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
Tesla M2090
CORE STATE
GF110
VRAM
6 GB
CLOCK SPEED
—
TDP
250 W
BUS WIDTH
384 bit
ARCHITECTURE
Fermi 2.0
PROCESS
40 nm
LAUNCH DATE
2011
PERFORMANCE BENCHMARKS
geekbench_opencl
geekbench_vulkan
DETAILED SPECIFICATIONS
SPECIFICATION
GB10
Tesla M2090
Core Specs
Shading Units
6,144
512
-91.7%
Shaders
6,144
512
-91.7%
TMUs
384
64
-83.3%
ROPs
48
48
0.0%
SM Count
48
16
-66.7%
Clocks
Base Clock
1665 MHz
—
Boost Clock
2418 MHz
—
GPU Clock
—
651 MHz
Shader Clock
—
1301 MHz
Memory Clock
1067 MHz
8.5 Gbps effective
924 MHz
3.7 Gbps effective
Memory
Memory Size
128 GB
6 GB
VRAM (MB)
131,072
6,144
-95.3%
Memory Type
LPDDR5X
GDDR5
Memory Bus
256 bit
384 bit
Bandwidth
273.2 GB/s
177.4 GB/s
Cache
L1 Cache
256 KB (per SM)
64 KB (per SM)
L2 Cache
50 MB
768 KB
Performance
Pixel Rate
116.1 GPixel/s
20.83 GPixel/s
Texture Rate
928.5 GTexel/s
41.66 GTexel/s
FP32 (TFLOPS)
29.71 TFLOPS
1,332.2 GFLOPS
FP64 (TFLOPS)
464.3 GFLOPS (1:64)
666.1 GFLOPS (1:2)
FP16 (TFLOPS)
29.71 TFLOPS (1:1)
—
AI/RT
RT Cores
48
—
Tensor Cores
384
—
Power
TDP
140 W
250 W
TDP (W)
140
250
+78.6%
Suggested PSU
300 W
600 W
Power Connectors
None
1x 6-pin + 1x 8-pin
Architecture
Architecture
Blackwell 2.0
Fermi 2.0
GPU Name
GB20B
GF110
Generation
Server Blackwell
(Bxx)
Tesla Fermi
(x20xx)
Process Size
5 nm
40 nm
Transistors
—
3,000 million
Die Size
—
520 mm²
Foundry
TSMC
TSMC
Density
—
5.8M / mm²
API Support
DirectX
—
12 (11_0)
OpenGL
—
4.6
OpenCL
3.0
1.1
CUDA
12.1
2.0
Shader Model
—
5.1
Physical
Slot Width
IGP
Dual-slot
Length
150 mm
5.9 inches
248 mm
9.8 inches
Height
51 mm
2 inches
—
Outputs
1x HDMI
No outputs
Bus Interface
PCIe 5.0 x16
PCIe 2.0 x16
Other
Launch Price
3,999 USD
—
Production
Active
End-of-life
Predecessor
Server Hopper
Tesla
Successor
Server Rubin
Tesla Kepler