Overclock.net › Forums › Graphics Cards › Graphics Cards - General › I don't see any real benefits in gaming performance when i OC
New Posts  All Forums:Forum Nav:

I don't see any real benefits in gaming performance when i OC

post #1 of 20
Thread Starter 
So, lately i haven't bothered much with OC/Gaming, but i now i have some time on my hands so i got busy:

Dirt Rally is my current poison, so first of all i didn't like the high 70's temps my GPU was getting now i the summer (personal but informed opinion), so i gutted my sweet little 10yo PC and improved the GPU cooling; i replaced the stock heatsink with the Big Typhoon i had on my CPU since 2007.
*CPU now has a Raijintek Themis, thanks for the help and suggestions guys!

After a couple of days and while the mentioned "mod" is not 100% complete, i got back to gaming and i was happy that the GPU was ~20c cooler.
But then i got the itch to improve the tiny little issues i was getting in that game (rare drops to <60fps now and then). Surely a bit of OC can cure this, since at stock i am 95% of time right in the 60FPS neighborhood. Right?

Well as it turns out (and at least without help?) no, no my modest OC will not change much if anything frown.gif

Quick PC specs before i go any further: PC specs! (Click to show)
Asus P5Q3 MB (s775, P45 chipset, 4xDDR3 only)
Xeon X5460 CPU (45nm, Quad 3.16GHz, 9.5multi/333bus/1333FSB, 2x6MB Cache)
Crucial Ballistix Sport 8GB Kit (budget 2x4GB 1600MHz 9-9-9-24, running at 1333/8-8-8-22 on stock CPU)
Sapphire Dual-X R9 270X 2G OC (1070MHz Core stock)
Corsair TX750W PSU
1x40GB SDD with only Win7 on it :/
1x2TB slow-as-hell HDD for anything else
sound cards come and go...

First of all, CPU OC does nothing in my (very particular?) case, so let's get that out of the way:
Yes other people and other tasks may be benefited by raising CPU clocks, but for me maybe 1fps min/max gained from 3.16GHz to 3.8GHz is not worth mentioning, more so because i can get 5 times that by only raising the GPU core by 30MHz....

So in this day and age, GPU it is then:
Well i am a complete noob to GPU overclocking, but i did check were stock-cooled (and yes not all stock coolers are the same ofc) can reach, and even if i only got to 20% of that OC potential (100% being max stock-cooling OC and cooking eggs on your card), i still don't really see and benefit in my case (and here "my case" is really important).

So i started my "2016 OC journey":
Since room temps at this time of year are really high (30c and up) and i know from years of experience that will give me horrible PC temps, i didn't go crazy:
I very hastily read generic OC guides, just so i can spot any "if you have a R9 AMD chip don't you raise Vcore above X.XX volts or it WILL blow!' or any "head's up, FurMark can blow up your VRM MOSFETS" (that is nice to know because i used this util for ages lol). and then its "raise sliders!" time...
Using the Trixx util that came with the card (only 4.x versions because otherwise i can't adjust Vcore), i first raised only the core MHz, and that had an effect alright. But nothing to celebrate about, and honestly ok, i wasn't expecting miracles from a <3% clock increase.

But as i, yes hastily and without "proper methodology", tried to push more aggressively (= closer to the max my chip can take?), i became aware that this does makes a difference, but nothing really in real life gaming for me in this particular game, in this particular time or the year and whatever.

To summarize my OC experience:
CPU only OC did almost nothing for this game in numbers (maybe 1fps difference), ABSOLUTELY NO DIFFERENCE in real life feeling playing the game.
GPU very mild OC (raising Core from 1070 to 1100MHz only), did more than the CPU OC in Heaven test numbers (~+5fps Min, maybe +1fps Avg and Max), but also ABSOLUTELY NO DIFFERENCE in real life feeling playing the game.
Pushing GPU further only gave minuscule results and PC crashes:


*results are in order of testing
Stock: 1070 core / 1400 mem / 1,238 Vcore / 0% Power
1100 / 1400 mem / 1,238 Vcore / 0% Power - significantly higher Min FPS in Heaven bench, ZERO difference in Dirt Rally gaming
1120 / 1500 mem / 1,250 Vcore / +20% Power - not worth mentioning better Min FPS in Heaven bench, ZERO to marginal but 1000% not worth the OC difference in Dirt Rally
1155 / 1550 mem / 1,270 Vcore / +20% Power - PC Crashed within 5sec of gaming
1140 / 1520 mem / 1,260 Vcore / +20% Power - PC Crashed ~4mins into Dirt Rally, right at the point that is ALWAYS "stutters" and FPS plummets (see the big dip in the graph at ~250sec)
Max GPU temps on these tests? Only 60c with 30-31 room temp !!!

Test some more? F@ck no, i will not cook my GPU to gain another 1fps that will do absolutely nothing about the 10 FPS "dips" i am experiencing in fixed points of the game.

This is what i experienced, now i would appreciate the help/opinions/comments:
Did i do something (horribly or otherwise) wrong?
Is that all i can expect to get out of this GPU and/or in this particular game?
Games are on a separate partition on the HDD that is 100% fragment-free.

Don't tell me to get an 290X instead, i know that much XD

Thanks for the reading people!

Edit: I am using a LG IPS235 Monitor and game at 1080p/60fps if that's relevant...
Edit2: Graphs are from a single ~6:30 Dirt Rally stage using these settings.
Edited by JohnnyG01 - 7/23/16 at 6:04am
Thy Old Lady
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon X5460 Asus P5Q3 (P45/ICH10R) Sapphire R9 270X Dual-X 2GB OC Crucial Ballistix Sport 8GB Kit 
Hard DriveHard DriveCoolingCooling
Kingston SSDNow 40GB Westrn Digital WD20PURX Purple 2TB Thermaltake Big Typhoon (GPU) Raijintek Themis (CPU) 
OSMonitorPowerCase
Windows 7 Ultimate x64 LG Flatron IPS235 Corsair TX750W Coolermaser Elite 332 
Other
Cat 
  hide details  
Reply
Thy Old Lady
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon X5460 Asus P5Q3 (P45/ICH10R) Sapphire R9 270X Dual-X 2GB OC Crucial Ballistix Sport 8GB Kit 
Hard DriveHard DriveCoolingCooling
Kingston SSDNow 40GB Westrn Digital WD20PURX Purple 2TB Thermaltake Big Typhoon (GPU) Raijintek Themis (CPU) 
OSMonitorPowerCase
Windows 7 Ultimate x64 LG Flatron IPS235 Corsair TX750W Coolermaser Elite 332 
Other
Cat 
  hide details  
Reply
post #2 of 20
Dirt Rally seem to prefer higher IPC

post #3 of 20
Yeah CPU overclocking is a not much help these days.
GPU is where its all at.
post #4 of 20
Thread Starter 
Quote:
Originally Posted by PontiacGTX View Post

Dirt Rally seem to prefer higher IPC


Well then, that graph tells me that a modern CPU can bottleneck two 980 Ti's in SLI...
In my case, gaming/benching Heaven and Dirt rally with CPU at 3.16 (mem @1333) and again at 3.8GHz (mem @1600) makes no difference. Unless the 'problem' is that i should OC higher, like over 4GHz, but i doubt that would yield any better results, in my case at least.

Quote:
Originally Posted by MuscleBound View Post

Yeah CPU overclocking is a not much help these days.
GPU is where its all at.

GPU OC it is, but are the benefits of OCing in gaming so small for other people too, or do i just did it wrong or is my setup bad in some way?
I know a 270X is entry-level gaming-wise (not that i am not satisfied with its performance), but still for the 50MHz Core that i could raise in a hurry i only got ZERO in-game difference, a 1-2 fps higher Minimum frame in that dip the game does in that Monaco stage, and a ~5fps gain in Minimum FPS in Heaven test that would only make someone happy if they played at 13FPS lol.

I mean sure 1120 Core / 1500 Memory is not a big OC and i could get some more if i kept testing and using proper technique (raising clocks in little steps/testing/repeat) but i just don't think all those PC freezes/Reboots/HHD chkdsk's and general messiness is worth that. What will i get if i go to 1200 Core? Another couple of higher Minimum FPS that may be noticeable in-game during that 'stuttering'?

The goal was to get rid of the stuttering altogether (or at least to make it dip a couple of frames and not 10), and that is clearly not going to happen in my case. So why should i OC then, i don't care about posting a higher 3DMarks score, i just want to improve my gaming experience...

This performance difference is not acceptable! thumbsdownsmileyanim.gif

This is OCing just so you can say "hey, i overclocked!" frown.gif
Edited by JohnnyG01 - 7/23/16 at 4:48pm
Thy Old Lady
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon X5460 Asus P5Q3 (P45/ICH10R) Sapphire R9 270X Dual-X 2GB OC Crucial Ballistix Sport 8GB Kit 
Hard DriveHard DriveCoolingCooling
Kingston SSDNow 40GB Westrn Digital WD20PURX Purple 2TB Thermaltake Big Typhoon (GPU) Raijintek Themis (CPU) 
OSMonitorPowerCase
Windows 7 Ultimate x64 LG Flatron IPS235 Corsair TX750W Coolermaser Elite 332 
Other
Cat 
  hide details  
Reply
Thy Old Lady
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon X5460 Asus P5Q3 (P45/ICH10R) Sapphire R9 270X Dual-X 2GB OC Crucial Ballistix Sport 8GB Kit 
Hard DriveHard DriveCoolingCooling
Kingston SSDNow 40GB Westrn Digital WD20PURX Purple 2TB Thermaltake Big Typhoon (GPU) Raijintek Themis (CPU) 
OSMonitorPowerCase
Windows 7 Ultimate x64 LG Flatron IPS235 Corsair TX750W Coolermaser Elite 332 
Other
Cat 
  hide details  
Reply
post #5 of 20
Quote:
Originally Posted by JohnnyG01 View Post

Well then, that graph tells me that a modern CPU can bottleneck two 980 Ti's in SLI...
In my case, gaming/benching Heaven and Dirt rally with CPU at 3.16 (mem @1333) and again at 3.8GHz (mem @1600) makes no difference. Unless the 'problem' is that i should OC higher, like over 4GHz, but i doubt that would yield any better results, in my case at least.
even with a single card it gets performance gains
https://www.computerbase.de/2015-12/dirt-rally-benchmarks/2/#abschnitt_prozessorbenchmarks
http://www.pcgameshardware.de/Dirt-Rally-Spiel-55539/Specials/Benchmark-Test-1182995/

and it seems to prefer CPUs with HT or more cores
post #6 of 20
Thread Starter 
Quote:
Originally Posted by PontiacGTX View Post

even with a single card it gets performance gains
https://www.computerbase.de/2015-12/dirt-rally-benchmarks/2/#abschnitt_prozessorbenchmarks
http://www.pcgameshardware.de/Dirt-Rally-Spiel-55539/Specials/Benchmark-Test-1182995/

and it seems to prefer CPUs with HT or more cores

Fresh out the oven that my room has become (31.9c atm): CPU 3.8GHz more tests (Click to show)
3.8GHz / 1600MHz setup


After CPU OC only run


After CPU/GPU OC run


Comparison

Please note the 64-67% Max CPU load shown in HWMonitor... On stock CPU i get 75-85% regularly (i am keeping an eye on sh|t like that).
Once my again my older generation and cheaper parts PC does not respond to OC the same way as those 4770K's frown.gif
Thy Old Lady
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon X5460 Asus P5Q3 (P45/ICH10R) Sapphire R9 270X Dual-X 2GB OC Crucial Ballistix Sport 8GB Kit 
Hard DriveHard DriveCoolingCooling
Kingston SSDNow 40GB Westrn Digital WD20PURX Purple 2TB Thermaltake Big Typhoon (GPU) Raijintek Themis (CPU) 
OSMonitorPowerCase
Windows 7 Ultimate x64 LG Flatron IPS235 Corsair TX750W Coolermaser Elite 332 
Other
Cat 
  hide details  
Reply
Thy Old Lady
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon X5460 Asus P5Q3 (P45/ICH10R) Sapphire R9 270X Dual-X 2GB OC Crucial Ballistix Sport 8GB Kit 
Hard DriveHard DriveCoolingCooling
Kingston SSDNow 40GB Westrn Digital WD20PURX Purple 2TB Thermaltake Big Typhoon (GPU) Raijintek Themis (CPU) 
OSMonitorPowerCase
Windows 7 Ultimate x64 LG Flatron IPS235 Corsair TX750W Coolermaser Elite 332 
Other
Cat 
  hide details  
Reply
post #7 of 20
Thread Starter 
So, after more tests to see what's up, i found out that in all but one of the games i have installed, max CPU usage is 60-80% (mostly laying around 60-65% when gaming):
Kerbal Space Program, SpinTires, Path of Exile, Endless Legend and Dirt Rally are all around 50-60% during gameplay, maybe with spikes that reach 70-80& utilization.
The exception is GTA V, which uses all of my CPU, that sits between 90+ and 99% while gaming.

The info from the German site shows that a faster than mine CPU can make a huge difference in a machine with a also much faster GTX980 Ti or Fury X.

If all of my games didn't use all of my CPU then it would make sense (something is broken and i can't use all of my CPU).
If i had a 8x4770K setup it would make sense (i got so much CPU that i can't use it all).
Not that one and only one game (and CineBench/IntelBurnTest/CPU-Z bench and cr@p) is using the CPU to the max and the others don't,i have no idea what is going on... But i bet it's not good mad.gif

I had a cunning plan to get more info, like if with my 270X, 60% of the X5460 is all it takes to do the best it can on Dirt Rally (GPU load on said game is >90%), so i reduced my CPU speed to see if it would use more of it:
Nope, maybe barely notable difference in graphs at 2.66 and 2.33GHz but game performance got worst with each frequency reduction...

After some Googling about low CPU use in games, i tried setting the game to run at 'High' or 'Real Time' even Priority and 'unparking my CPU cores in Win7' (even though this is not an issue with my CPU generation), but neither of those made any difference at all...

Anybody has any ideas?
How much CPU does your rig use usually in games?
Screenshots (Click to show)
Dirt Rally vs GTA V CPU usage


Dirt rally CPU clock vs CPU Usage


Drive all games are installed



computerbase.de tests results

Thy Old Lady
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon X5460 Asus P5Q3 (P45/ICH10R) Sapphire R9 270X Dual-X 2GB OC Crucial Ballistix Sport 8GB Kit 
Hard DriveHard DriveCoolingCooling
Kingston SSDNow 40GB Westrn Digital WD20PURX Purple 2TB Thermaltake Big Typhoon (GPU) Raijintek Themis (CPU) 
OSMonitorPowerCase
Windows 7 Ultimate x64 LG Flatron IPS235 Corsair TX750W Coolermaser Elite 332 
Other
Cat 
  hide details  
Reply
Thy Old Lady
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon X5460 Asus P5Q3 (P45/ICH10R) Sapphire R9 270X Dual-X 2GB OC Crucial Ballistix Sport 8GB Kit 
Hard DriveHard DriveCoolingCooling
Kingston SSDNow 40GB Westrn Digital WD20PURX Purple 2TB Thermaltake Big Typhoon (GPU) Raijintek Themis (CPU) 
OSMonitorPowerCase
Windows 7 Ultimate x64 LG Flatron IPS235 Corsair TX750W Coolermaser Elite 332 
Other
Cat 
  hide details  
Reply
post #8 of 20
Its not like the days of Sandy Bridge where stock 3.4 ghz and u nca OC to 5Ghz.
6700k is 4.4k and can mostly OC to 4.8
post #9 of 20
Thread Starter 
Quote:
Originally Posted by MuscleBound View Post

Its not like the days of Sandy Bridge where stock 3.4 ghz and u nca OC to 5Ghz.
6700k is 4.4k and can mostly OC to 4.8

Well since even at 4GHz my CPU is still no were close to a stock 3770K as you will see in the AIDA bench below, i know that it would be a poor match for a Fury X. But i don't have a Fury X lol.
I just don't know why most, well all but one now really but i don't have Project Cars installed atm which i could not play with my C2D E8500, of my games don't use more of my CPU like GTA V does...
Is that normal?
Is the 'stuttering' (frames drop) i feel now and then in fixed parts of certain stages in Dirt Rally, caused by the not-max-CPU-usage or its just my 270X can't do much more?

Thy Old Lady
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon X5460 Asus P5Q3 (P45/ICH10R) Sapphire R9 270X Dual-X 2GB OC Crucial Ballistix Sport 8GB Kit 
Hard DriveHard DriveCoolingCooling
Kingston SSDNow 40GB Westrn Digital WD20PURX Purple 2TB Thermaltake Big Typhoon (GPU) Raijintek Themis (CPU) 
OSMonitorPowerCase
Windows 7 Ultimate x64 LG Flatron IPS235 Corsair TX750W Coolermaser Elite 332 
Other
Cat 
  hide details  
Reply
Thy Old Lady
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon X5460 Asus P5Q3 (P45/ICH10R) Sapphire R9 270X Dual-X 2GB OC Crucial Ballistix Sport 8GB Kit 
Hard DriveHard DriveCoolingCooling
Kingston SSDNow 40GB Westrn Digital WD20PURX Purple 2TB Thermaltake Big Typhoon (GPU) Raijintek Themis (CPU) 
OSMonitorPowerCase
Windows 7 Ultimate x64 LG Flatron IPS235 Corsair TX750W Coolermaser Elite 332 
Other
Cat 
  hide details  
Reply
post #10 of 20
Quote:
Originally Posted by MuscleBound View Post

Yeah CPU overclocking is a not much help these days.
GPU is where its all at.

Nah, entirely depends on the usage case. Go play some CPU-bound games and you'll know what I mean wink.gif.
Petrichor
(23 items)
 
Placid
(12 items)
 
Poised
(6 items)
 
CPUMotherboardGraphicsRAM
Intel i5-3570K @ 4.7GHz / 1.28V ASRock Z77 Extreme4 Gainward GTX 1060, 6GB TeamGroup, 8GB, DDR3-1600MHz / 1.5V, 9-9-9-24 1T 
Hard DriveHard DriveHard DriveOptical Drive
Samsung SSD 830 Series, 128GB Seagate Barracuda ST31000524AS, 1TB, 7200RPM Hitachi Deskstar IDK 0S03661, 3TB, 7200RPM LG GH22NS50 
CoolingCoolingCoolingOS
Lamptron FC5 V3 Noctua NH-D15 Noctua NF-A14 FLX (140mm) (x5)  Windows 10 Pro (64-bit) 
MonitorKeyboardPowerCase
Dell UltraSharp U2312HM (x3) Lioncast LK20, TKL, Cherry MX-Blue, DE-Layout EVGA SuperNOVA G3 750 750W Fractal Define R4 Black 
MouseMouse PadAudioAudio
Mionix Castor Akasa Venom XXL Denon PMA 520 AE Dali Zensor 3 (x2) 
AudioAudioAudio
Magnat Monitor Supreme Sub 301 A Audioengine D1 Beyerdynamic DT 770 Pro 250 Ω 
CPUMotherboardRAMHard Drive
Intel Pentium G3440T ASRock E3C226D2I Kingston ValueRAM, 16GB, DDR3-1333MHz, ECC HGST IDK Deskstar H3IK40003272SE, 4TB, 7200RPM 
Hard DriveOSPowerCase
Seagate ST4000DM000, 4TB, 5900RPM (x3) FreeNAS be quiet! Straight Power E9 400W NZXT Source 210 Elite 
CPUGraphicsRAMHard Drive
Intel Core i3-2367M Intel HD Graphics 3000 4GB 500GB 
OSMonitor
Windows 10 Enterprise (64-bit) 11.6" WXGA non-glare LED TFT, 1366x768 
  hide details  
Reply
Petrichor
(23 items)
 
Placid
(12 items)
 
Poised
(6 items)
 
CPUMotherboardGraphicsRAM
Intel i5-3570K @ 4.7GHz / 1.28V ASRock Z77 Extreme4 Gainward GTX 1060, 6GB TeamGroup, 8GB, DDR3-1600MHz / 1.5V, 9-9-9-24 1T 
Hard DriveHard DriveHard DriveOptical Drive
Samsung SSD 830 Series, 128GB Seagate Barracuda ST31000524AS, 1TB, 7200RPM Hitachi Deskstar IDK 0S03661, 3TB, 7200RPM LG GH22NS50 
CoolingCoolingCoolingOS
Lamptron FC5 V3 Noctua NH-D15 Noctua NF-A14 FLX (140mm) (x5)  Windows 10 Pro (64-bit) 
MonitorKeyboardPowerCase
Dell UltraSharp U2312HM (x3) Lioncast LK20, TKL, Cherry MX-Blue, DE-Layout EVGA SuperNOVA G3 750 750W Fractal Define R4 Black 
MouseMouse PadAudioAudio
Mionix Castor Akasa Venom XXL Denon PMA 520 AE Dali Zensor 3 (x2) 
AudioAudioAudio
Magnat Monitor Supreme Sub 301 A Audioengine D1 Beyerdynamic DT 770 Pro 250 Ω 
CPUMotherboardRAMHard Drive
Intel Pentium G3440T ASRock E3C226D2I Kingston ValueRAM, 16GB, DDR3-1333MHz, ECC HGST IDK Deskstar H3IK40003272SE, 4TB, 7200RPM 
Hard DriveOSPowerCase
Seagate ST4000DM000, 4TB, 5900RPM (x3) FreeNAS be quiet! Straight Power E9 400W NZXT Source 210 Elite 
CPUGraphicsRAMHard Drive
Intel Core i3-2367M Intel HD Graphics 3000 4GB 500GB 
OSMonitor
Windows 10 Enterprise (64-bit) 11.6" WXGA non-glare LED TFT, 1366x768 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Graphics Cards - General
Overclock.net › Forums › Graphics Cards › Graphics Cards - General › I don't see any real benefits in gaming performance when i OC