GPU Comparison
GEFORCE
NVIDIA A2
CORE STATE
GA107
VRAM
16 GB
CLOCK SPEED
1770 MHz
TDP
60 W
BUS WIDTH
128 bit
ARCHITECTURE
Ampere
PROCESS
8 nm
LAUNCH DATE
2021
VS
GEFORCE
GB10
CORE STATE
GB20B
VRAM
128 GB
CLOCK SPEED
2418 MHz
TDP
140 W
BUS WIDTH
256 bit
ARCHITECTURE
Blackwell 2.0
PROCESS
5 nm
LAUNCH DATE
2025
PERFORMANCE BENCHMARKS
geekbench_opencl
geekbench_vulkan
Analysis: NVIDIA A2 vs NVIDIA GB10
When gaming capability is remarkably tightly balanced, the analysis between NVIDIA A2 from NVIDIA and NVIDIA GB10 by NVIDIA grows more multifaceted. The two options trade victories over different games. The total assessment is one of remarkable equality. None of the stands as a undisputed champion. This selection hinges on specific needs.
For people conflicted between NVIDIA A2 from NVIDIA and NVIDIA GB10 by NVIDIA, feel certain that any selection achieves comparable rendering results. This benchmark results supports no obvious leader. Make your choice on alternative considerations and experience excellent visual results in any case. Both are excellent in this particular sense. Choose assuredly.
DETAILED SPECIFICATIONS
SPECIFICATION
A2
GB10
Core Specs
Shading Units
1,280
6,144
+380.0%
Shaders
1,280
6,144
+380.0%
TMUs
40
384
+860.0%
ROPs
32
48
+50.0%
SM Count
10
48
+380.0%
Clocks
Base Clock
1440 MHz
1665 MHz
Boost Clock
1770 MHz
2418 MHz
Memory Clock
1563 MHz
12.5 Gbps effective
1067 MHz
8.5 Gbps effective
Memory
Memory Size
16 GB
128 GB
VRAM (MB)
16,384
131,072
+700.0%
Memory Type
GDDR6
LPDDR5X
Memory Bus
128 bit
256 bit
Bandwidth
200.1 GB/s
273.2 GB/s
Cache
L1 Cache
128 KB (per SM)
256 KB (per SM)
L2 Cache
2 MB
50 MB
Performance
Pixel Rate
56.64 GPixel/s
116.1 GPixel/s
Texture Rate
70.80 GTexel/s
928.5 GTexel/s
FP32 (TFLOPS)
4.531 TFLOPS
29.71 TFLOPS
FP64 (TFLOPS)
70.80 GFLOPS (1:64)
464.3 GFLOPS (1:64)
FP16 (TFLOPS)
4.531 TFLOPS (1:1)
29.71 TFLOPS (1:1)
AI/RT
RT Cores
10
48
+380.0%
Tensor Cores
40
384
+860.0%
Power
TDP
60 W
140 W
TDP (W)
60
140
+133.3%
Suggested PSU
250 W
300 W
Power Connectors
None
None
Architecture
Architecture
Ampere
Blackwell 2.0
GPU Name
GA107
GB20B
Generation
Workstation Ampere
(Ax000)
Server Blackwell
(Bxx)
Process Size
8 nm
5 nm
Transistors
8,700 million
—
Die Size
200 mm²
—
Foundry
Samsung
TSMC
Density
43.5M / mm²
—
API Support
DirectX
12 Ultimate (12_2)
—
OpenGL
4.6
—
Vulkan
1.4
—
OpenCL
3.0
3.0
CUDA
8.6
12.1
Shader Model
6.8
—
Physical
Slot Width
Single-slot
IGP
Length
—
150 mm
5.9 inches
Height
—
51 mm
2 inches
Outputs
No outputs
1x HDMI
Bus Interface
PCIe 4.0 x8
PCIe 5.0 x16
Other
Launch Price
—
3,999 USD
Production
End-of-life
Active
Predecessor
Quadro Turing
Server Hopper
Successor
Workstation Ada
Server Rubin