Performance Rating
The RTX PRO 4000 Blackwell SFF is a professional graphics card by NVIDIA, launched on August 11th, 2025. Built on the 5 nm process, and based on the GB203 graphics processor, the card supports DirectX 12 Ultimate. The GB203 graphics processor is a large chip with a die area of 378 mm² and 45,600 million transistors. Unlike the fully unlocked GeForce RTX 5080, which uses the same GPU but has all 10752 shaders enabled, NVIDIA has disabled some shading units on the RTX PRO 4000 Blackwell SFF to reach the product's target shader count. It features 8960 shading units, 280 texture mapping units, and 96 ROPs. Also included are 280 tensor cores which help improve the speed of machine learning applications. The card also has 70 raytracing acceleration cores. NVIDIA has paired 24 GB GDDR7 memory with the RTX PRO 4000 Blackwell SFF, which are connected using a 192-bit memory interface. The GPU is operating at a frequency of 577 MHz, which can be boosted up to 1432 MHz, memory is running at 1125 MHz (18 Gbps effective).
Contents:
Memory ML Performance Compute Power Architecture & Compatibility ML Software Support Clocks & Performance Power Consumption Rendering AdditionalMemory
Memory Size
Memory Type
Memory Bandwidth
Memory Bus Width
ML Performance
FP16 (Half Precision)
BF16 (Brain Float)
TF32 (TensorFloat)
Compute Power
FP32 (Single Precision)
FP64 (Double Precision)
CUDA Cores
RT Cores
Architecture & Compatibility
GPU Architecture
SM (Streaming Multiprocessor)
PCIe Version
ML Software Support
CUDA Version
CUDA Toolkit (first supported)
CUDA Toolkit status
Clocks & Performance
Base Clock
Boost Clock
Memory Clock
Power Consumption
Recommended PSU
Power Connector
TDP/TGP
Rendering
Texture Units (TMU)
ROP
L2 Cache
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
