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
Tesla P4
CORE STATE
GP104
VRAM
8 GB
CLOCK SPEED
1114 MHz
TDP
75 W
BUS WIDTH
256 bit
ARCHITECTURE
Pascal
PROCESS
16 nm
LAUNCH DATE
2016
PERFORMANCE BENCHMARKS
geekbench_opencl
geekbench_vulkan
DETAILED SPECIFICATIONS
SPECIFICATION
H200 NVL
Tesla P4
Core Specs
Shading Units
16,896
2,560
-84.8%
Shaders
16,896
2,560
-84.8%
TMUs
528
160
-69.7%
ROPs
24
64
+166.7%
SM Count
132
20
-84.8%
Clocks
Base Clock
1365 MHz
886 MHz
Boost Clock
1785 MHz
1114 MHz
Memory Clock
1593 MHz
6.4 Gbps effective
1502 MHz
6 Gbps effective
Memory
Memory Size
141 GB
8 GB
VRAM (MB)
144,384
8,192
-94.3%
Memory Type
HBM3e
GDDR5
Memory Bus
6144 bit
256 bit
Bandwidth
4.89 TB/s
192.3 GB/s
Cache
L1 Cache
256 KB (per SM)
48 KB (per SM)
L2 Cache
50 MB
2 MB
Performance
Pixel Rate
42.84 GPixel/s
71.30 GPixel/s
Texture Rate
942.5 GTexel/s
178.2 GTexel/s
FP32 (TFLOPS)
60.32 TFLOPS
5.704 TFLOPS
FP64 (TFLOPS)
30.16 TFLOPS (1:2)
178.2 GFLOPS (1:32)
FP16 (TFLOPS)
241.3 TFLOPS (4:1)
89.12 GFLOPS (1:64)
AI/RT
Tensor Cores
528
—
Power
TDP
600 W
75 W
TDP (W)
600
75
-87.5%
Suggested PSU
1000 W
250 W
Power Connectors
8-pin EPS
None
Architecture
Architecture
Hopper
Pascal
GPU Name
GH100
GP104
Generation
Server Hopper
(Hxx)
Tesla Pascal
(Pxx)
Process Size
5 nm
16 nm
Transistors
80,000 million
7,200 million
Die Size
814 mm²
314 mm²
Foundry
TSMC
TSMC
Density
98.3M / mm²
22.9M / mm²
API Support
DirectX
—
12 (12_1)
OpenGL
—
4.6
Vulkan
—
1.4
OpenCL
3.0
3.0
CUDA
9.0
6.1
Shader Model
—
6.8
Physical
Slot Width
Dual-slot
Single-slot
Length
267 mm
10.5 inches
168 mm
6.6 inches
Height
111 mm
4.4 inches
—
Outputs
No outputs
No outputs
Bus Interface
PCIe 5.0 x16
PCIe 3.0 x16
Other
Production
Active
End-of-life
Predecessor
Server Ada
Tesla Maxwell
Successor
Server Blackwell
Tesla Volta