GPU Comparison
GEFORCE
NVIDIA P104-100
CORE STATE
GP104
VRAM
4 GB
CLOCK SPEED
1733 MHz
TDP
—
BUS WIDTH
256 bit
ARCHITECTURE
Pascal
PROCESS
16 nm
LAUNCH DATE
2017
VS
GEFORCE
PG506-232
CORE STATE
GA100
VRAM
24 GB
CLOCK SPEED
1440 MHz
TDP
165 W
BUS WIDTH
3072 bit
ARCHITECTURE
Ampere
PROCESS
7 nm
LAUNCH DATE
2021
PERFORMANCE BENCHMARKS
3dmark_3dmark_steel_nomad_dx12
geekbench_opencl
geekbench_vulkan
DETAILED SPECIFICATIONS
SPECIFICATION
P104-100
PG506-232
Core Specs
Shading Units
1,920
3,584
+86.7%
Shaders
1,920
3,584
+86.7%
TMUs
120
224
+86.7%
ROPs
64
96
+50.0%
SM Count
15
56
+273.3%
Clocks
Base Clock
1607 MHz
930 MHz
Boost Clock
1733 MHz
1440 MHz
Memory Clock
1251 MHz
10 Gbps effective
1215 MHz
2.4 Gbps effective
Memory
Memory Size
4 GB
24 GB
VRAM (MB)
4,096
24,576
+500.0%
Memory Type
GDDR5X
HBM2
Memory Bus
256 bit
3072 bit
Bandwidth
320.3 GB/s
933.1 GB/s
Cache
L1 Cache
48 KB (per SM)
192 KB (per SM)
L2 Cache
2 MB
24 MB
Performance
Pixel Rate
110.9 GPixel/s
138.2 GPixel/s
Texture Rate
208.0 GTexel/s
322.6 GTexel/s
FP32 (TFLOPS)
6.655 TFLOPS
10.32 TFLOPS
FP64 (TFLOPS)
208.0 GFLOPS (1:32)
5.161 TFLOPS (1:2)
FP16 (TFLOPS)
104.0 GFLOPS (1:64)
10.32 TFLOPS (1:1)
AI/RT
Tensor Cores
—
224
Power
TDP
—
165 W
TDP (W)
—
165
Suggested PSU
200 W
450 W
Power Connectors
1x 8-pin
8-pin EPS
Architecture
Architecture
Pascal
Ampere
GPU Name
GP104
GA100
Generation
Mining GPUs
Server Ampere
(Axx)
Process Size
16 nm
7 nm
Transistors
7,200 million
54,200 million
Die Size
314 mm²
826 mm²
Foundry
TSMC
TSMC
Density
22.9M / mm²
65.6M / mm²
API Support
DirectX
12 (12_1)
—
OpenGL
4.6
—
Vulkan
1.4
—
OpenCL
3.0
3.0
CUDA
6.1
8.0
Shader Model
6.8
—
Physical
Slot Width
Dual-slot
Dual-slot
Length
267 mm
10.5 inches
267 mm
10.5 inches
Height
—
112 mm
4.4 inches
Outputs
No outputs
No outputs
Bus Interface
PCIe 1.0 x4
PCIe 4.0 x16
Other
Production
End-of-life
End-of-life
Predecessor
—
Tesla Turing
Successor
—
Server Ada