AMD has just introduced its new flagship, the Radeon R9 Fury X. And we got a chance to take a closer look at its cooling design and PCB. AMD has just launched its entirely new family of Fury graphics cards based on the groundbreaking Fiji GPU and HBM Technology here at the PCGaming event in E3. We've already talked about all of these graphics cards in detail as well as the dual Fiji, 17 TERAFLOP behemoth that will be coming later in the autumn.

In this particular piece we'll take a closer look at what Fury X looks like from the inside. How AMD managed to design the liquid cooling block to cool all the components and what these components look like on the printed circuit board of the graphics card.

AMD Radeon R9 Fury X Teardown - PCB, Liquid Cooling Pump And Block Pictured

The PCB is 7.5 inches long making it by far the smallest of any high end graphics card. Primarily thanks to the new stacked memory technology which saves 50% of PCB area that would otherwise be occupied by GDDR5 memory chips.

Credit @ Legitreviews.com

The card features a 6 phase power delivery design that's capable of delivering a whopping 400 amps of power. The card is intentionally overbuilt to enable the highest possible overclocks.

Additionally the card's also equipped with a full set of activity LEDs that light up to reflect the GPU load while the card is running.

Cooling the card is a closed loop liquid cooler that's capable of dissipating 500W of heat. On the Fury X, which has a typical board power dissipation of 250-275W the result is an operating temperature of 50c at an unheard of whisper quiet 32dB.

Board size comparison 11.5 inch R9 290X vs 7.5 inch R9 Fury X .....Credit : @HardwareCanucks

In comparison the reference designed Nvidia NVTTM cooler used for the GTX Titan X and 980 Ti maintains temperatures at 84c at a noise level of 40dB. The combination of the over-engineered power delivery system as well as the incredibly capable cooling system make Fury X what AMD described as "an overclocker's dream" graphics card.

The liquid cooling block cools three major components, the Fiji XT GPU, the HBM modules and the regulator via a single loop that's connected to a 120mm radiator. So users should have no fears of VRMs or any other components overheating. Connected to the 120mm radiator is a quiet 120mm high static pressure fan that would exhaust the heat directly outside of the Chassis to keep the rest of the components nice and cool. Housing the card is a di-cast aluminum body with soft black finish and accented with a high polish black nickel exterior exoskeleton.

The Radeon R9 Fury X will be available to buy on June 24th for $649. And that's when official reviews and benchmarks will be out.

However if you're anxious and can't wait until the 24th don't worry. One can very easily estimate the performance of Fury X in a number of in-game benchmarks using a series of simple mathematical formulas.

It’s a very straight forward process. We look at the core count and the clock speed of Fiji and extrapolate its potential performance from those two data points. All structures inside the GPU will scale with the core count, so they’re accounted for. The huge increase in memory bandwidth brought by HBM will definitely have a positive impact on performance. However it will vary based on the workload and the resolution, and thus we’ll have to leave it out for now

The results will be based on TPU’s latest GTX 980 Ti review, click on the drop-down menu below for the math and performance graphs.

[spoiler title="The Math + Performance Graphs"]

#1 We Divide the number of stream processors of Fiji by the number of stream processors of Hawaii, trusting that the performance will scale linearly with each additional SP. Based on the scaling we observed with all other GCN GPUs we found that the performance scales with perfect linearity at 4K.

You can go back back to the R9 290 vs 7870 data as a reference point.

For example the 2560 GCN unit R9 290 was precisely 2.0x faster than the 1280 GCN unit HD 7870. Even though the 290 is clocked slightly below the 7870, so this would account for the architectural improvements that AMD introduced with GCN 1.1 vs GCN 1.0. In this particular exercise around Fiji we do not account for any possible architectural improvements because they're difficult to quantify.

#2 We then multiply the result by Fiji's clock speed, which is purported to be 1050Mhz vs 1000Mhz for Hawaii. Which makes Fiji's clock speed 1.05 that of Hawaii.

Performance = core count x clock speed

= (4096 / 2816) x 1.05

= 1.45 x 1.05

= 1.53

This makes Fiji exactly 53% faster than Hawaii. We then use the data of Hawaii's average performance in the 21 tested games and multiply it by 1.53. Below is the final result laid out neatly in two bar graphs for your viewing pleasure.

[/spoiler]

Or you can check out these leaked Fury X and Fury benchmarks below.