Performance Rating
NVIDIA CMP 70HX outperforms NVIDIA NVS 810 by 909.21% in the overall GPU ARK performance rating
NVIDIA NVS 810
Expert Comparison
NVIDIA CMP 70HX明显优于NVIDIA NVS 810。CMP 70HX具有更强大的Ampere架构，更多的3840个核心，8GB GDDR6X内存和更高的性能指标（如608.3 GB/s的带宽和10.71 TFLOPS的FP32性能），适用于高端游戏和专业图形任务。而NVS 810基于较旧的Maxwell架构，只有512个核心，2GB DDR3内存，性能指标较低，更适合基本图形处理和轻量级任务。
Contents:
Memory ML Performance Compute Power Architecture & Compatibility ML Software Support Clocks & Performance Power Consumption Rendering AdditionalMemory
Memory Size
+300%
8
ГБ
Memory Type
Memory Bandwidth
Memory Bus Width
ML Performance
FP16 (Half Precision)
10.71
TFLOPS
BF16 (Brain Float)
TF32 (TensorFloat)
Compute Power
FP32 (Single Precision)
+912%
10.71
TFLOPS
FP64 (Double Precision)
+406%
0.1674
TFLOPS
CUDA Cores
+650%
3,840
RT Cores
30
Architecture & Compatibility
GPU Architecture
SM (Streaming Multiprocessor)
30
PCIe Version
ML Software Support
CUDA Version
8.6
CUDA Toolkit (first supported)
CUDA Toolkit (deprecated)
CUDA Toolkit (support removed)
CUDA Toolkit status
Clocks & Performance
Base Clock
+51%
1,365
Boost Clock
+35%
1,395
Memory Clock
+32%
1,188
Power Consumption
Recommended PSU
-20%
200 W
Power Connector
TDP/TGP
Rendering
Texture Units (TMU)
+275%
120
ROP
30
L2 Cache
4 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
