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
GeForce RTX 4070 SUPER
CORE STATE
AD104
VRAM
12 GB
CLOCK SPEED
2475 MHz
TDP
220 W
BUS WIDTH
192 bit
ARCHITECTURE
Ada Lovelace
PROCESS
5 nm
LAUNCH DATE
2024
PERFORMANCE BENCHMARKS
geekbench_opencl
geekbench_vulkan
3dmark_3dmark_steel_nomad_dx12
passmark_directx_10
passmark_directx_11
passmark_directx_12
passmark_directx_9
passmark_g2d
passmark_g3d
passmark_gpu_compute
Analysis: NVIDIA GB10 vs NVIDIA GeForce RTX 4070 SUPER
the NVIDIA RTX 4070 SUPER and GeForce NVIDIA GB10 provide almost matching gaming results in our comparative test. The score variations between both GPUs generally sit within single digits, positioning either a solid selection. This balance positions the decision interesting. Other elements become increasingly important. Value usually sways the choice.
At the end of the day: the NVIDIA RTX 4070 SUPER and GeForce NVIDIA GB10 are similarly matched. No earns a undisputed preference over the competitor. Make availability and driver factors guide your eventual choice. Either options will serve you well. This is a win-win scenario.
DETAILED SPECIFICATIONS
SPECIFICATION
GB10
RTX 4070 SUPER
Core Specs
Shading Units
6,144
7,168
+16.7%
Shaders
6,144
7,168
+16.7%
TMUs
384
224
-41.7%
ROPs
48
80
+66.7%
SM Count
48
56
+16.7%
Clocks
Base Clock
1665 MHz
1980 MHz
Boost Clock
2418 MHz
2475 MHz
Memory Clock
1067 MHz
8.5 Gbps effective
1313 MHz
21 Gbps effective
Memory
Memory Size
128 GB
12 GB
VRAM (MB)
131,072
12,288
-90.6%
Memory Type
LPDDR5X
GDDR6X
Memory Bus
256 bit
192 bit
Bandwidth
273.2 GB/s
504.2 GB/s
Cache
L1 Cache
256 KB (per SM)
128 KB (per SM)
L2 Cache
50 MB
48 MB
Performance
Pixel Rate
116.1 GPixel/s
198.0 GPixel/s
Texture Rate
928.5 GTexel/s
554.4 GTexel/s
FP32 (TFLOPS)
29.71 TFLOPS
35.48 TFLOPS
FP64 (TFLOPS)
464.3 GFLOPS (1:64)
554.4 GFLOPS (1:64)
FP16 (TFLOPS)
29.71 TFLOPS (1:1)
35.48 TFLOPS (1:1)
AI/RT
RT Cores
48
56
+16.7%
Tensor Cores
384
224
-41.7%
Power
TDP
140 W
220 W
TDP (W)
140
220
+57.1%
Suggested PSU
300 W
550 W
Power Connectors
None
1x 16-pin
Architecture
Architecture
Blackwell 2.0
Ada Lovelace
GPU Name
GB20B
AD104
Generation
Server Blackwell
(Bxx)
GeForce 40
Process Size
5 nm
5 nm
Transistors
—
35,800 million
Die Size
—
294 mm²
Foundry
TSMC
TSMC
Density
—
121.8M / mm²
API Support
DirectX
—
12 Ultimate (12_2)
OpenGL
—
4.6
Vulkan
—
1.4
OpenCL
3.0
3.0
CUDA
12.1
8.9
Shader Model
—
6.8
Physical
Slot Width
IGP
Dual-slot
Length
150 mm
5.9 inches
267 mm
10.5 inches
Height
51 mm
2 inches
112 mm
4.4 inches
Outputs
1x HDMI
1x HDMI 2.13x DisplayPort 1.4a
Bus Interface
PCIe 5.0 x16
PCIe 4.0 x16
Other
Launch Price
3,999 USD
599 USD
Production
Active
End-of-life
Predecessor
Server Hopper
GeForce 30
Successor
Server Rubin
GeForce 50