10+ In stock
Was: £359.99 £299.99
10+ In stock
10+ In stock
Radeon RX VEGA 64
AMD Radeon RX Vega 64 graphics cards represent the flagship of gaming graphics cards from the Texas-based chip manufacturer and are based on AMD's "Vega" GPU architecture. Coming with 64 Next-gen Compute Units (NCUs) and with a total of 4.096 Shaders, Radeon RX Vega 64 graphics cards offer a massive amount of pure processing power. RX Vega 64 also comes with 8 GB of second generation High Bandwidth Memory (HBM2) and a 2048-bit memory interface. AMD Radeon RX Vega 64 graphics cards have been designed from the ground up for 4K/UHD resolution gaming as well as virtual reality.
AMD has prepared to launch three variants in the Radeon RX Vega 64 series:
- AMD Radeon RX Vega 64
- AMD Radeon RX Vega 64 Limited Edition
- AMD Radeon RX Vega 64 Liquid Cooled Edition
Apart from a standard version with a black plastic cooler shroud, AMD is also offering a "Limited Edition" that comes with an anodised aluminium shroud. Another version with an aluminium cover is also available, however this comes with integrated All-in-One water-cooling to enable higher stock clock frequencies.
Watercooled AMD Radeon RX Vega 64 Graphics Cards with 2 Year Guarantee
The Overclockers UK Tech Labs Team naturally has watercooled AMD Radeon RX Vega 64 graphics cards available. To this end the Tech Labs Team is offering an RX Vega graphics card of your choice with a monoblock and (optionally) a backplate. The advantage of a watercooled Radeon RX Vega 64 from the Tech Labs Service is the 2 year Tech Labs guarantee that replaces the manufacturer guarantee. Watercooling offers optimal cooling, thereby allowing the graphics card to reach its full potential. Gaming enthusiasts in particular will appreciate graphics cards with watercooling, and watercooled Radeon RX Vega 64 graphics cards will, as a rule, offer better overclocking than traditional air-cooled graphics cards.
AMD's Radeon Vega Graphics Architecture
The Vega graphics architecture not only contains a variety of optimisations compared to the Graphics Core Next architecture (GCN) of previous generations, it also introduces entirely new components. The new Vega GPU architecture is based on four new technologies:
- The High Bandwidth Cache Controller (HBCC) enables a markedly more effective use of HBM (High Bandwidth Memory) in comparison to Fiji GPUs, which were the first to use this type of memory. The HBC Controller allows highly efficient control of the memory and increases the minimum Frames per Second (FPS) in many games by up to 100% and the maximum FPS by up to 50%. This means the 8 GB HBM2-VRAM on the AMD Radeon RX Vega 64 graphics cards performs to a similar level as 16 GB of GDDR5!
- Rapid Packed Math (RPM) is a feature of the Next-Gen Compute Unit (NCU) and exploits the extremely flexible core structure of the new Compute Units. RPM also accelerates FP16 calculations for example, thereby dramatically increasing the processing speed of scientific calculations. Activated RPM doubles the number of calculations carried out per second without requiring any additional resources. Physics-based calculations in particular benefit from Rapid Packed Math, and certain scenarios in games e.g. large explosions utilise this.
- The new programmable Geometry Pipeline of the Vega Architecture allows the efficient load balancing of many millions of polygons simultaneously. The AMD Radeon RX Vega 64 graphics cards offer geometry calculations with 11 polygons (when using 4 Geometry Units) per clock and therefore twice as many as the Radeon R9 Fury X with its 4 polygons per clock with 4 Geometry Units. Modelling applications in particular benefit from this and this allows large 3D models to be rendered in near real-time.
- The term Enhanced Pixel Engine also refers to the optimised structure of the Shader Units which massively improves performance at high resolutions (WQHD, UHD) and in virtual reality. The Enhanced Pixel Engine relates to a new rasterizer technology that optimises cache localisation and performs rendering processes more efficiently.