Nvidia has launched the successor to its RTX A2000 GPU for compact workstations and edge computers, giving channel partners a new affordable chip to sell businesses for AI-accelerated and graphics-intensive applications.
Announced Monday, the new Nvidia RTX 2000 Ada Generation GPU packs 16GB of GDDR6 memory—an upgrade from the A2000’s 12GB—and provides a performance boost ranging from 30 percent for 3-D modeling, video editing and virtual-reality design software, to 60 percent for Stable Diffusion’s text-to-image generative AI model, according to the AI chip giant.
Nvidia said the RTX 2000 Ada is available now through global distributors such as Arrow Electronics, Ingram Micro, PNY and TD Synnex, and it will arrive in systems from Dell Technologies, HP Inc. and Lenovo starting in April.
With a suggested price of $625, the RTX 2000 Ada is positioned as a budget-friendly workstation GPU for architects, urban planners, product designers, engineers, content creators who rely on compact form factor computers for demanding applications, according to Nvidia.ADVERTISEMENT
This is half the suggested price of the next workstation GPU on the performance spectrum, the 20GB RTX 4000, and nearly 10 times less expensive than Nvidia’s most powerful workstation GPU, the 48GB RTX 6000.
Nvidia said the RTX 2000 Ada’s dual-slot, low-profile form factor also makes it a good fit for edge computing, whether it’s real-time data processing for medical devices, predictive maintenance for manufacturing systems or AI-driven analytics for retail stores.
As the name suggests, the RTX 2000 Ada uses the same Ada Lovelace architecture as other workstation and gaming GPUs, like the RTX 6000 and GeForce RTX 4090, released since 2022.
Compared to the RTX A2000 from 2021, RTX 2000 Ada features third-generation RT Cores that provide up to 70 percent faster ray-tracing performance, fourth-generation Tensor Cores that offer up to 80 percent faster AI throughput, CUDA cores with an up to 50 percent boost in 32-bit floating-point (FP32) throughput and up to double the power efficiency, according to Nvidia.
The RTX A2000 also takes advantage of Nvidia’s DLSS 3 AI-powered image upscaling software and comes with an AV1 video encoder that it said is 40 percent more efficient than H.264.
Nvidia said its latest RTX Enterprise Driver supports the RTX 2000 Ada, among other GPUs, and comes with new features, such as the AI-based VideoHDR, which “expands the color range and brightness levels when viewing content in Chrome or Edge Browsers.”
Also new to the driver are TensorRT-LLM, an open-source library that optimizes inference performance for large language models on Nvidia GPUs; an API to offload workloads from the CPU to the GPU; and video quality and coding efficiency improvements to video codecs.