Nvidia has officially revealed the full specifications of its RTX 30 series, revealing specifications which are very different to what was previously leaked. With Ampere, Nvidia promises up to a 1.9x efficiency boost over today's Turing products, as well as performance boosts of up to 2x over today's Turing GPUs.

With Ampere, Nvidia's using a customised version of Samsung's 8nm manufacturing process to deliver higher levels of transistor density and increased efficiency. This upgrade is how Nvidia has delivered such a vast increase in CUDA core counts, enabling a true generational leap in graphics performance.

One of the main changes with Ampere is Nvidia's new streaming multiprocessors, an innovation which has allowed Nvidia to deliver a 2x increase in FP32 throughput. This alteration makes Nvidia's Ampere architecture their most efficient yet, offering users over 30 Shader-TPFLOPS of graphics performance.

With Nvidia's second-generation RT cores, Nvidia plans to deliver a 2x increase in RT throughput. At this to Ampere's ability to offer concurrent ray tracing, shading and compute, and it is clear that Ampere will be a force to be reckoned with in future RT workloads. Nvidia's 3rd Generation Tensor cores will also ensure that Nvidia offers fast AI performance when using technologies like DLSS 2.0.

Nvidia has also confirmed that its RTX 3000 series will support HDMI 2.1, which will allow Ampere to support 4K 120FPS framerates on modern TVs. This connection will also support 8K screens when they come to market.

Nvidia has also partnered with Micron to offer the world's fastest gaming memory with GDDR6X, delivering data rates that are close to 1TB/s with their RTX 3090 graphics card.

Below are the specifications of Nvidia's RTX 30 series of products, as well as their UK prices and PSU wattage recommendations.