What a few months it’s been for AMD. Eternal hope arose after AMD began dropping details about its latest GPU architecture update, dubbed Fiji, which includes an all-new, much-hyped technology called High Bandwidth Memory (HBM). After years of playing second fiddle to Nvidia in terms of raw performance and power consumption, HBM’s promised power savings and massive bandwidth would be AMD’s saving grace, and a turning point for a company desperately needing a win after a streak of multiple lose million dollars.
Unfortunately for AMD, the GPU that emerged, the R9 Fury X, wasn’t quite worth it. Nvidia’s preemptive strike with the GTX 980 Ti, no doubt propelled by the imminent launch of the Fury X, put a damper on AMD’s groundbreaking technology. The Fury X cost the same as a 980 Ti, but it wasn’t faster, had less memory, and consumed more power, even though its excellent water-cooling system kept temperatures in check. Team Red, it seemed, was in trouble again.
While the flagship R9 Fury X made headlines, the release of its little brother – the R9 Fury – a few weeks later was much more interesting. At £450 ($560), the R9 Fury is priced below the £550 ($650) of the Fury X and the GTX 980 Ti, but above the £400 ($520) of the GTX 980 and the £350 ($430 ) of the R9 390X. That puts it, and AMD, in a slightly odd position: if you can get the Fury (and that tasty HBM) for £450, is there any point in switching to the Fury X? And then again, if the Fury is only slightly better than a 980 or 390X, why spend the extra money?
|Specifications at a glance||R9 Anger X||R9 rage||R9 390X||R9 290X|
|Raise the clock||1050MHz||1000MHz||1050MHz||1000MHz|
|Memory bus width||4096 bits||4096 bits||512-bit||512-bit|
|Memory bandwidth||512 GB/sec||512 GB/sec||384 GB/sec||320GB/s|
|Memory size||4GB HBM||4GB HBM||8GB DDR5||4GB GDDR5|
|Typical board power||275W||275W||250W||250W|
Like the Fury X, the Fury is based on AMD’s Fiji architecture, which combines a massive 8.9 billion transistor, 596 square millimeter chip built on a 28nm process with 4GB of stacked HBM. We’ve covered HBM extensively before, but to recap, HBM uses stacked memory chips along with a silicon interposer and through-silicon vias (a connection that runs from top to bottom through the chip) to move the DRAM closer to the GPU. This shortens the tracks, allowing for greater bandwidth and lower power consumption. As an added bonus, the circuit board itself is also much smaller.
However, as this is a first-generation design, there are a few compromises. For starters, the Fury (as well as its bigger brother the Fury X) is limited to 4 GB of memory. AMD might say 4GB is more than enough even for 4K gaming, but tellingly the significantly cheaper 390X comes with 8GB of GDDR5 memory, while Nvidia’s 980 Ti comes with 6GB.
Perhaps the 390X’s 8GB of VRAM is just AMD’s way of covering up that, other than the RAM, the card is virtually identical to the 290X (which had the same GPU but only 4GB of RAM). At least while it is currently While it’s hard to completely fill 4GB, that smaller amount of memory is likely to become a problem before it’s beneficial to upgrade.
To be fair to AMD, Nvidia’s GTX 980 will also face that problem, even though that card has the advantage of being older and a tad cheaper. On paper, however, the Fury should be significantly faster than the 980. Compared to the Fury X, the Fury’s Fiji chip lowers the compute units from 64 to 56, reducing the total number of stream processors from 4096 to 3584. The clock speed also drops slightly to 1000 MHz, a five percent reduction from the Fury X. That’s all there is to cuts, keeping the number of geometry units and ROPs unchanged. That should put the Fury’s performance right between that of the Fury X and 980 Ti and the 980 and 390X.
In our own testing of an Asus Strix version of the R9 Fury (some of which are shown below), as well as a quick look at other benchmarks reported by other sites, the Fury is indeed faster than a 390X and 980, but slower then a Fury X and 980 Ti. But that does not tell the whole story.
Given the Fury’s odd price positioning, that’s key How much faster or slower than its competitors. In Anandtech’s comparison tests against the Fury X (unfortunately, Ars doesn’t have access to a Fury X to run his own tests), they saw an average performance difference of about seven percent in favor of the Fury X. Since there are about 20 percent difference in price between the two cards (14 percent for the US ones), you have to wonder if the Fury X is really worth it.
Flagship cards, of course, rarely offer value for money. Nvidia does a little better with the GTX 980 Ti, which costs about 30 percent more than a 980 for a performance boost of about 30 percent. But it’s also worth comparing to the 970, which offers similar performance to the 980, with the 980 Ti costing a whopping 70 percent more.