NVIDIA has expanded its range of professional video cards with the new compact model RTX 2000 ADA. It is the seventh and youngest card in a series of specialized graphics accelerators based on the Ada Lovelace architecture. Similar to the RTX 4000 SFF model on the same architecture, the new product has a low-profile design and is intended for use in compact PCs (SFF).
The RTX 2000 ADA also has in common with the previous RTX 4000 SFF its low power consumption of only 70 W. The video card does not have an additional power connector. Unlike the RTX 4000 SFF, which is based on the AD104 graphics chip with 6144 CUDA cores, the RTX 2000 ADA model received the AD107 GPU with 2816 CUDA cores. The chip also includes 88 fourth-generation tensor cores and 22 third-generation RT cores. In terms of characteristics, the new product turns out to be slightly weaker than the gaming GeForce RTX 4060.
NVIDIA RTX 2000 ADA is the direct successor to the RTX 2000 Ampere, which was equipped with 12 GB of memory. In turn, the new video card received 16 GB of memory, and this, by the way, is more than what the rumors attributed to the widow. At the same time, the RTX 2000 ADA supports a 128-bit memory bus and has a bandwidth of up to 224 GB/s. The card uses the PCIe 4.0 interface, but the manufacturer has allocated only 8 lines of this interface.
The set of external RTX 2000 ADA connectors includes four Mini-DisplayPort 1.4a, so those who need more modern DisplayPort 2.0 or 2.1 as part of a specialized GPU of the same level will have to look towards solutions from AMD and Intel.
NVIDIA has priced the RTX 2000 ADA at $625. It will be available from the company’s partners, including Arrow Electronics, Ingram Micro, Leadtek, PNY, Ryoyo Electro, TD SYNNEX, Dell, HP and Lenovo from April this year.