Performance Rating
NVIDIA B200 outperforms GIGABYTE AORUS RTX 5090 INFINITY by 39.46% in the overall GPU ARK performance rating
NVIDIA B200
The NVIDIA B200 significantly outperforms the GIGABYTE AORUS RTX 5090 INFINITY, leading by 39% in the overall GPI performance index. For machine learning, the GIGABYTE AORUS RTX 5090 INFINITY has the edge in FP16 performance (+9873%). For gaming, the GIGABYTE AORUS RTX 5090 INFINITY delivers 60% more FP32 throughput, and the GIGABYTE AORUS RTX 5090 INFINITY supports hardware ray tracing. In performance per watt, the GIGABYTE AORUS RTX 5090 INFINITY is 178% more efficient at 575W TDP vs 1000W for the NVIDIA B200.
Contents:
Memory ML Performance Compute Power Architecture & Compatibility ML Software Support Clocks & Performance Power Consumption Rendering Benchmarks AdditionalMemory
Memory Size
+181%
90
ГБ ×2 (180 ГБ)
Memory Type
Memory Bandwidth
+129%
4.10 TB/s
×2 (8.2 TB/s)
Memory Bus Width
ML Performance
FP16 (Half Precision)
+9,873%
118.8
TFLOPS
BF16 (Brain Float)
TF32 (TensorFloat)
Compute Power
FP32 (Single Precision)
+60%
118.8
TFLOPS
FP64 (Double Precision)
+1,905%
37.22
TFLOPS
CUDA Cores
+15%
21,760
RT Cores
170
Architecture & Compatibility
GPU Architecture
SM (Streaming Multiprocessor)
+15%
170
PCIe Version
ML Software Support
CUDA Version
12.0
CUDA Toolkit (first supported)
CUDA Toolkit status
Clocks & Performance
Base Clock
+188%
2,017
Boost Clock
+39%
2,730
Memory Clock
+14%
2,000
Power Consumption
Recommended PSU
-32%
950 W
Power Connector
TDP/TGP
-42%
575 W
Rendering
Texture Units (TMU)
+15%
680
ROP
170
L2 Cache
+92%
96 MB
Benchmarks
llama.cpp, llama 7B Q4_0
Additional
Slots
SXM Module
Release Date
Display Outputs
Renting is cheaper than buying
H100 80Gb SXM (80GB)
NVIDIA
Флагманская GPU на архитектуре Hopper с 80 ГБ памяти HBM3 в форм-факторе SXM. Создана для обучения и инференса больших языковых моделей, HPC задач. Поддерживает FP64, FP32, TF32, FP16, BF16, FP8 и INT8 вычисления с Tensor Cores 4-го поколения.
from 299.11 RUB/hour
