The RTX A5000 and A4000 are aimed at professionals at slightly more affordable levels. The Turing-based RTX A series will be available on Max-Q-design laptops.
On the server and AI side, Nvidia, which made important announcements this week, has not forgotten professional users. Renamed the Quadro series RTX A, the company announced new mid-range graphics cards.
Introducing the RTX A8000 and RTX A6000 accelerators last year, Nvidia introduced the RTX A5000 and A4000 graphics cards. Ampere-based graphics cards are aimed at professionals.
The RTX A5000 graphics card features an 8192 CUDA processor and 24GB of GDDR6 ECC RAM. With NVLINK, you can connect the graphics card with 240W thermal design power to another A5000 card. This achieves a total bandwidth of 48GB of GDDR6 and 112GB/s.
The RTX A4000 has 6144 CUDA cores and 16GB of GDDR6 ECC RAM memory. The graphics card with 140W thermal design power is not available in SLI style. There are also RTX A2000, A3000, A4000 and A5000 graphics cards based on turing architecture. These cards are Max-Q designed and supported by Nvidia Studio.
There will also be A10 and A16 graphics cards that will work in harmony with grace architecture for servers, as well as mobile Nvidia T1200 and T600 graphics cards focused on multitasking.