NVIDIA Quadro RTX 5000

The NVIDIA Quadro RTX 5000 is a workstation graphics card that launched in Q3 2018 with a MSRP of $2,299. It is built on the Turing GPU microarchitecture (codename TU104) and is manufactured on a 12 nm process.

Memory

The RTX 5000 has 16 GB of GDDR6 memory, with a 1,750 MHz memory clock and a 256 bit interface. This gives it a memory bandwidth of 448 Gb/s, which affects how fast it can transfer data to and from memory. GPU memory stores temporary data that helps the GPU with complex math and graphics operations. More memory is generally better, as not having enough can cause performance bottlenecks.

Cores and Clock Speeds

The RTX 5000 includes 3,072 CUDA cores, the processing units for handling parallel computing tasks. The GPU operates at a core clock speed of 1,620 MHz and can dynamically boost its clock speed up to 1,815 MHz. Complementing the processing units are 192 texture mapping units (TMUs) for efficient texture filtering and 64 render output units (ROPs) for pixel processing. Additionally, the GPU features 384 tensor cores optimized for AI-accelerated workloads and 48 RT cores dedicated to real-time ray tracing calculations.

Compatibility & Power Consumption

The RTX 5000 occupies 2 PCIe expansion slots. It supports DisplayPort 1.4a, USB Type-C display connections. NVIDIA recommends a power supply of at least 550 W to handle the GPU's thermal design power (TDP) of 230 W.