• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

*** Proper 390X review & discussion thread ***

Soldato
Joined
22 Aug 2008
Posts
8,338
HardOCP

Link: http://www.hardocp.com/article/2015/06/18/msi_r9_390x_gaming_8g_video_card_review

IxxSiPP.gif

Guru3D

Link: http://www.guru3d.com/articles-pages/msi-radeon-r9-390x-gaming-8g-oc-review,1.html

9ICmyqy.jpg

Anandtech [no benchmarks]

Link: http://www.anandtech.com/show/9387/amd-radeon-300-series

AMD had told us that there are a number of small changes from the 290 series to the 390 series that should improve performance by several percent on a clock-for-clock, apples-to-apples basis. That means along with the 20% memory clockspeed increase and 5% GPU clockspeed increase, we should see further performance improvements from these lower-level changes, which is also why we can’t just overclock a 290X and call it a 390X.

The first optimization is that AMD has gone back and refined their process for identifying the operating voltages of Hawaii chips, with the net outcome being that Hawaii voltages should be down a hair, reducing power and/or thermal throttling. The second optimization mentioned is that the 4Gb GDDR5 chips being used offer better timings than the 2Gb chips, which can depending on the timings improve various aspects of memory performance. Most likely AMD has reinvested these timing gains into improving the memory clockspeeds, but until we get our hands on a 390X card we won’t know for sure.

Techspot [HIS MUSTARD COOLER]

Link: http://www.techspot.com/review/1019-radeon-r9-390x-390-380/

We'd like to thank HIS for supplying us with their Radeon R9 IceQ X² graphics cards. The company's IceQ X² cooler performed exceptionally on all three cards and they look great, too.

Hardware Canucks

Link: http://www.hardwarecanucks.com/foru...69646-amd-r9-390x-8gb-performance-review.html

Regardless of whether you want to call this a rebrand or refresh (I’m firmly on the refresh side), the R9 390X is an undeniably appealing card for anyone who can’t justify spending over $450 for a GPU. It is truly amazing to see that a Hawaii-based derivative can be so competitive this far into its life. I’m just not sure if that represents a ringing endorsement for the versatility of AMD’s GCN 1.1 architecture or an honest critique about how the graphics performance yardsticks haven’t moved all that much in almost two years. Maybe it’s both.

ETeknix

Link: http://www.eteknix.com/sapphire-tri-x-r9-390x-8gb-graphics-card-review/

fw9vHKY.png
 
Last edited:
If you see any more post 'em here and I will add to OP.

So far we have a pretty good collection of info, Anands has some details on Grenada, techspot covered the HIS cooler which was popular last gen, and the benchies are there for everyone to pore over.
 
But, allegedly, you can mod the new 300 series drivers to work on 200 series cards, and get a boost in performance... The cards themselves are nothing new at all.

That's an unsubstantiated rumour afaik.

Anandtech talked to AMD themselves and they report the 390X is tweaked in some way.

Last but certainly not least however, we want to talk a bit more about the performance optimizations AMD has been working on for the 390 series. While we’re still tracking down more details on just what changes AMD has made, AMD had told us that there are a number of small changes from the 290 series to the 390 series that should improve performance by several percent on a clock-for-clock, apples-to-apples basis. That means along with the 20% memory clockspeed increase and 5% GPU clockspeed increase, we should see further performance improvements from these lower-level changes, which is also why we can’t just overclock a 290X and call it a 390X.

So what are those changes? From our discussions with AMD, we have been told that the clock-for-clock performance gains comes from a multitude of small factors, things the company has learned from and been able to optimize for over the last 2 years. AMD did not name all of those factors, but there were a couple of optimizations in particular that were pointed out.

The first optimization is that AMD has gone back and refined their process for identifying the operating voltages of Hawaii chips, with the net outcome being that Hawaii voltages should be down a hair, reducing power and/or thermal throttling. The second optimization mentioned is that the 4Gb GDDR5 chips being used offer better timings than the 2Gb chips, which can depending on the timings improve various aspects of memory performance. Most likely AMD has reinvested these timing gains into improving the memory clockspeeds, but until we get our hands on a 390X card we won’t know for sure.
 
iv been trying to find a chart i read earlier, it shows the teraflops difference between 300 & 200 series, but cant find it anymore QQ, i need to learn to bookmark links...i will post the link if i stumble on it again

http://videocardz.com/56676/amd-officially-introduces-radeon-300-caribbean-islands-series

The AMD Caribbean Islands Family: Not Just a Rebrand (..)

AMD has been hard at work over the past year-and-a-half optimizing and re-architecting the microcontrollers within the ASICs themselves. Combined with the improvements to their manufacturing process, AMD has been able to squeeze more performance out of each of their cards and increase performance while maintaining the same price tier as its predecessor.

  • The R9 390 and R9 390X replace the R9 290 and R9 290X and are both 300 GFLOPS faster than their predecessors (5,100 GFLOPS and 5,900 GFLOPS respectively) without increasing power in typical workloads.
  • The R9 380 also benefits from the maturing of the 28nm process technology and AMD’s optimizations. It gains roughly 200 GFLOPS in performance: from 3,290 GFLOPS to 3,480 GFLOPS in compute performance.
  • The R7 370’s compute capability of 2,000 GFLOPS is also faster by 200 GFLOPS than its predecessor’s (R7 265) 1,800 GFLOPS.
  • The R7 360 has a compute performance of 1,610 GFLOPS, slightly more than the 1,536 GFLOPS of the R7 260.

In all cases, AMD increased performance and also added many features that previous generations did not have. Some of those features are enabled through the driver and others are done in hardware. But all of the GPUs listed above will support DirectX 12, Vulkan, and Mantle graphics APIs.
 
The hardocp review also point to greatly improved tessellation performance on the witcher "apples apples" paragraph

http://www.hardocp.com/image.html?image=MTQzNDYxMjU0OWwxR0JRekpFNXFfM180X2wuZ2lm

We ran two separate tests here because we couldn't believe how much faster the MSI R9 390X was over the R9 290X, but it was that much faster. It comes much closer to GTX 980 performance in this game, and that is with HairWorks enabled in both tests above. There just might be something to that improvement in tessellation performance noted in the introduction.

The sentence in question:

AMD Radeon 300 Series Graphics fully support Microsoft DirectX 12, with the following enhancements over earlier products, Faster Tessellation, Tiled Resources – Support for massive virtual textures, enabling dynamic loading of tiles into graphics RAM for expansive game world details

We need people to do more specific testing.
 
Yes but a lot of that % increase is due to clock bumps... it's great that they've been able to improve power consumption, and it's great that they've enabled some stuff in the drivers that wasn't enabled (and won't be?!) for 200 series cards (tho it could be??).

And given how other sites are reporting that the silicon is exactly the same, why should anyone trust Videocardz saying that the hardware is different? Not exactly the go to site for trusted information, is it?

Videocardz isn't the one saying it. It's just a link where you can read it.

Why are you trying so hard to crap on it? DP already got banned for doing what you're doing.
 
Some of us prefer to work it out for ourselves rather than being told. :)

We have heard tell of improvements, we have a vector for our investigation and we are going to pursue it. :)

The argument over "rebrand vs refresh" doesn't interest me as much as the tech itself.
 
And why are you so obsessed with getting people banned. It is pathetic that if you don't like what someone says you start trying to infer they will be banned. Just grow up.

One comment is an obsession? :confused:

I would prefer that people stick to the topic, do you want to talk about the 300 series?
 
If you're going to have 'proper' reviews and maintain this separate thread, you can remove Tom's.

They're one of the sites astroturfing for NVIDIA. They used the older, non-recommended drivers (15.5) rather than 15.15 (which shows a large performance gain for all AMD cards and particularly the 390X and are the 3xx / Fury launch drivers).

Also, remove any other reviews using 15.5 for 3xx please.

Toms is removed for now, if they update with newer drivers let me know and I'll put it back up. Thanks for pointing it out. :)
 
We have to remember with catalyst drivers your able to control the amount of tessellation that being applied in games. Moving the slider from 64x to 16x apprantly results in discernable drop in visual quality but gives a dramatic boost in performance.

I thought quality only suffered going below 8x? As in no difference above 8x.

If AMD have simply upped core speeds surely the 290x 8gb should be seeing the same boosts with the newer drivers...?

Apparently they are. They're saying you can hack the pre-release ones to work on 290X.

I'm running a Titan X atm or I would check myself.

edit: Thanks to Greebo for the link

http://forums.guru3d.com/showthread.php?t=399956
 
Last edited:
In the Moor report they stated AMD had redone the microcontrollers in the Hawaii silicon itself. Aren't the tesselator, scaler etc integrated within the die?

So perhaps tess has been improved in this way?
 
In the Moor report they stated AMD had redone the microcontrollers in the Hawaii silicon itself. Aren't the tesselator, scaler etc integrated within the die?

So perhaps tess has been improved in this way?

From reading some reddit comments it seems they might simply be binning 290X cores that have tesselator logic blocks that can be OC'd, which are then clocked accordingly with the new Grenada BIOS. That might be the true meaning of Grenada. This might also allow them to offer higher VSR options. Rather clever as it allows them to avoid a respin or even a stepping change and thus the costs associated.

Thoughts?
 
Last edited:
Back
Top Bottom