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
Tesla M40
CORE STATE
GM200
VRAM
12 GB
CLOCK SPEED
1112 MHz
TDP
250 W
BUS WIDTH
384 bit
ARCHITECTURE
Maxwell 2.0
PROCESS
28 nm
LAUNCH DATE
2015
PERFORMANCE BENCHMARKS
geekbench_opencl
geekbench_vulkan
DETAILED SPECIFICATIONS
SPECIFICATION
B200
Tesla M40
Core Specs
Shading Units
18,944
3,072
-83.8%
Shaders
18,944
3,072
-83.8%
TMUs
592
192
-67.6%
ROPs
24
96
+300.0%
SM Count
148
—
Clocks
Base Clock
700 MHz
948 MHz
Boost Clock
1965 MHz
1112 MHz
Memory Clock
2000 MHz
8 Gbps effective
1502 MHz
6 Gbps effective
Memory
Memory Size
90 GB
12 GB
VRAM (MB)
92,160
12,288
-86.7%
Memory Type
HBM3e
GDDR5
Memory Bus
4096 bit
384 bit
Bandwidth
4.10 TB/s
288.4 GB/s
Cache
L1 Cache
256 KB (per SM)
48 KB (per SMM)
L2 Cache
50 MB
3 MB
Performance
Pixel Rate
47.16 GPixel/s
106.8 GPixel/s
Texture Rate
1,163.3 GTexel/s
213.5 GTexel/s
FP32 (TFLOPS)
74.45 TFLOPS
6.832 TFLOPS
FP64 (TFLOPS)
37.22 TFLOPS (1:2)
213.5 GFLOPS (1:32)
FP16 (TFLOPS)
1,191.2 TFLOPS (16:1)
—
AI/RT
Tensor Cores
592
—
Power
TDP
1000 W
250 W
TDP (W)
1,000
250
-75.0%
Suggested PSU
1400 W
600 W
Power Connectors
—
8-pin EPS
Architecture
Architecture
Blackwell
Maxwell 2.0
GPU Name
GB100
GM200
Generation
Server Blackwell
(Bxx)
Tesla Maxwell
(Mxx)
Process Size
5 nm
28 nm
Transistors
104,000 million
8,000 million
Die Size
—
601 mm²
Foundry
TSMC
TSMC
Density
—
13.3M / mm²
API Support
DirectX
—
12 (12_1)
OpenGL
—
4.6
Vulkan
—
1.4
OpenCL
3.0
3.0
CUDA
10.0
5.2
Shader Model
—
6.8
Physical
Slot Width
SXM Module
Dual-slot
Length
—
267 mm
10.5 inches
Outputs
No outputs
No outputs
Bus Interface
PCIe 5.0 x16
PCIe 3.0 x16
Other
Production
Active
End-of-life
Predecessor
Server Hopper
Tesla Kepler
Successor
Server Rubin
Tesla Pascal