The professional visualization graphics cards from NVIDIA’s RTX A6000 series, dubbed “Ada,” are about to go on sale. These cards are aimed at the same market segment as the previous NVIDIA Quadro series: professionals who create 3D content. The 4 nm “AD102” silicon-based RTX A6000 is the market leader (the same one powering the GeForce RTX 4090). Despite operating at slower GPU clock speeds, the A6000 has stronger silicon capabilities than the RTX 4090 because to its stricter 300 W power constraint (compared to 450 W of the RTX 4090).

In comparison to the RTX 4090’s 16,384 CUDA cores across 128 SM, the A6000 “Ada” has 18,176 CUDA cores spread across 142 SM. Additionally, it receives 568 more Tensor cores. Memory has to be the key difference between the A6000 and RTX 4090, with the pro-vis card receiving 48 GB of ECC GDDR6 memory across the chip’s 384-bit memory bus, clocked at 20 Gbps (960 GB/s memory bandwidth); as opposed to the RTX 4090 receiving 24 GB of 21 Gbps GDDR6X (1008 GB/s). Additionally, the card offers six separate accelerated transcoding streams using all three of the NVDEC and NVENC video hardware-accelerators that are physically present on the AD102.

As the card’s usual board power is only 300 W, it has a 16-pin ATX 12VHPWR connector and probably comes with three 8-pin to 16-pin adapters. The reference board has a 2-slot layout, a lateral-flow blower, and a high-density channel-type heatsink with a vapor-chamber plate. No NVLink support is available. The projected board partners for this card are Leadtek, PNY, and ELSA. Leadtek and PNY both published prices, with the PNY card being less expensive at USD $7,377.