Compared to the Nvidia Quadro RTX 8000 GPU offering 4608 CUDA cores, the NVIDIA RTX A6000 runs on the latest Ampere architecture and delivers more than twice the CUDA cores at 10,752. This GPU also supports 336 Tensor cores, 112 ROPs, and 84 RT cores. With this new architecture, the Nvidia A6000 can leverage 38.7 TFLOPS of single-precision performance, 75.6 TFLOPS of RT core performance, and 309.7 TFLOPS² of Tensor performance to address the most complex applications. With 28.3 billion transistors on Samsung’s 8nm process chip, the A6000 is up to 2x faster in ray tracing compared to the previous generation.
Under the hood, the RTX A6000 has 48GB of GDDR6 memory running at 2000 MHz with a 384-bit memory interface and error correcting code (ECC). ECC helps meet the uncompromised, accurate, and reliable standards for mission critical applications. Intended for professional workstations, this A6000 GPU operates with the base clock running at 1455 MHz and a boost clock of 1860 MHz. If systems need more power, users can deploy the 3rd generation NVLink bridge to link two A6000 cards together for a total of 96GB of memory.
The dual slot card is built on an 8nm manufacturing process, which supports the latest technological advancements. The Ampere architecture incorporates 2nd generation ray tracing engines to render complex professional models with physically accurate shadows and reflections. The GPU also supports PCIe Gen 4.0 technology for double the bandwidth and faster data-transfer speeds compared to PCIe Gen 3.0. To take on the most demanding 4K or 8K video encoding workloads, the 7th generation NVENC engine provides better encoding quality than software-based x264 encoders.
Designed for digital content creation, the NVIDIA RTX A6000 GPU offers nearly twice the performance of the RTX 8000 GPU. Built on the Ampere architecture, the GPU also supports new RT cores, Tensor cores, PCIe 4.0, and CUDA cores to deliver the best performance for creators and developers.
