NVIDIA L40S vs NVIDIA GB200 Full Specs

Summary

18,176 Shaders 2.52GHz	37,888 Shaders 2.11GHz
48GB GDDR6864GB/s	384GB HBM3e16TB/s
Peak·FP32·FP16 91.61 TFLOPS	Peak·FP32·FP16 160 TFLOPS
Form Factor PCIe Card	Form Factor Superchip
TDP 300W	TDP 2700W
Power Connectors 1x 16-Pin 12VHPWR	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 FP64 Tensor FP64 BF16 Tensor BF16 BF16 Tensor Sparse Tensor TF32 Tensor INT8 Tensor INT4 Memory Bandwidth TDP

L40S2.93 POPSINT4 Tensor Sparse

x1

GB20040 PFLOPSFP4 Tensor Sparse (2x 20 PFLOPS)

x13.65

Show Top Score

Theoretical Performance

Clock Speed2520MHz UC·Base·Boost·OC	Clock Speed2112MHz UC·Boost·OC
Peak OPS 2.93 POPSINT4 Tensor Sparse	Peak OPS 40 PFLOPSFP4 Tensor Sparse
-	Tensor FP4 20 PFLOPS FP4 Tensor Sparse 40 PFLOPS
Tensor FP8-16 732.9 TFLOPS FP8-16 Tensor Sparse 1.47 PFLOPS Tensor FP8-32 366.4 TFLOPS FP8-32 Tensor Sparse 732.9 TFLOPS	Tensor FP8-16 10 PFLOPS FP8-16 Tensor Sparse 20 PFLOPS Tensor FP8-32 10 PFLOPS FP8-32 Tensor Sparse 20 PFLOPS
Tensor FP16-16 366.4 TFLOPS FP16-16 Tensor Sparse 732.9 TFLOPS Tensor FP16-32 183.2 TFLOPS FP16-32 Tensor Sparse 366.4 TFLOPS	Tensor FP16-16 5 PFLOPS FP16-16 Tensor Sparse 10 PFLOPS Tensor FP16-32 5 PFLOPS FP16-32 Tensor Sparse 10 PFLOPS
BF16 91.61 TFLOPS Tensor BF16 183.2 TFLOPS BF16 Tensor Sparse 366.4 TFLOPS	BF16 320.1 TFLOPS Tensor BF16 5 PFLOPS BF16 Tensor Sparse 10 PFLOPS
Tensor TF32 91.61 TFLOPS	Tensor TF32 2.5 PFLOPS
FP32 91.61 TFLOPS	FP32 160 TFLOPS
FP64 1.43 TFLOPS Tensor FP64 -	FP64 80.02 TFLOPS Tensor FP64 78.13 TFLOPS
Tensor INT4 1.47 POPS	Tensor INT4 -
Tensor INT8 732.9 TOPS	Tensor INT8 10 POPS
Ray 211.7 TOPS	Ray -
Pixel Rate 483.8 GPixel/s	Pixel Rate 67.6 GPixel/s
Texture Rate 1.43 TTexel/s	Texture Rate 1.25 TTexel/s

Core Config

Shaders 18,176 Shaders	Shaders 37,888 Shaders
TMUs 568 TMUs	TMUs 1,184 TMUs
ROPs 192 ROPs	ROPs 64 ROPs
Tensor Cores 568 T-Cores	Tensor Cores 1,184 T-Cores
RT Cores 142 RT-Cores	RT Cores -
SMs 142 SMs	SMs 296 SMs

Clocks

Base Clock 1.11GHz	Base Clock -
Boost Clock 2.52GHz	Boost Clock 2.11GHz
Tensor Clock -	Tensor Clock 2.06GHz

Cache

L2 Cache 98.3MB shared

L2 Cache 65.5MB shared

Memory

48GB GDDR6	384GB HBM3e
Memory Bus 384-bit	Memory Bus 8192-bit
Memory Speed 18GT/s	Memory Speed 7.8GT/s
Memory Bandwidth 864GB/s	Memory Bandwidth 16TB/s

Power & Thermals

TDP 300W

TDP 2700W

Video Output

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

Ports

4x DisplayPort 1.4

-

Encoder

Encoder Model 2x NVENC 8

-

Decoder

Decoder Model NVDEC 5

Decoder Model 7x NVDEC 6

Card

Form Factor PCIe Card	Form Factor Superchip
PCIe 2-Slots	PCIe -
Height 111 mm (4.37") Width 267 mm (10.51") Depth 40 mm (1.57")	-
Cooling Passive	Cooling Passive
Power Connectors 1x 16-Pin 12VHPWR	Power Connectors -

Chip

Manufacturer NVIDIA	Manufacturer NVIDIA
Chip Designer NVIDIA	Chip Designer NVIDIA
Architecture Ada Lovelace	Architecture Blackwell
Family Server	Family Server
Branding	Branding
Codename NV182	Codename NV190
Chip Variant AD102-200-A1	Chip Variant Umbriel
Market Segment Server	Market Segment Server
Release Date Oct 13, 2022	Release Date Mar 18, 2024

Fabrication

Foundry TSMC	Foundry TSMC
Fabrication Node 4N	Fabrication Node 4NP
Die Size 608mm²	Die Size 1620mm²
Transistor Count 76.3 Billion	Transistor Count 208 Billion
Transistor Density 125.4 MTr/mm²	Transistor Density 128.4 MTr/mm²

Images

No images available

No images available

Competitors

NVIDIAL40S vs GPU NVIDIAB200 NVIDIAL40S vs GPU NVIDIARTX PRO 5000 Blackwell NVIDIAL40S vs GPU NVIDIARTX 5880 Ada Show 12 more

NVIDIAGB200 vs GPU NVIDIAB200 NVIDIAGB200 vs GPU NVIDIAB100 NVIDIAGB200 vs GPU AMDInstinct MI300X Show 8 more