Review: PowerColor Radeon RX 480 RED DEVIL edition

Published by

Click here to post a comment for Review: PowerColor Radeon RX 480 RED DEVIL edition on our message forum
#5313179
https://forums.guru3d.com/data/avatars/m/254/254338.jpg
Thanks for adding Vulkan results. The Vulkan results are very impressive for this card, moving the card up from just over GTX970 to exceeding GTX980! Not bad for a mid range card. Hopefully we'll see more developers picking up Vulkan over DX12 going forwards so that users who don't want to use Windows 10 can benefit from a next gen API.
#5313197
https://forums.guru3d.com/data/avatars/m/115/115710.jpg
I'm surprised how "hot" it runs... Even though it's quiet but still. I would have expected a cooler like that to keep it cooler. Also this pretty much confirms that these cards don't overclock that well, only small gains.
#5313202
https://forums.guru3d.com/data/avatars/m/16/16662.jpg
Administrator
Like 2 sentences before you stated that it costs 269 USD. I guess that quoted sentence is from the reference card review ?
No that price is correct as the MSR sales price is: Suggested retail price of 269USD with availability on the 29th
#5313203
https://forums.guru3d.com/data/avatars/m/248/248627.jpg
I expected a much higher overclock but still a great card I would take this over 1060 Great review as always
#5313206
https://forums.guru3d.com/data/avatars/m/180/180081.jpg
AIB 1060s are faster, cooler, more efficient and have more overclocking headroom at the same or even lower prices. I don't think this card is a wise choice at this price.
But a 1060 doesn't feel as futureproof due to the non-implementation of asynchronous compute.
#5313217
https://forums.guru3d.com/data/avatars/m/260/260114.jpg
He He Sooner than Later, thats Great. Fiji is better ;-)
#5313220
https://forums.guru3d.com/data/avatars/m/164/164033.jpg
you are wrong. Pascal supports full DX 12.1 features including asynchronous compute.
It does not really. It does support something similar tho. Even when comparing 12.0 features nvidia and amd have differences, asynchronous compute is not part of that feature set tho. Best support is to be had from Intel and skylake igpus. Vulkan results are tad lower then they would be with async but it's a really nice bump already with only dx12 🙂 Nice review as always.
#5313221
https://forums.guru3d.com/data/avatars/m/16/16662.jpg
Administrator
Don't say nvidia doesnt have async enabled... who cares, we want to see max performance of this card. This got me pissed off and wondering who pays your bills. Be impartial.
Doom Vulkan was tested at SMAA, it's going to stay that way. Do you really think I have 48 hours per day time or something just to meet any and every settings ? You wonder who pays my bills ... I do and work roughly 70+ hours a week to do so. I cannot remember an instance in the past ten years that Nvidia paid us anything whatsoever. You state 'be impartial', yet it seems I just stumbled into the biggest fanboy with a massive AMD BIAS. Why I test with HBAO+ ? It's technology supported by both camps. The fact that is might work our a tiny bit less for AMD is reason to disable it ? Now that's bias. Should I go ahead and disable some AMD preferred settings as well ? Tomb Raider will be retested once when I have the time, the new driver was released hours ago. No, Timespy does not allow a generic score when you disable async. Spend 2 minutes and and 5 bucks to try for yourself ? Seriously grow the hell up, don't throw around accusations like that as your message here is colored completely red by modeling settings that work out the best for AMD. Our settings and tests are balanced properly and evenly in-between both AMD and Nvidia. You might not like it, but that's how it is.
#5313222
data/avatar/default/avatar32.webp
Was Vulkan in Doom tested with SMAA also? Because if it was... Test Doom with AA completely off or on TSSAA. And don't be so hesitant to test Vulkan. Doom may have come out with OpenGL but it is a Vulkan game and not OpenGL. Since I installed the game it was set to Vulkan and not OpenGL. Don't say nvidia doesnt have async enabled... who cares, we want to see max performance of this card. This got me pissed off and wondering who pays your bills. Be impartial. WHY in the world would you test Tomb raider with nvidias HBAO+ is another question. its a recipe for lower performance on AMD. please Retest Tomb Raider with newer drivers from AMD that promise 10 percent performance increase. Timespy, why are there only graphics score async on/off results? cant you show overall test results with async on and off?
wow... just wow can anyone think of any other brand drawing as much lunatics as AMD ban the mofo please
#5313232
https://forums.guru3d.com/data/avatars/m/16/16662.jpg
Administrator
Yes but in the quote I posted from the review there is 239$. That's what I was trying to point out.
Oh crap I misread your post, sorry let me check that out.
#5313233
https://forums.guru3d.com/data/avatars/m/16/16662.jpg
Administrator
HH is wattmann so bad for OC.. can you use msi afterburner ?
I am OCing an MSI RX480 at the moment with an internal beta of AfterBurner. Which already works nicely yet needs more refining. However it could take weeks before we can publish an advanced build to the public as there isn't only Polaris support, we're (well Alex is ) working on DirectX 12 overlay compatibility support as well.
#5313244
data/avatar/default/avatar15.webp
But a 1060 doesn't feel as futureproof due to the non-implementation of asynchronous compute.
That's relative ... lacking CR and ROV's is probably less futureproof than AC since they (CR, ROV's) are DX12_1 requirements.
#5313252
https://forums.guru3d.com/data/avatars/m/16/16662.jpg
Administrator
You are awesome HH, tnx as usual. If I may ask just one more thing does the msi 480 AIB go over 1400Mhz with afterburner?.. Tnx
All AIB cards seem to halt at roughly 1360~1375 MHz, after that number they become more unstable. I did reach 1400 on the MSI card, but it just wasn't stable long term. AB is not going to change the overall result, other then the tweaking process is way more handy.
#5313259
https://forums.guru3d.com/data/avatars/m/16/16662.jpg
Administrator
Is that with stock voltages, I am just asking because a user over at OCUK managed to get his Nitro + 1400 mhz with stock voltages and 1450 i believe with voltage tweaks..
+100 mV on the GPU stock fan settings though. But it depends highly on ASIC quality ... have a peek here.
#5313265
https://forums.guru3d.com/data/avatars/m/164/164033.jpg
Tnx again it clears it up.. I am getting the devil for my GF... HH as usual you are awesome 🙂
1.3v might get those 1400+ but then again the amount of power the gpu would use. Uhhh. I ordered nitro+ for my partner 😀
#5313267
https://forums.guru3d.com/data/avatars/m/263/263841.jpg
Nice review Hilbert! Out of curiosity how does the devilish award stack up against the recommended award? 😀
#5313293
https://forums.guru3d.com/data/avatars/m/209/209146.jpg
http://www.guru3d.com/articles-pages/powercolor-radeon-rx-480-red-devil-review,10.html Going by those results then this third-party 480 would with the 16.7.3 drivers actually perform above a Fury? Not bad if the changelog is accurate with the performance gains in that game. 🙂 (EDIT: Well it generally never is I suppose, it's always "up to" I believe. 😛 ) http://forums.guru3d.com/showthread.php?t=409004
Performance Improvements: Rise of the Tomb Raider™ performance increase up to 10% versus Radeon Software Crimson Edition 16.7.2 on Radeon™ RX 480 graphics
A gain of ~6 FPS at 1080, 4 FPS at 1440 and 2 FPS at 2160 would make it possibly squeeze just between the Fury's performance levels. 🙂 (Granted it's only for a single game and it's taken AMD a while to optimize their drivers for said game but it's a nice first glimpse of what more mature drivers could potentially do for this GPU model.) (And what Vega might be capable off as a more high-end 14nm model...we'll see when it's actually available though. 😀 )
#5313303
https://forums.guru3d.com/data/avatars/m/252/252334.jpg
So... Does this card use a reference pcb with default power delivery as the reference cards ? Oh btw, I agree with you HH, that's a David's star definitely 😀
#5313340
https://forums.guru3d.com/data/avatars/m/165/165326.jpg
So... Does this card use a reference pcb with default power delivery as the reference cards ?
No it does not , it uses an 8 Pin power connector plus superior 6+1 multi-phases where each phase supplies 25W instead of 22.5W found in other boards. Source: http://www.powercolor.com/Global/News.asp?id=1251 @ Hilbert H. , nice review. Great job as always 🙂.

Next Page