Cores	RT Cores 64 RT-Cores
SMs 72 SMs	CUs 64 CUs

Clocks

Base Clock 1.3GHz	Base Clock 2.4GHz
Boost Clock 1.56GHz	Boost Clock 2.97GHz

Cache

L2 Cache 6.1MB shared	L2 Cache 8.2MB shared
L3 Cache -	L3 Cache 64MB shared
L3 Bandwidth -	L3 Bandwidth 2.25TB/s

Memory

24GB GDDR6	16GB GDDR6
Memory Bus 384-bit	Memory Bus 256-bit
Memory Speed 13GT/s	Memory Speed 20GT/s
Memory Bandwidth 624GB/s	Memory Bandwidth 640GB/s

Power & Thermals

TDP 150W

TDP 304W

Video Output

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

Ports

-	3x DisplayPort 2.1
-	1x HDMI 2.1

Encoder

Encoder Model NVENC 7

Encoder Model VCN 4.0

Decoder

Decoder Model NVDEC 4

Decoder Model VCN 4.0

Card

Form Factor PCIe Card	Form Factor PCIe Card
PCIe 1-Slots	PCIe 2.5-Slots
Height 111 mm (4.37") Width 267 mm (10.51") Depth 20 mm (0.79")	Height 111 mm (4.37") Width 267 mm (10.51") Depth 50 mm (1.97")
Cooling Passive -	Cooling Open-Air 3x Fans
Power Connectors 1x 8-Pin, 1x 6-Pin	Power Connectors 2x 8-Pin

Chip

Manufacturer NVIDIA	Manufacturer AMD
Chip Designer NVIDIA	Chip Designer AMD
Architecture Turing	Architecture RDNA 4
Family Server	Family Radeon RX 9000
Branding	Branding
Codename NV162	Codename -
Chip Variant TU102-300A-K1-A1	Chip Variant Navi 48 XL
Market Segment Server	Market Segment Desktop
Release Date Sep 13, 2018	Release Date Jan 6, 2025

Fabrication

Foundry TSMC	Foundry TSMC
Fabrication Node 12FFN	Fabrication Node N4P
Die Size 754mm²	Die Size 357mm²
Transistor Count 18.6 Billion	Transistor Count 53.9 Billion
Transistor Density 24.67 MTr/mm²	Transistor Density 151.2 MTr/mm²

Images

No images available

No images available

Competitors

NVIDIAT40 vs GPU AMDRadeon PRO W7700 NVIDIAT40 vs GPU AMDRadeon AI PRO R9700 NVIDIAT40 vs GPU AMDRadeon RX 7800 XT Show 12 more

AMDRadeon RX 9070 XT vs GPU NVIDIAGeForce RTX 5070 Ti AMDRadeon RX 9070 XT vs GPU NVIDIAGeForce RTX 4080 Super AMDRadeon RX 9070 XT vs GPU NVIDIAGeForce RTX 4070 Ti Super Show 12 more