Introduction
ASUS is one of the biggest names in the enthusiast technology market; with a product range including motherboards, AIO PC’s, and even internet routers, they know a thing or two about what makes a good product. They produce some of the best graphics cards on the planet with such innovative designs and production processes, it can make other companies shake.
Today we have in an ASUS STRIX R9 380 2GB. ASUS has introduced an entirely new production method with the R9 300 series, called Auto-Extreme. This new process introduces a highly automated production line, increasing longevity by removing all possible human error and with the eradication of flux; this increases the quality of joints and components. The R9 380 also brings new technology and features to the table, with the full support of DX12, Free sync technology, and passive cooling; this is set to be a great enthusiast graphics card. Why not take a look at the Auto-Extreme process here:
https://youtu.be/4gRpuurPsuc
The R9 380 range, in general, is based on the slightly ageing R9 285 graphics card PCB, this can only lead you to think why AMD want to push older technologies; is it to push more resources into new technology? We’ll find out today in this review.
The box follows a similar design to the rest of the newly introduced STRIX range. A striking outer sleeve with the key information and the essential documents to get you up and running. 
The cooling shroud follows a similar design to that of the previous STRIX cooler designs, this model also runs passively under low load circumstances.
Something ASUS always pleases me with is the backplate. Even on the lower models such as this, you get a stunning backplate that not only strengthens the card, but provides cooling support for the rear mounted components. Here we can see the behemoth 10mm cooling pipe, iconic of the DIRECT CU coolers from ASUS.
A close up on the power connection shows that this card gets all of its power from a single 8-pin PCI cable, great news for those looking for an upgrade with a lower wattage power supply.
At the business end of the card, we see the usual connections of 2x DVI, 1x HDMI and 1x DisplayPort.
Something’s not right here… how can a chip which is essentially the same as the one in the 285 (although clocked a bit higher) deliver over twice the average framerate at 1080p and 1440p in Metro Last Light? I thought the VRAM was the problem while reading the review of the sapphire 380, but here the VRAM is 2GB – same as the 285… the only difference is about a 100MHz higher core clock. How can that result in such a large difference in performance?
Guess these cards are more of a refresh than a rebadge then, I was thinking this when I saw my first 390x review yesterday, which seems to gives pretty impressive results similar to a 980, from what is apparently an old 290x core.
Maybe When everyone was reporting about this rebadge being scandalous etc they were wrong? I guess that’s also why AMD got fed up and pulled loads of reviewer’s samples? Who knows, but so far all these “rebrands” have been pretty surprising and I was expecting them to suck – colour me surprised
my thought exactly !
I was blocked from kitguru for posting that video on vimeo haha
my video was removed too 😛
Whatever useless website.
😛
In this case, it’s more likely that there was something wrong with the tests on that game, because in other games the difference is small, and in other reviews of the card the difference between it and 285 on Metro Last Light is a couple of frames per second, nothing nearly as bad as this. I guess it could be a driver problem or something…
It’s really weird seeing this good results. I have much lower both averages and minimums on my 280X in BF4. Minimum can go even to 35ish, but only on death cam, and average is about 55-60 on big maps.
If you remember, this is all Hawaii architecture, since HD7970 everything is almost 100% the same chip. All R7 and R9 200 and 300 series, and now this fps results, like what the hell? 🙂
But, it’s good to see that it’s not just a name changer, cause, that would be silly of AMD.
Even Fiji is not a new architecture, it’s Hawaii with “tweaks” -> GCN 3.0 to work with HBM.