GPU Comparison
GEFORCE
NVIDIA L40S
CORE STATE
AD102
VRAM
48 GB
CLOCK SPEED
2520 MHz
TDP
300 W
BUS WIDTH
384 bit
ARCHITECTURE
Ada Lovelace
PROCESS
5 nm
LAUNCH DATE
2022
VS
GEFORCE
P106-090
CORE STATE
GP106
VRAM
3 GB
CLOCK SPEED
1531 MHz
TDP
75 W
BUS WIDTH
192 bit
ARCHITECTURE
Pascal
PROCESS
16 nm
LAUNCH DATE
2017
PERFORMANCE BENCHMARKS
geekbench_opencl
geekbench_vulkan
3dmark_3dmark_steel_nomad_dx12
DETAILED SPECIFICATIONS
SPECIFICATION
L40S
P106-090
Core Specs
Shading Units
18,176
768
-95.8%
Shaders
18,176
768
-95.8%
TMUs
568
48
-91.5%
ROPs
192
48
-75.0%
SM Count
142
6
-95.8%
Clocks
Base Clock
1110 MHz
1354 MHz
Boost Clock
2520 MHz
1531 MHz
Memory Clock
2250 MHz
18 Gbps effective
2002 MHz
8 Gbps effective
Memory
Memory Size
48 GB
3 GB
VRAM (MB)
49,152
3,072
-93.8%
Memory Type
GDDR6
GDDR5
Memory Bus
384 bit
192 bit
Bandwidth
864.0 GB/s
192.2 GB/s
Cache
L1 Cache
128 KB (per SM)
48 KB (per SM)
L2 Cache
48 MB
1536 KB
Performance
Pixel Rate
483.8 GPixel/s
73.49 GPixel/s
Texture Rate
1,431.4 GTexel/s
73.49 GTexel/s
FP32 (TFLOPS)
91.61 TFLOPS
2.352 TFLOPS
FP64 (TFLOPS)
1,431.4 GFLOPS (1:64)
73.49 GFLOPS (1:32)
FP16 (TFLOPS)
91.61 TFLOPS (1:1)
36.74 GFLOPS (1:64)
AI/RT
RT Cores
142
—
Tensor Cores
568
—
Power
TDP
300 W
75 W
TDP (W)
300
75
-75.0%
Suggested PSU
700 W
250 W
Power Connectors
1x 16-pin
1x 6-pin
Architecture
Architecture
Ada Lovelace
Pascal
GPU Name
AD102
GP106
Generation
Server Ada
(Lxx)
Mining GPUs
Process Size
5 nm
16 nm
Transistors
76,300 million
4,400 million
Die Size
609 mm²
200 mm²
Foundry
TSMC
TSMC
Density
125.3M / mm²
22.0M / mm²
API Support
DirectX
12 Ultimate (12_2)
12 (12_1)
OpenGL
4.6
4.6
Vulkan
1.4
1.4
OpenCL
3.0
3.0
CUDA
8.9
6.1
Shader Model
6.8
6.8
Physical
Slot Width
Dual-slot
Dual-slot
Length
267 mm
10.5 inches
250 mm
9.8 inches
Height
111 mm
4.4 inches
—
Outputs
1x HDMI 2.13x DisplayPort 1.4a
No outputs
Bus Interface
PCIe 4.0 x16
PCIe 1.0 x1
Other
Production
End-of-life
End-of-life
Predecessor
Server Ampere
—
Successor
Server Hopper
—