GPU Comparison
GEFORCE
NVIDIA 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
VS
GEFORCE
RTX A400
CORE STATE
GA107
VRAM
4 GB
CLOCK SPEED
1762 MHz
TDP
50 W
BUS WIDTH
64 bit
ARCHITECTURE
Ampere
PROCESS
8 nm
LAUNCH DATE
2024
PERFORMANCE BENCHMARKS
geekbench_opencl
geekbench_vulkan
Analysis: NVIDIA GB10 vs NVIDIA RTX A400
When gaming capability is this evenly balanced, the matchup between the NVIDIA RTX A400 and GeForce NVIDIA GB10 grows particularly interesting. The two options exchange leads throughout various scenarios. This overall outcome is one of notable parity. Not one appears as a undisputed winner. The choice depends on personal preferences.
In the end, builders could decide based on availability over pure performance solely. Either the NVIDIA RTX A400 and GeForce NVIDIA GB10 achieve strong rendering capability for their respective price points. This determination depends on personal priorities. Each option is solid. Buyers won't go amiss here.
DETAILED SPECIFICATIONS
SPECIFICATION
GB10
RTX A400
Core Specs
Shading Units
6,144
768
-87.5%
Shaders
6,144
768
-87.5%
TMUs
384
24
-93.8%
ROPs
48
16
-66.7%
SM Count
48
6
-87.5%
Clocks
Base Clock
1665 MHz
1417 MHz
Boost Clock
2418 MHz
1762 MHz
Memory Clock
1067 MHz
8.5 Gbps effective
1500 MHz
12 Gbps effective
Memory
Memory Size
128 GB
4 GB
VRAM (MB)
131,072
4,096
-96.9%
Memory Type
LPDDR5X
GDDR6
Memory Bus
256 bit
64 bit
Bandwidth
273.2 GB/s
96.00 GB/s
Cache
L1 Cache
256 KB (per SM)
128 KB (per SM)
L2 Cache
50 MB
2 MB
Performance
Pixel Rate
116.1 GPixel/s
28.19 GPixel/s
Texture Rate
928.5 GTexel/s
42.29 GTexel/s
FP32 (TFLOPS)
29.71 TFLOPS
2.706 TFLOPS
FP64 (TFLOPS)
464.3 GFLOPS (1:64)
42.29 GFLOPS (1:64)
FP16 (TFLOPS)
29.71 TFLOPS (1:1)
2.706 TFLOPS (1:1)
AI/RT
RT Cores
48
6
-87.5%
Tensor Cores
384
24
-93.8%
Power
TDP
140 W
50 W
TDP (W)
140
50
-64.3%
Suggested PSU
300 W
250 W
Power Connectors
None
None
Architecture
Architecture
Blackwell 2.0
Ampere
GPU Name
GB20B
GA107
Generation
Server Blackwell
(Bxx)
Workstation Ampere
(Ax000)
Process Size
5 nm
8 nm
Transistors
—
8,700 million
Die Size
—
200 mm²
Foundry
TSMC
Samsung
Density
—
43.5M / mm²
API Support
DirectX
—
12 Ultimate (12_2)
OpenGL
—
4.6
Vulkan
—
1.4
OpenCL
3.0
3.0
CUDA
12.1
8.6
Shader Model
—
6.8
Physical
Slot Width
IGP
Single-slot
Length
150 mm
5.9 inches
163 mm
6.4 inches
Height
51 mm
2 inches
69 mm
2.7 inches
Outputs
1x HDMI
4x mini-DisplayPort 1.4a
Bus Interface
PCIe 5.0 x16
PCIe 4.0 x8
Other
Launch Price
3,999 USD
—
Production
Active
Active
Predecessor
Server Hopper
Quadro Turing
Successor
Server Rubin
Workstation Ada