Performance Rating
MSI RTX 5090 LIGHTNING Z outperforms NVIDIA RTX PRO 4000 Blackwell Embedded by 213.97% in the overall GPU ARK performance rating
NVIDIA RTX PRO 4000 Blackwell Embedded
The MSI RTX 5090 LIGHTNING Z dramatically outperforms the NVIDIA RTX PRO 4000 Blackwell Embedded, leading by 214% in the overall GPI performance index. For machine learning, the MSI RTX 5090 LIGHTNING Z has the edge in FP16 performance (+354%) and tensor cores (680 vs 240). For gaming, the MSI RTX 5090 LIGHTNING Z delivers 354% more FP32 throughput, and the MSI RTX 5090 LIGHTNING Z has 183% more RT cores. In performance per watt, the NVIDIA RTX PRO 4000 Blackwell Embedded is 175% more efficient at 80W TDP vs 1000W for the MSI RTX 5090 LIGHTNING Z.
Contents:
Memory ML Performance Compute Power Architecture & Compatibility ML Software Support Clocks & Performance Power Consumption Rendering AdditionalMemory
Memory Size
+100%
32
ГБ
Memory Type
Memory Bandwidth
1.79 TB/s
Memory Bus Width
ML Performance
FP16 (Half Precision)
+354%
118.8
TFLOPS
BF16 (Brain Float)
TF32 (TensorFloat)
Compute Power
FP32 (Single Precision)
+354%
118.8
TFLOPS
FP64 (Double Precision)
+354%
1.856
TFLOPS
CUDA Cores
+183%
21,760
RT Cores
+183%
170
Architecture & Compatibility
GPU Architecture
SM (Streaming Multiprocessor)
+183%
170
PCIe Version
ML Software Support
CUDA Version
CUDA Toolkit (first supported)
CUDA Toolkit status
Clocks & Performance
Base Clock
+107%
2,017
Boost Clock
+60%
2,730
Memory Clock
1,750
Power Consumption
Recommended PSU
Power Connector
2x 16-pin
TDP/TGP
-92%
80 W
Rendering
Texture Units (TMU)
+183%
680
ROP
+183%
170
L2 Cache
+50%
96 MB
Additional
Slots
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
