GPU Comparison
GEFORCE
NVIDIA B200
CORE STATE
GB100
VRAM
90 GB
CLOCK SPEED
1965 MHz
TDP
1000 W
BUS WIDTH
4096 bit
ARCHITECTURE
Blackwell
PROCESS
5 nm
LAUNCH DATE
—
VS
GEFORCE
P106-100
CORE STATE
GP106
VRAM
6 GB
CLOCK SPEED
1709 MHz
TDP
120 W
BUS WIDTH
192 bit
ARCHITECTURE
Pascal
PROCESS
16 nm
LAUNCH DATE
2017
PERFORMANCE BENCHMARKS
geekbench_opencl
3dmark_3dmark_steel_nomad_dx12
geekbench_vulkan
DETAILED SPECIFICATIONS
SPECIFICATION
B200
P106-100
Core Specs
Shading Units
18,944
1,280
-93.2%
Shaders
18,944
1,280
-93.2%
TMUs
592
80
-86.5%
ROPs
24
48
+100.0%
SM Count
148
10
-93.2%
Clocks
Base Clock
700 MHz
1506 MHz
Boost Clock
1965 MHz
1709 MHz
Memory Clock
2000 MHz
8 Gbps effective
2002 MHz
8 Gbps effective
Memory
Memory Size
90 GB
6 GB
VRAM (MB)
92,160
6,144
-93.3%
Memory Type
HBM3e
GDDR5
Memory Bus
4096 bit
192 bit
Bandwidth
4.10 TB/s
192.2 GB/s
Cache
L1 Cache
256 KB (per SM)
48 KB (per SM)
L2 Cache
50 MB
1536 KB
Performance
Pixel Rate
47.16 GPixel/s
82.03 GPixel/s
Texture Rate
1,163.3 GTexel/s
136.7 GTexel/s
FP32 (TFLOPS)
74.45 TFLOPS
4.375 TFLOPS
FP64 (TFLOPS)
37.22 TFLOPS (1:2)
136.7 GFLOPS (1:32)
FP16 (TFLOPS)
1,191.2 TFLOPS (16:1)
68.36 GFLOPS (1:64)
AI/RT
Tensor Cores
592
—
Power
TDP
1000 W
120 W
TDP (W)
1,000
120
-88.0%
Suggested PSU
1400 W
300 W
Power Connectors
—
1x 6-pin
Architecture
Architecture
Blackwell
Pascal
GPU Name
GB100
GP106
Generation
Server Blackwell
(Bxx)
Mining GPUs
Process Size
5 nm
16 nm
Transistors
104,000 million
4,400 million
Die Size
—
200 mm²
Foundry
TSMC
TSMC
Density
—
22.0M / mm²
API Support
DirectX
—
12 (12_1)
OpenGL
—
4.6
Vulkan
—
1.4
OpenCL
3.0
3.0
CUDA
10.0
6.1
Shader Model
—
6.8
Physical
Slot Width
SXM Module
Dual-slot
Length
—
250 mm
9.8 inches
Outputs
No outputs
No outputs
Bus Interface
PCIe 5.0 x16
PCIe 1.0 x16
Other
Production
Active
End-of-life
Predecessor
Server Hopper
—
Successor
Server Rubin
—