Cat's little mini-Adventure with a GTX1080(Fin)! Now with MOAR benchmarks!

The prologue!

A while back DR sent me a PM whether I would like to try a “monster” graphics card in my PC – now after giving it all of like 5 seconds thought,I instantly said “hell yeah”(well my head voice did),and a few weeks later I had a little surprise delivered to me.

A very shiny looking GTX1080 Founders Edition card.


The story begins


Now looking at the box - it comes across as quite understated and conservative looking - almost understated unlike some graphics card boxes in the past which seem to rather blingy and full of various icons and emblems professing how l33t they are.

http://i.imgur.com/jYhSa9Z.jpg

Even the included bumpf seemed more of the same - it reminded me of some expensive watch,etc in terms of the little wallet it is included in.

http://i.imgur.com/Ol68ago.jpg


Now I got to say looking at the review pictures of the GTX1080FE,I was not sure what to make of the angular design,but in person it is actually quite a nice looking card,and I would say it does feel very well constructed,especially compared to my last higher end cards in the form of the 8800GTS 512MB and HD5850 1GB which were mostly plastic fantastics.

http://i.imgur.com/V6cczLk.jpg


Now at the time I had somewhat “slightly” weaker card,and I had recently upgraded my ancient 1680X1050 jobbie to a nice new qHD UP2516D. However,with over a doubling off resolution over my previous monitor,I still needed to run things at 1920X1080.

The prospect was I could actually run things at native resolution AND AT MAXIMUM DETAILS!!

Now,as some might realise I am a “slight” SFF PC fan,so my current box is as follows:
1.)Xeon E3 1230 V2/Core i7 3770
2.)ASRock B75 mini-ITX motherboard(had a better one but it died,and I had not much choice)
3.)16GB 1600MHZ DDR4
4.)Corsair SF 450 SF-X 450W PSU
5.)Sandisk 480GB SSD
6.)2X2TB 7200RPM HDDs
7.)Aluminium Cubitek Mini Cube case

Its not state of the art by stretch of the imagination,but the old box still does a good job.

Now some of you might thinking – CAT,have you lost your banonos?? A pinted sized 450W PSU powering a monster like the GTX1080 – are you honestly going to try and blow that poor card up – its barely even had a full life yet.

Well,no folks,you see the GTX1080 might be a higher end card,but it certainly does not need a higher end power supply as you will see as my review trundles forward at its somewhat moose like pace only limited by my 7MB/1MB superfast Sky BB connection.

Prelude to Axe....sorry benchmarks.

Now I won't try to say my system will win any awards for build of the year,but as you can see the GTX1080 just about fits in,and since it is a blower cooler works very well in a SFF system like mine.

http://i.imgur.com/95yhz48.jpg

Zoomed in a bit closer you can see the illuminated Nvidia logo which is present on the card.

http://i.imgur.com/5gmHQAU.jpg

My initial thoughts about the cooler were very positive – despite some negative press about it,compared to the ones I had in the 8800GTS 512MB and HD5850 1GB it was actually subjectively less annoying. The fan seems to have a much lower pitch(although helped by me getting older and probably more tone deaf) and the cooler seems to not have the need to spin up all the time like a screaming banshee.

It is easily the best stock cooler and best blower cooler I have experienced on any card.

I already had an Nvidia card before,so setting it up was not an issue,and for the first time I decided to actually install GFE just to try out this new fangled recording milarky.


The testing methodology

What I do intend to to is test a few of the games I have installed at qHD and at 1080p. Due to a number of issues I was intending to get all this done earlier(sorry about taking longer than I expected,DR!),but I will try to benchmark as many games as possible,to give a good indication of performance. I might also try and compare it against one or two other cards,to show the difference at qHD and show why it is such a taxing resolution,but those come later.

It is also a realworld test,of whether an older system like mine can benefit from a decent GPU upgrade.

What I will try and do is use a mixture of playthroughs and benchmark tools to test game performance.

I also will included a video of the test sequence played recorded using ShadowPlay,as it had a lowish performance penalty. However,YouTube seems to have overcompressed the videos,so sorry for the artefacts.

Now I did intend to test Doom,but the game is a bit of a pain in the arse to benchmark due to the way it can randomise stuff. Also in some of the games,I might not go into CPU bound situations since this will be more a test of CPU processing power than anything to do with the card which would be unfair. Remember,I am also using ShadowPlay and FRAPS and these can reduce performance slightly.

Unless otherwise stated I am using the Nvidia 376.53 driver set which is a few months old. I tend to stay a few months behind latest drivers,to avoid hitting any issues with newer drivers,and because I don't always buy games on release. I will only change over to newer drivers after that period if I feel I can get performance improvements.


Now for something completely different...go! Go! Gadget benchmarks!

Everything to the max,or stutter trying.


14/03/2017 Update!!

I added some pictures of the GTX1080 card itself - evidently I forgotten to do before.

http://i.imgur.com/v8bqevF.jpg

http://i.imgur.com/pr0h87Z.jpg

All of the cooler shroud is made of metal,including the cap on the fan,and the card has a metal back-plate too.

It feels and looks really nice.

Benchmarks, First blood – the fantasy RPGs.

The Witcher 3

One of the most graphical intensive and system punishing titles out there,the Witcher 3 was launched to critical acclaim in 2015,and on max settings is a fantastic test of the mettle of a card.

You also get to see Geralt's locks in full dynamic real-time motion and with hair like that I do wonder why he can't smile more often??!!

As shown below the settings were maxed out at maximum everything.

http://i.imgur.com/hJQZU3j.jpg

http://i.imgur.com/Lz6iMPE.jpg

http://i.imgur.com/yb0Z7cG.jpg

http://i.imgur.com/b2Vjrba.jpg

I did three averages of the playthrough using FRAPs and the video indicates the playthough tested. The video is with v-sync enabled to reduce screen-tearing which is generally what I do for games as standard - I just forgot to switch it off for the video. I switched it off for the benchmarking OFC!!

https://www.youtube.com/watch?v=cNWZeS62iK8

http://i.imgur.com/tsw26kr.png

Now at 2560X1400 the game averaged 60.56FPS and was quite smooth to play with minimums in the late 50 FPS range. Now,you might ask,CAT,we noticed some mean arsed chaps at the end of the video - how do things fare during a battle? Very well TBH,as framerates were between the early 50s to around 60FPS despite Geralt going all Crouching Tiger,Hidden Dragon,but due to the variability in fights it didn't make sense to try and benchmark them.

Now lets see if we half the number of pixels,whether we can get close to the magical 120FPS mark for those 120HZ screens.

http://i.imgur.com/qLB73VL.png

Umm,nope - the game averaged 80.44FPS and that boys and girls looks like a CPU bottleneck there.

Looks like you might need a shiny new Core i7 7700K or R7 1800X instead! ;)

Now I told you something about the power consumption earlier,right??

During the benchmark run,according to my trust SilverCrest power meter,the system was drawing between 245W to just under 280W at the wall.

Told you,that one of those 900W £1000 PSUs was not required.

:)

Dragon Age:Inquisition

Coming out around six months before The Witcher 3,Dragon Age:Inquisition was one of the best looking titles of 2014,and you had the added advantage of playing as somebody a little bit happier than Geralt. Its also using the Frostbite engine,which will power Mass Effect:Andromeda which I am eagerly awaiting.

The settings used are as shown below.

http://i.imgur.com/kezRpSC.jpg

http://i.imgur.com/pleO0RD.jpg

I did three averages of the playthrough using FRAPs and the video indicates the playthough tested.

https://www.youtube.com/watch?v=qFWRcAhjazY

http://i.imgur.com/NYHcVck.png

Now as you can see Miss sole survivor Inquisitor lady moves around quite smoothly at an average FPS of 65.85FPS,with the worst minimums in the mid 50FPS mark.

Now lets see if we half the number of pixels,whether we can get close to the magical 120FPS mark for those 120HZ screens.

http://i.imgur.com/ZspUi4S.png

The average was 83.16 FPS.

I think I need a faster CPU.

Now what did Mr Power Meter say - the system consumed between 243W to 276W at the wall. Dragon Age:Inquisition is slightly better for the environment than The Witcher 3.

Now,some of you might say,CAT what about ditching MSAA 4X at 1080P?

It didn't do much to increase FPS,and power consumption even dropped as low as 192W and peaked at around 263W,which means the card seem to be less taxed overall.

Skyrim:Special Edition

The rejigged version of the seminal 2011 CPU sim....sorry RPG from Bethesda now with even more HD cows(the cake is a lie) and even more Nords who seem to have all gone to the same school of voice training.

OFC,benchmarking Skyrim without any mods will probably deserve the wrath of those ancient Daedric Princes,so I decided it needed to be lightly modded. Skyrim:Special Edition does not have the breath of mods the original version has,so there is room for it to be more taxing in the future.

The list of mods used are:
Trees in Cities - Special Edition
JK's Riverwood
Open Cities Skyrim
JK's Whiterun
Bijin All in One-SSE
Mature skin texture and body for UNP UNPB 7BASE CBBE Vanilla
Static Mesh Improvement Mod - SMIM
Climates Of Tamriel Special Edition - Weather - Lighting - Effects - Audio
Skyrim Better Roads
Skyrim Better Roads (SkyBeRoads)
Enhanced Lights and FX
SSE Texture Pack - Osmodius
Realistic Ragdolls and Force
Fores New Idles in Skyrim - FNIS
Wet and Cold
XP32 Maximum Skeleton Special Extended - XPMSSE
Unofficial Skyrim Special Edition Patch
Improved Bards - Special Edition
Immersive Citizens - AI Overhaul SE
Skyrim Sizes
Realistic Water Two
Relationship Dialogue Overhaul - RDO SE
Amazing Follower Tweaks SSE (Dheuster's Original)
BodySlide and Outfit Studio
ApachiiSkyHair SSE
Total Character Makeover
TRUE EYES SE

The game was run on maximum settings as shown below.

http://i.imgur.com/3rfkYk5.jpg

http://i.imgur.com/ui0zfHk.jpg

I did three averages of the playthrough using FRAPs and the video indicates the playthough tested.

https://www.youtube.com/watch?v=eeSEkJHVSPI&feature=youtu.be

Like all Creation Engine based games,Skyrim:Special Edition is locked to 60FPS,as the game physics are tied to framerates and things can get a bit wonky otherwise - its useful to unlock FPS to see the relative positions of CPUs though as the game can hit performance issues,since it only uses one maybe two cores intensively(it can cause more issues when you mod Skyrim).

http://i.imgur.com/MRpjXq1.png

As you can see my loveable Rogue of a character(you literally do play a criminal in the game) moves around quite smoothly at an average of 59FPS with an occasional movement to around 55FPS,however FRAPS might not be helping here.

Power consumption varied between 169W to 221W and the GTX1080 probably was having a nap it seems.

There was no real need for a 1080p set of benchmarks as the system would be capable of around 60FPS at that resolution anyway.

Benchmarks II - The Sequel! MMO FPS

Planetside 2

Now for a total change of pace from Medieval Land - to future present techno sci-fi land with Planetside 2. In an endless war with a near vertical learning curve you play one of multiple factions fighting over 4 distinct continents with upto 200 people bashing each other at any one point - eat your heart out Battlefied and COD,this is what PCMR people play! :p I hence decided to play as the spandex faction,also known as the Vanu - a player who wears spandex into battle knows no fear!!

Now regarding the game - it is quite CPU limited even when you probably don't think you are,but it is also very taxing on cards.

http://i.imgur.com/TM6ID8k.jpg

It seems Shadows were tanking framerates by 20FPS when set to Ultra(!) so I pushed them down to medium which meant framerates whilst standing stabilised around 70FPS to 75FPS,and this was in Snow Land,aka,as Esamir!

The following benchmarks were during one of the smaller battles with "only" 40 to 50 people. Larger battles can have 200 people and a whole lot of aircraft and tanks attacking the same point simultaneously and during an earlier battle framerates could sink to around 35FPS to 45FPS and that was with around 100 people,but I was hitting a CPU limit there and the in-game FPS counter will handily tell you if you are CPU limited or GPU limited.

I decided to record two sequences.

https://www.youtube.com/watch?v=U_yFGtvBSSs&feature=youtu.be

http://i.imgur.com/DqbfbtD.jpg

So in this first sequence I was mostly GPU limited and was most of the time above 60FPS.

https://www.youtube.com/watch?v=tx1W83msaxA&feature=youtu.be

http://i.imgur.com/6uayiTf.jpg

In the second sequence I was starting to hit some CPU limitations but was still able to mostly hit above 60FPS.

Remember I was using both ShadowPlay and FRAPS recording at the same time when getting the numbers with the test sequences,so there was an added CPU load.

So as long as you don't hit a CPU limit even at almost maximum settings you can still hit over 60FPS on Planetside 2.

Power consumption varied at between 140W to 160W at the wall according to Mr SilverCrest power meter extraordinaire.

Benchmarks III - The Return of the Benchmarks. Action-Adventure.

Rise of the Tombraider

In RoTTR you play as the young Lara Croft with her TRESemmé treated locks,which seem to stay immaculate what ever the world throws at them, whilst she raids tombs,to preserve the ancient knowledge by destroying half of them and causing mass extinction level events for half the local animal populations. I wonder if we will see a Green Peace Raider vs Tomb Raider game,soon? ;)

Unlike Geralt who seems bored half the time,Lara seems to be in a rush all of time,yet seems to be still able to crack out a slight smile - see Geralt just because you have full dynamic real-time motion animated hair,you shouldn't be so grumpy -it could be worse you could be Agent 47 and be a baldy man....this actually might explain a lot.

Now for the next few tests I upgraded to the new fangled 378.78 WHQL driver which has added performance improvements in a number of newer games. I tried The Witcher 3 and didn't see much improvement but it is from 2015 and using DX11.

The settings used are shown down below.

https://i.imgur.com/1Qo6aRs.jpg

https://i.imgur.com/FXPCJAR.jpg

Its important to enable Fullscreen and Exclusive Fullscreen together as it actually is meant to help performance a bit.

The first set of benchmarks is from the internal demo,which tests three different scenes - a snowy one with Lara just standing around looking starstruck in the freezing Siberian wastes,a second where she is standing starstruck looking at some ancient tomb,and a third scene where she is so starstruck you don't seem to see her at all,but we do get to see some scenery.

First up the qHD DX11 results.

http://i.imgur.com/sDOYQ5I.jpg

The qDH results don't look too bad - I am not sure what its up with the internal benchmark but the FPS counter I had running didn't seem to show these very low FPS and the animation seemed smooth. However,despite the fact multiple runs produced similar results,sometimes you could see situations where the minimums looked worse,etc but the averages look around the same.

http://i.imgur.com/bV8yF6s.jpg

1080p looks better,but I might be hitting some bottlenecks there - I wonder what it is?? We will soon find out.

Now moving to DX12,will things get better,I wonder??

Well the problem is I hit a snag - being an utter noob at DX12 testing,I soon discovered to my chagrin most popular software does not support DX12 testing yet.

I soon found two packages which did work PresentMon and Mirillis Action! The problem with PresentMon is its finicky to setup and is CLI,which is not ideal for the way I am testing things ATM(but I also a lazy git too). However Mirillis Action! at times can cause some greater FPS drops than FRAPS.

https://i.imgur.com/gZ0RLFr.jpg

Not much difference at qHD,minimums did seem on average better.

http://i.imgur.com/P0fi4rZ.jpg

However,at 1080p,look at the GTX1080 fly!!

Well that is all fine and dandy,CAT,you might say - but what about when Lara is flying around tombraiding and traipsing around the countryside without a care in the world?

OK,lets do some time runs then!!

The first one is in the Geothermal Valley in the Village of the....Remnants.

Its known as one of the most intensive parts of the game,so I thought why not??

https://www.youtube.com/watch?v=PVwzZcYVOso&feature=youtu.be

That is a video of the time run under DX11.

Now lets look at some MOAR charts because everybody loves MOAR charts right?? Wait,Wut,you don't?? TF2 is that away-----------------------------> ;)

So firstly lets do some under everybodies favourite API,DX11.

https://i.imgur.com/KcAkpkB.jpg

At qHD I get an average of 51.32FPS,which is not bad considering it is one of the most taxing parts of the games. Power consumption varied between 245W and 278W.

https://i.imgur.com/IcDp6PR.jpg

At 1080p,I only get 54.51FPS despite dropping the number of pixels to be rendered in half??

Emm,I didn't realise it was the CPU being the main issue here - but come to think of it it does do a lot of stuff like have animated chickens which randomly move about. Who could have thought Tombraider had such realistic realtime chicken animations,eh?? ;)

Now,lets see if Microsoft's new kid on the block API,DX12,which is meant to be the saviour of PC gaming will swoon in and save the day??

https://i.imgur.com/gjlWVWz.jpg

Umm,I am not so sure. At first I thought it was the recording software causing some problems,but even switching it off,restarting the PC and just using the Nvidia FPS counter,I saw the same thing. Certain runs would be fine and certain other ones would stutter at one point and that was the end of any hopes of decent framerates - I am loading the game off an SSD connected to a SATA III port BTW.

I decided to do six runs for the DX12 tests to see how often it would happen,and it happened quite a bit!! :(

It does make me wonder,whether it was related to the animal animations in the village,as they don't always seem to follow the exact same path,but even when I tried a different path something similar would happen. It would randomly drop during one time run and the next one it would go up again.

Anyway,lets see if the same issues happened at 1080p.

https://i.imgur.com/I8yB2A3.jpg

It is not as bad but still happens - it is shame since when things do work properly the FPS jump is HUGE.

Having said that the Rise of the Tombraider DX12 mode is not heralded as the best showing of the API and has been known to be buggy.

15/03/2017 UPDATE - I think I have an answer to what is happening - check post 22!

However,I thought ,lets try another part of the game to see if the issue was just related to the section I was testing.

Or it could be DX12 thinks its better than my old computer,so said screw you hippy - Skylake or newer for me!!

MOAR TESTS!

So,I decided to find an area with less chickens,so settled on the area just before the final confrontation in the game.

The area tested is shown below.

https://www.youtube.com/watch?v=O9ArNWOr7s0&feature=youtu.be

This type of scene is actually quite common in the snowy sections of the game where you are running through areas with not much happening and trying to avoid all the scenery which is out to get you.

https://i.imgur.com/XmL4Mcb.jpg

In DX11 you get a very smooth 74.05 FPS despite all the lighting and snow effects in place.

Now lets move to a less taxing resolution.

https://i.imgur.com/CBlBxFc.jpg

Again not quite doubling the framerates as the system is hitting some bottlenecks there,but still 102.33 FPS is nothing to complain about!!

Now lets see how DX12 fares.

https://i.imgur.com/Qr7o9Hr.jpg

At qHD framerates seemed around the same at 72.68 FPS,and there apparently was none of that weird behaviour I saw in the earlier test scene.

Now lets shift to 1080p.

https://i.imgur.com/JScQ68U.jpg

Again slightly more consistent framerates and a higher average of 106.46 FPS.

Remember,I am using Mirillis Action! for DX12 performance recording which seems a tad worse than FRAPs for the performance hit.

So I am still a bit perplexed at what was happening.

Benchmarks IV - The Benchmarks Strike Back.Again. Cyberpunk RPG.

Deus Ex:Mankind Divided.

You are agent Adam Jensen,mechanically augmented super person,and the only good mechanical man left in the world,after all the other ones had a BSOD,who is trusted with saving the world,a second time,against the Lizard people who want to make everybody NOT great again,as once was not enough. Seriously isn't a competitive game of Croquet not enough for these people??

I did try ramping all the settings up,until I hit the SSAA setting,which seems the most pointless form of AA ever - sure lets render the image at a silly resolution just to remove some jaggies,so we can tank performance at a much lower resolution. TAA it is then.

The settings used are detailed as below.

https://i.imgur.com/KW3Cg0q.jpg

https://i.imgur.com/g9A4pks.jpg

Now I used the internal benchmark in the game for testing - I have not got far enough in the game to do a custom test sequence,which will be nothing but me sneaking down a corridor trying not to get shot at by a 100 people.

I used the same settings as the Bit-tech except I used higher AF,which seems to have not made any difference to the scores when I turned it down. Apparently the benchmark is actually more taxing than most sections in the game anyway,so is a good test for the worst case scenario.

Now,I made sure I did three runs of each set of settings used. The benchmark seemed to be very consistent in performance whether DX11 or DX12 was used.

https://i.imgur.com/18sm89i.jpg

Average Minimum Maximum
DX11 1080p 72.23 55.40 93.30
DX11 1440p 51.67 39.80 64.60
DX12 1080p 71.70 53.77 92.37
DX12 1440p 49.90 39.07 61.10

There was not much difference between DX11 and DX12,with DX11 being slightly ahead in performance,but there was nothing serious. It is an AMD sponsored game,so I would expect to see greater uplift in their cards,and performance seems in line with some reviews I have seen.

Now how does performance compare with in the game,so far - during the final battle during the initial part of the game,which has many NPCs,sand and dust effects with people shooting everywhere(and my uber agent character not being nearly bulletproof enough) I was getting between 50FPS to 60FPS.

Power consumption during the DX11 qHD test varied between 229W and 263W.

Clockspeed to the near max - a little bit of overclocking!

Now TBH,I didn't really feel the need to overclock the card,since stock performance was good enough for my purposes,and I didn't want to have to ramp up the fan,however I decided to overclock......FOR SCIENCE!!

Looking at some reviews I decided to try a "safe" overclock of +200MHZ to the core and +100MHZ for the memory. There are rumours that Nvidia will be refreshing the GTX1070 and GTX1080 with versions with higher clocked RAM,and I suspect they would also come with a higher core clockspeed too,so it will be interesting to see how much even a modest overclock will make.

http://i.imgur.com/nPSIAvh.jpg

The core boosted up to 2.1GHZ during the benchmarks.

http://gpuz.techpowerup.com/17/03/12/4ux.png

Thats the first time I have used a 2GHZ+ GPU.

I started with Deus Ex:Mankind Divided as that is one of the more taxing titles I owned.

http://i.imgur.com/7ssECYI.jpg

Average Minimum Maximum
DX11 1440p 51.67 39.80 64.60
DX11 1440p(overclocked) 57.97 44.87 72.93


That is a 6 FPS improvement in the average framerate and a 5 FPS improvement in minimums. With a few more overclocking tweaks,you should be able to get to around 60FPS during the benchmark. Power consumption was now 288W to 314W.

Now,we move over to Rise of the Tomb Raider.

http://i.imgur.com/XJW8Lri.jpg

At qHD we see an average of 56.74FPS,which is an increase of 5.45FPS. Now if you look at the bottom chart,it is difference of the average value between each set of runs.

The area with the lowest FPS,seems to have a tiny but consistent improvement,so I am not sure now if it is an entire CPU bottleneck at that point or a combination of a CPU and GPU issue. Power consumption was between 255W to 314W.

http://i.imgur.com/yck1BZM.jpg

At 1080p we see an average of 57.97FPS,which is an increase of 3.42FPS,with a similar pattern to the previous set of results,but again it is not massively higher than the qHD results.

Next,we look at The Witcher 3.

http://i.imgur.com/9yOiWjx.jpg

The average framerate increased to 66.82FPS which was a 6.26FPS increase over stock clockspeeds. Power consumption increased to between 299W to 314W.

Summary

So what do I think of the GTX1080 Founders Edition after using it for a while - I am generally impressed. In all of the titles I tried it generally has meant I can up settings to silly levels and still get a perfectly smooth experience - over my previous card I not only got higher framerates,but also at a higher resolution and with more settings turned up. The cooler is also pretty decent too - I found it generally unobtrusive,and since the card didn't draw a huge amount of power,it also meant by extension my PSU could stay in passive mode longer.

However,it has exposed that in some titles my CPU is starting to become an issue,but OTH still for plenty of games like The Witcher 3 for example,it will breath plenty of life into an older system. I can't make much comment about DX12,but looking at the review results for the games I tested especially Rise of the Tomb Raider it is all over the place,so I think it does not help that DX12 is a work in progress currently.

Even though I didn't show results for DOOM,Vulkan seems to work much better overall from my experience of running the game. However,I uninstalled it recently,since I was running short of space on my SSD and its mahoosive for an FPS game.

Apparently Nvidia is launching refreshed versions of the GTX1070 and GTX1080 cards onto the market soon with faster RAM and probably bumps to the core clockspeeds too,so this level of performance will be cheaper than ever,as this is a criticism people had about the card,ie,the cost. However as cards go,it does feel like Nvidia have got inspiration from Apple in making a product which looks and feels very expensive,so I can appreciate people paying a high price. It certainly is in a class above what I usually buy!

Regarding the Nvidia ecosystem,I did find Shadow Play quite easy to use - however some criticisms I have is the fact I didn't see a way to change the encoding codec used,and even more importantly it would have been awesome if Nvidia included a FRAPS like ability to record framerates and frametimes,as it was a bit fiddly to get measurements for DX12 games.

Anyway,thanks again DR for giving me a chance to try a high end piece of kit in my system!

Interesting run down CAT, thanks for that :)

One thing I think is worth mentioning though is that virtually no GPUs today will hit their theoretical pixel fill rates in games, which are far more geometry and shading contrained than pixel fill constrained. In pure pixel terms a 1080 could render over 30,000 qHD frames per second ;) So you shouldn't expect to get double frame rates by halving the rendered pixels - the pixel pipelines aren't the bottleneck on the GPU ;)

What's interesting to me is the dips below 60FPS in some titles at 1080p. You often hear that buying a 1080 (or even 1070) for 1080p is pointless, which I've always disagreed with.
Given that games only get more demanding as time goes on, and some titles are already dipping below 60fps, surely that shows the need for such a fast card even at a relatively low resolution? Ok CAT's CPU could be a bit of a bottleneck, but it's not that slow.

Do you need a 1080 at 1080p? Absolutely not, but with limited funds I think I'd rather buy a 1080 with a 1080p monitor and crank the settings up than a 1070 with a 1440p monitor and some settings turned down.

Quote:

---

Originally Posted by Bagnaj97

... Do you need a 1080 at 1080p? Absolutely not, but with limited funds I think I'd rather buy a 1080 with a 1080p monitor and crank the settings up than a 1070 with a 1440p monitor and some settings turned down.

---

Interesting - I'm quite the reverse. I'll happily turn settings down to match the GPU power available. If a game needs highest IQ settings to be worth playing, then to me it's not worth playing ;)

Addendum!


Benchmarks Re-dux!


Now my mate said,CAT why not compare it to another card?? Well,I thought enough,fair point,so I decided to test against a card I had access too, a Powercolor Devil RX470 4GB which is one of the best RX470 4GB cards out there currently and has one of the highest pre-overclocks of an RX470 as it matches the stock clocks of a reference RX480,and reviews show it can maintain higher clocks too.

http://i.imgur.com/4nmsx6n.jpg

It also continues the AMD tradition of having massive midrange cards - its nearly as big as the GTX1080. Also the GTX1080 does feel a step-up in build quality although for a midrange card the RX470 seems fine TBH.

Now to the benchmarks. The AMD card had the latest Radeon 17.3.1 drivers and the GTX1080 had the latest 378.78 ones.

The first came to be tested was the AMD sponsored Deus Ex:Mankind Divided.

http://i.imgur.com/L58B1Sh.jpg

http://i.imgur.com/VTrgkpU.jpg

The pre-overclocked RX470 puts up a valiant fight but just about can muster slightly more than half the performance of a GTX1080 in the title,and the GTX1080 is upto 70% faster.

Interestingly enough even with DX12,the AMD card either shows no performance improvements or a slight regression,which is in-line with what we saw with the GTX1080.

Power consumption was between 223W to 238W as opposed to the 229W and 263W for the GTX1080 during the qHD DX11 run.

Another thing - the GTX1080 was whisper quiet in comparison with the RX470 which was very noisy.

Maximum Geralt! MOAR The Witcher 3 testing!

Even though Geralt was still grumpy from his last encounters with The Wild Hunt I decided I needed to wheel him and his super animated locks out for another set of benchmarking.

Now lets start at qHD.

http://i.imgur.com/iaDvgZ3.jpg

The GTX1080 is just over the 60FPS mark on average and the RX470 is around 32.83 FPS or almost half the speed - The Witcher 3 is not a VRAM heavy title,so it is simply the case that the RX470 is running out of grunt here.

Lowering the resolution to 1080p does help the RX470 a tad more though.

http://i.imgur.com/4hiC08K.jpg

It is 80.44 FPS against 45.17 FPS,this time but if you do want to have the game at max settings and get 60FPS at 1080p I think you need at least a GTX1070 or a Fury X.

Power consumption for the RX470 varied at between 224W to 238W at qHD as opposed to the GTX1080 which around 245W to just under 280W at the wall.

Dragon Age:Inquisition

I thought Miss sole survivor Inquisitor lady probably need to go for another stroll round the ramparts,after all being stuck in that War Room can't be that great for her health,can it?? ;)

Now let's start with the qHD benchmarks again.

http://i.imgur.com/yhwfllC.jpg

Instead of a PCMR 65.85 FPS experience with the GTX1080 you get a cinematic 33.71 FPS experience with the RX470 instead,together with more HD fan noise to boot. Power consumption varied between 235W to 243W instead of the 243W to 276W at the wall for the GTX1080.

Now,some of you Eagle eyed folk might ask,but CAT doesn't the game support Mantle?? Yes,you would be right,but since none of the software I had could be used to record FPS and since the AMD software package didn't seem to have any obvious ways to do it,I decided to test the card in DX11. Mantle might help but I don't have the eye sight of a Pigeon.

Considering qHD was a bit too much for the RX470,how about 1080p you might ask??

http://i.imgur.com/PRyx2a6.jpg

The GTX1080 runs through the game at 83.16 FPS,and the RX470 almost gets there at 51.06 FPS.

Considering this is the closest the RX470 gets to 60FPS in the titles I tested so far,lets see if some overclocking might help.

AMD,has radically changed its menu system,so I had to consult the very useful HardOCP review of the RX470 Devil:

http://www.hardocp.com/article/2016/...cking_review/3

http://i.imgur.com/CzVisRs.jpg

So I settled on +115MHZ for the core and +100MHZ for the memory.

Now lets see if it makes much difference.

http://i.imgur.com/4y7Mbm9.jpg

The FPS went from 51.06 FPS to 55.93 FPS,which is almost 10% extra but not quite enough to get to 60FPS either. The power consumption went up by around 30W to between 265W to 276W,which is what the GTX1080 was consuming at qHD whilst pushing more FPS.

http://i.imgur.com/fbz0l7x.jpg

As you can see here,at 90% fan speed,which was very noisy,the core was only at 65C,and GPU utilisation was at 100% and the core had no problem maintaining its overclock.

LARA doing even more traipsing around!

I must ask her what brand of walking boots she wears....!

Lets look at the internal benchmark before looking at the time runs.

http://i.imgur.com/e3NRlTG.jpg

At qHD the RX470 produces around 40FPS with the same caveats as seen with the GTX1080 testing. The GTX1080 has double the score.

http://i.imgur.com/4RM5XlC.jpg

At 1080p things do improve and the RX470 is just shy of the magical 60 FPS mark. The GTX1080 is just under 100 FPS in the same benchmark.

However,the RX470 does seem to have more of the very quick framerate drops when compared to the GTX1080 even at 1080p.

Now,I left the best for last - Rise of the Tomb Raider seemed very taxing in the Village of the.....Chickens,so lets see how the RX470 stacks up. Will it be clucked into oblivion??

Baby steps first,lets see how DX11 performance is for the RX470.

First the qHD results.

http://i.imgur.com/XpMkAsR.jpg

The RX470 produces a sub-cinematic 19.84 FPS against the 51.32 FPS of the GTX1080. The RX470 probably ran out of VRAM. Power consumption varied at between 134W to 238W which kind of supports this theory.

Lets look at the less VRAM taxing resolution of 1080p.

http://i.imgur.com/tZwbxDD.jpg

The RX470 does a bit better,producing a framerate of 27.84 FPS against the 54.51 FPS of the GTX1080 - VRAM usage for the RX470 peaked at just under 3.9GB of VRAM.

Sorry for the slightly inconsistent time-runs,framerates got so low for the RX470 it was hard to control the mouse.

Now for some DX12 testing - will it save the RX470??

http://i.imgur.com/WWyIiKo.jpg

At qHD there seems to be no improvement.

http://i.imgur.com/WVdV6l4.jpg

At 1080p it is the same - remember the software I use for DX12 benchmarking has a worse performance hit than FRAPS. However the GTX1080 went from 95 FPS to 113 FPS,whereas the RX470 got slightly worse.

Now I started the Village benchmark run in DX12 on the RX470,and the FPS was hovering around the 20FPS mark,so it wasn't really doing much.

Hence,I quickly went to 1080p instead.

http://i.imgur.com/O3eoSSb.jpg

The RX470 did see a decent improvement from 27.84 FPS to 36.67 FPS which was a 32% improvement,and even though the deviations were worse on the GTX1080,you can still see some variability in results,ie,sometimes framerates would drop for no reason.

Now,the GTX1080 improved from 54.51 FPS to 66.81 FPS overall which was a 22.6% improvement,but if you took the outliers out the average increased to 70.23 FPS,which was a 28.8% improvement.

However,as a percentage of the average framerate,the RX470 did seem to suffer much worse drops though down to 10 FPS to 20 FPS. The GTX1080 in the worse case scenario was just under 40 FPS at its lowest point.

Summary Re-dux!

Having tested out the Powercolor RX470 Devil 4GB,its not a bad card for its segment,but when you look at things like the power consumption of the GTX1080 and the acoustic profile of the GTX1080,it really does seem a step down especially when you consider the fact,even in its weakest moments,it can still give an overclocked RX470 4GB a good thrashing.

Its one thing AMD probably needs to learn from Nvidia - a half decent blower cooler on its cards,since it really is useful for SFF PC users like me.

Thanks for the review Cat. Can I ask some questions about the green teams image quality and drivers? How do you rate them, primarily? Are the AA modes and texture filtering options good? And do they have anything like AMD's negative LoD bias when using adaptive AA to keep textures sharp?

Quote:

---

Originally Posted by kalniel

Thanks for the review Cat. Can I ask some questions about the green teams image quality and drivers? How do you rate them, primarily? Are the AA modes and texture filtering options good? And do they have anything like AMD's negative LoD bias when using adaptive AA to keep textures sharp?

---

Well this will be the first AMD card I have tried out since an HD7850,since I had a GTX660 and a GTX960 after that. Is there a good way to test this,so I can give a definite answer??

Possible explanation for DX12 inconsistencies!


Right,I started mucking around with some settings - I think I can explain some of the weirdness I saw under DX12. It looks like if I turn down texture settings when I hit the runs which are a bit crap things get much better. I have a feeling there is some intermittent issue between the very high texture setting and DX12,which isn't apparent under DX11 in the area I am testing. However,looking at VRAM usage under DX12,it is a bit higher than DX11,and under DX11 at qHD at 1080p it is between 7GB and 7.5GB,and DX12 mode seems to be a bit higher than that. It would also explain why at 1080p the issue happened less often,as VRAM usage should be a bit less overall.

Wow, now that's an amazing bit of work CAT. My stock Zotac founders 1080 also hits over 2GHz with a small OC (similar to yours) and power draw from the wall is in the ~310W range as well.

Quote:

---

Originally Posted by CAT-THE-FIFTH

Well this will be the first AMD card I have tried out since an HD7850,since I had a GTX660 and a GTX960 after that. Is there a good way to test this,so I can give a definite answer??

---

Hmm I guess try some of the different AA and texture modes and see what happens ;) Novigrad in TW3 might be a good test scene as plenty of objects with jaggies yet also high res textures to try and keep sharp. Skyrim and trees/vegetation is another one. You may encounter some issues with various AA modes not making any difference where a game uses non-forward rendering - sometimes drivers are clever and manage to inject some form of AA in there, so it's a good test of the driver teams ingenuity, but where that fails the only remaining options are post process AA methods like FXAA which blur textures, and super sampling/downscaling.

Quote:

---

Originally Posted by kalniel

Hmm I guess try some of the different AA and texture modes and see what happens ;) Novigrad in TW3 might be a good test scene as plenty of objects with jaggies yet also high res textures to try and keep sharp. Skyrim and trees/vegetation is another one. You may encounter some issues with various AA modes not making any difference where a game uses non-forward rendering - sometimes drivers are clever and manage to inject some form of AA in there, so it's a good test of the driver teams ingenuity, but where that fails the only remaining options are post process AA methods like FXAA which blur textures, and super sampling/downscaling.

---

I will have a look,but I need to get to Novigrad first!! :p

Edit!!

It might take a few days though,due to stuff.

Quote:

---

Originally Posted by bsodmike

Wow, now that's an amazing bit of work CAT. My stock Zotac founders 1080 also hits over 2GHz with a small OC (similar to yours) and power draw from the wall is in the ~310W range as well.

---

Thanks - good to see that my numbers are kind of confirmable regarding clockspeed and power consumption.

https://youtu.be/4aM8CMSPfWQ?t=438

Seems like this thread has gone mainstream :)