NVIDIA released new sets of GPUs for PC workstations for desktop and several for professional laptops.
Nvidia RTX for desktop workstations
The NVIDIA RTX A5500 for desktops is based on GA102 GPU running 10,240 CUDA cores—same configuration as the RTX 3080 Ti. However, the said GPU utilizes Error Correction Code (ECC) memory with a whopping 24GB ECC GDDR6 VRAM. ECC can detect and correct errors itself which is significant for professional tasks such as 3D modelling, machine learning, simulations, and the likes.
The said VRAM can be expanded up to 48GB via NVLink to connect a second GPU. It promises doubled performance over its predecessor and can be up to 12 times as fast for machine learning.
Two Nvidia RTX A5500 connected via NVLink
Furthermore, RTX A5500 has a laptop equivalent based on GA103 chip with 7,424 CUDA cores (parallel with the laptop RTX 3080 Ti), which is nearly less than 30% from its desktop counterpart. This mobile GPU comes with 16 GB of GDDR6 VRAM non-ECC.
Other NVIDIA offerings include RTX A500 having the smallest Ampere based GPU for laptops with 2,048 CUDA cores and 4 GB of VRAM. The RTX A1000, A2000, and A3000 with 4 GB, 8 GB and 12 GB of VRAM respectively.
Lastly, the mid-ranger RTX 4500 based on GA104 with 5,888 CUDA cores (resembling the RTX 3070) and 16GB of VRAM.
Pricing details are yet to be announced but these new sets of GPUs should be available prebuilt on machines in the second quarter of 2022.
Watch the videos below showcasing the power of the desktop and laptop version of NVIDIA RTX A5500.
