Nvidia has unveiled the RTX A2000, the smallest and cheapest graphics card to date in its RTX-A graphics card series (which has replaced the Quadro series), which is based on Ampere technology and aimed at professional users in the design and video sectors. The RTX A2000, like its sister models, was produced via 8 nanometer process with its dimensions of 6.9 x 16.8 cm is specifically designed for PCs that are short on space but still need to offer as much GPU power as possible.
Nvidia RTX A2000
It consumes a maximum of 70 watts of power and uses one PCIe 4.0 x16 slot (but occupies two slots) and has 3,328 CUDA cores, 104 Tensor cores for machine learning and 26 RT (RayTracing) cores, a memory bandwidth of 288 GB/s and 6 GB ECC GDDR6 memory. External monitors can be connected via four mini DisplayPorts 1.4.
The Nvidia RTX A2000 should start shipping in October via Nvidia&s distributors as well as integrated into workstations at a price of . However, in view of the global demand, the actual price could be significantly higher than the MSRP. In April, Nvidia had already introduced the top models of the RTX-A series with the RTX A5000 and RTX A6000, which deliver several times the performance of the RTX A2000 (but also cost many times more).
Nvidia RTX-A series, the new A2000 on the lower left.
Compared to the A2000, the next model up, the RTX A4000, has nearly twice as many CUDA/Tensor/RT cores and 16 GB of memory, the A5000 has 2.5x as many (and 24 GB), and the A6000 has 3x as many (as well as 48 GB of RAM).