Introduction

Gainward GeForce RTX 5080 Phoenix GS is the company’s value custom-design take on the NVIDIA GeForce RTX 5080. The Phoenix line of graphics cards strike a balance, featuring a factory overclock, low noise, and are generally priced closer to the NVIDIA MSRP, while the company’s Phantom line of graphics cards take the upper crust in product design and performance. The Phoenix series is best for those looking for an RTX 5080 that they can install and get gaming. There are still plenty of premium touches, such as a vapor chamber-based cooling solution, dual-BIOS, double ball-bearing fans, and a touch of RGB LED lighting. The card also offers a fairly high factory overclock that can level up to some premium custom designs from other brands.

The GeForce RTX 5080 is designed to serve the same gaming use-case as the RTX 5090, which we reviewed last week. It enables you to enjoy today’s AAA games at 4K Ultra HD resolution with maximum settings and ray tracing enabled. While the RTX 5090 caters to broader markets, including AI development and compute thanks to its lavish 32 GB memory over a 512-bit memory bus, the RTX 5080 is squarely aimed at gamers seeking top-tier performance. Its starting price is precisely half that of the RTX 5090, and when you examine its specifications, you’ll notice that many figures are either approximately or exactly half of what appears on the RTX 5090 spec sheet.

The RTX 5080 is based on the GB203 silicon, which is NVIDIA’s second-largest chip to implement the GeForce Blackwell graphics architecture. This lean piece of silicon is built on the exact same NVIDIA 4N foundry node at TSMC, which is derived from its 5 nm EUV process. The RTX 5080 maxes out the GB203 enabling all 84 streaming multiprocessors (SM) present on the silicon. This yields 10,752 CUDA cores, 336 Tensor cores, 84 RT cores, 336 TMUs, and 112 ROPs. It also features all 64 MB of the on-die L2 cache, and two sets of NVDEC and NVENC video accelerators that support 4:2:2 formats for AV1 and HEVC. The chip features a 256-bit wide GDDR7 memory bus, which drives 16 GB of memory. NVIDIA is running this memory at 30 Gbps on the RTX 5080, yielding 960 GB/s of bandwidth—a 34% increase over that of the RTX 4080.

The GeForce Blackwell graphics architecture debuts a revolutionary new concept called Neural Rendering. The easiest way to describe this would be to use a generative AI model to create objects, materials, and textures that are then combined with a conventionally rendered raster 3D scene, just like RTX combines ray traced objects to it. This is exclusive to Blackwell because the chip has the capability to accelerate large AI models and render 3D graphics concurrently, with fine-grained performance control, thanks to a new on-die hardware scheduler called AI management processor (AMP). The new Blackwell SM has all the silicon-level groundwork for neural rendering, and NVIDIA even worked with Microsoft to standardize this at an API level, letting 3D applications directly address Tensor cores. The new 4th Gen RT core comes with hardware acceleration for Mega Geometry, the ability to give ray traced objects exponentially higher poly counts.

The Gainward RTX 5080 Phoenix GS comes with a fairly hefty aluminium fin-stack cooling solution which uses a trio of large fans to push air through the heatsink. Airflow from the third fan goes through the heatsink and out the back of the card through a large cutout in the backplate. The new Cyclone Series fan from Gainward features wedged fan impeller blades that work to increase airflow. Gainward has given the RTX 5080 Phoenix GS a factory overclock of 2700 MHz boost (compared to 2617 MHz reference), while leaving the memory speed untouched. We couldn’t get a price point from Gainward, we’re expecting that the card will hit $1150.