GPU Comparison
GEFORCE
NVIDIA H200 NVL
CORE STATE
GH100
VRAM
141 GB
CLOCK SPEED
1785 MHz
TDP
600 W
BUS WIDTH
6144 bit
ARCHITECTURE
Hopper
PROCESS
5 nm
LAUNCH DATE
2024
VS
GEFORCE
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
PERFORMANCE BENCHMARKS
geekbench_opencl
geekbench_vulkan
DETAILED SPECIFICATIONS
SPECIFICATION
H200 NVL
L40S
Core Specs
Shading Units
16,896
18,176
+7.6%
Shaders
16,896
18,176
+7.6%
TMUs
528
568
+7.6%
ROPs
24
192
+700.0%
SM Count
132
142
+7.6%
Clocks
Base Clock
1365 MHz
1110 MHz
Boost Clock
1785 MHz
2520 MHz
Memory Clock
1593 MHz
6.4 Gbps effective
2250 MHz
18 Gbps effective
Memory
Memory Size
141 GB
48 GB
VRAM (MB)
144,384
49,152
-66.0%
Memory Type
HBM3e
GDDR6
Memory Bus
6144 bit
384 bit
Bandwidth
4.89 TB/s
864.0 GB/s
Cache
L1 Cache
256 KB (per SM)
128 KB (per SM)
L2 Cache
50 MB
48 MB
Performance
Pixel Rate
42.84 GPixel/s
483.8 GPixel/s
Texture Rate
942.5 GTexel/s
1,431.4 GTexel/s
FP32 (TFLOPS)
60.32 TFLOPS
91.61 TFLOPS
FP64 (TFLOPS)
30.16 TFLOPS (1:2)
1,431.4 GFLOPS (1:64)
FP16 (TFLOPS)
241.3 TFLOPS (4:1)
91.61 TFLOPS (1:1)
AI/RT
RT Cores
—
142
Tensor Cores
528
568
+7.6%
Power
TDP
600 W
300 W
TDP (W)
600
300
-50.0%
Suggested PSU
1000 W
700 W
Power Connectors
8-pin EPS
1x 16-pin
Architecture
Architecture
Hopper
Ada Lovelace
GPU Name
GH100
AD102
Generation
Server Hopper
(Hxx)
Server Ada
(Lxx)
Process Size
5 nm
5 nm
Transistors
80,000 million
76,300 million
Die Size
814 mm²
609 mm²
Foundry
TSMC
TSMC
Density
98.3M / mm²
125.3M / mm²
API Support
DirectX
—
12 Ultimate (12_2)
OpenGL
—
4.6
Vulkan
—
1.4
OpenCL
3.0
3.0
CUDA
9.0
8.9
Shader Model
—
6.8
Physical
Slot Width
Dual-slot
Dual-slot
Length
267 mm
10.5 inches
267 mm
10.5 inches
Height
111 mm
4.4 inches
111 mm
4.4 inches
Outputs
No outputs
1x HDMI 2.13x DisplayPort 1.4a
Bus Interface
PCIe 5.0 x16
PCIe 4.0 x16
Other
Production
Active
End-of-life
Predecessor
Server Ada
Server Ampere
Successor
Server Blackwell
Server Hopper