Radeon R9 FURY X 

The performance gap suddenly widens as soon as we start looking at lower resolutions; 4K vs. the GTX 980 Ti: -2%, 1440p: -9%, 1080p: -14%, 900p.

Real performance losses only become apparent in x8 1.1 and x4 2.0, where the performance drop becomes noticeable with around 6-10.

This results in a 4% increase in clocks or a slim 2% increase in overall performance at 1440p when compared to the ASUS Strix, which I've been using as baseline throughout this review because it runs at reference design clocks.

NVIDIA's GTX 980 Ti and Titan X reference designs were far from quiet, and the Radeon Fury X is much quieter during gaming and runs extremely cool, reaching only 60°C in games.

Well, Furmark will run faster, but your games won't.

Sure, there will be lazy game developers that will fill all the VRAM up with unnecessary junk, but those games will be the exception, not the norm.

When looking at individual game results, the effects of constrained PCIe bandwidth vary wildly.

This means you can build a compact low-noise gaming PC for the living room that has enough power to play all the latest games.

In gaming, noise levels are very good, much lower than competing reference designs from NVIDIA.

For any memory standard (including HBM), memory speed/bandwidth and power draw don't scale linearly.

It is complemented by a transfer of video memory from physical chips surrounding the GPU package - over 32-bit-wide data paths, each.

This is what necessitates a new memory standard.

On the following page, we will investigate what gains can be achieved from voltage adjustments to the Radeon Fury, how these overclocking results translate into real performance, and what this means for power consumption.

This configuration is specified for up to 375 W power draw.

Furmark maximum power is definitely a good way up from what we've seen on ASUS' card, but it unfortunately makes no difference in out-of-the-box performance because of the way AMD's Boost technology works.