Overclock.net › Forums › Industry News › Hardware News › [AnandTech] Choosing a Gaming CPU at 1440p: Adding in Haswell
New Posts  All Forums:Forum Nav:

[AnandTech] Choosing a Gaming CPU at 1440p: Adding in Haswell - Page 6

post #51 of 90
0.01% represent!!

Fireball
(20 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5 2500K @ 4.7GHz ASUS P8P67 PRO EVGA 780 SC ACX SLI G.Skill 8GB DDR3-1600 
Hard DriveCoolingOSMonitor
Samsung 840 Pro 256GB Noctua NH-U12S | 2x NF-F12 Windows 7 Ultimate x64 Dell U2913WM - 2560x1080 - 75Hz 
KeyboardPowerCaseMouse
Logitech Wireless Illuminated K800 Corsair AX750 Fractal Design Arc Midi R2 Logitech G700 
Mouse PadAudioAudioAudio
Steelseries 4HD Creative Soud Blaster X7 Sennheiser PC 360 Headset Sony SS-MB150H Bookshelf Speakers 
OtherOtherOtherOther
Thrustmaster T.Flight Hotas X Logitech G13 TrackIR 5 / TrackClip Pro Google Nexus 7 
  hide details  
Reply
Fireball
(20 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5 2500K @ 4.7GHz ASUS P8P67 PRO EVGA 780 SC ACX SLI G.Skill 8GB DDR3-1600 
Hard DriveCoolingOSMonitor
Samsung 840 Pro 256GB Noctua NH-U12S | 2x NF-F12 Windows 7 Ultimate x64 Dell U2913WM - 2560x1080 - 75Hz 
KeyboardPowerCaseMouse
Logitech Wireless Illuminated K800 Corsair AX750 Fractal Design Arc Midi R2 Logitech G700 
Mouse PadAudioAudioAudio
Steelseries 4HD Creative Soud Blaster X7 Sennheiser PC 360 Headset Sony SS-MB150H Bookshelf Speakers 
OtherOtherOtherOther
Thrustmaster T.Flight Hotas X Logitech G13 TrackIR 5 / TrackClip Pro Google Nexus 7 
  hide details  
Reply
post #52 of 90
Also, my 5600K was purposefully picked to serve as an HTPC, so it does great there. I just decided to run some benches before I put the HTPC together to see how it'd perform with some proper GPUs.

It ran like crap.

I'm not some spurned lover tongue.gif
zfz.Professional
(16 items)
 
zfz.Prodigy
(9 items)
 
zfz.Bench
(15 items)
 
CPUMotherboardGraphicsRAM
2500K ASUS P8Z77-I Deluxe EVGA GTX 670 8GB Samsung 30nm 
Hard DriveCoolingOSCase
Samsung 830 128GB H60 Win 7 64 Bit BitFenix Spectre Pro 
Audio
NuForce uDAC 2 
CPUMotherboardGraphicsRAM
3570K Gigabyte Z77X-UP7 ASUS RoG Matrix 7970 8GB Samsung DDR3 1600 
Hard DriveCoolingOSMonitor
128GB Crucial M4 Prolimatech Megahalems Black Win 7 64bit BenQ XL2420T 
PowerCase
Corsair AX650 TripT Custom Bench 
  hide details  
Reply
zfz.Professional
(16 items)
 
zfz.Prodigy
(9 items)
 
zfz.Bench
(15 items)
 
CPUMotherboardGraphicsRAM
2500K ASUS P8Z77-I Deluxe EVGA GTX 670 8GB Samsung 30nm 
Hard DriveCoolingOSCase
Samsung 830 128GB H60 Win 7 64 Bit BitFenix Spectre Pro 
Audio
NuForce uDAC 2 
CPUMotherboardGraphicsRAM
3570K Gigabyte Z77X-UP7 ASUS RoG Matrix 7970 8GB Samsung DDR3 1600 
Hard DriveCoolingOSMonitor
128GB Crucial M4 Prolimatech Megahalems Black Win 7 64bit BenQ XL2420T 
PowerCase
Corsair AX650 TripT Custom Bench 
  hide details  
Reply
post #53 of 90
So I wonder how all these people with FX CPU's are not killing themselves with frustration, because their CPU can't play any games. I mean from what i'm gathering from all these debates, the FX CPU is so weak when you boot up a game it just craps itself, right? rolleyes.gif

/Sarcasm
Upgrayedd
(8 items)
 
  
CPUMotherboardGraphicsRAM
Ryzen 1700 @ 3.95GHz 1.387v Asus Prime B350 Plus  RX480 XFX RS 1350/2200 Corsair Vengeance LPX 16GB 2933Mhz cas 14 
Hard DriveOSMonitorPower
Samsung 950 Pro M.2 Windows 10 pro x64 Viotek curved FHD 144Hz EVGA 750w 80 Bronze 
  hide details  
Reply
Upgrayedd
(8 items)
 
  
CPUMotherboardGraphicsRAM
Ryzen 1700 @ 3.95GHz 1.387v Asus Prime B350 Plus  RX480 XFX RS 1350/2200 Corsair Vengeance LPX 16GB 2933Mhz cas 14 
Hard DriveOSMonitorPower
Samsung 950 Pro M.2 Windows 10 pro x64 Viotek curved FHD 144Hz EVGA 750w 80 Bronze 
  hide details  
Reply
post #54 of 90
Thread Starter 
Quote:
Originally Posted by Michalius View Post

Those are Tech Report's benches.

As I said, I think there is a lot of interesting data in the article, but using it as a basis for consumer advice on which CPU to buy is really limited. I think Ian knows this, which is why he included all of the 'yeah I know I'm missing this data, but it's hard' caveats in the introduction.

I normally don't comment on articles like this, as you find bad methodology the norm on places like HardOCP and Tom's, but Anand is one of those few sites that normally has an amazingly high bar for content. Seeing something like this on the front page with a conclusion that isn't backed up by anything substantial at all is a bit disheartening to me.

yes i know those are TR benches i do read the site. but what was tested for what reason under what conditions being unknown just makes them an image. i don't mean to be rude or condescending and i apologize if you are offended. the data albeit not the most desirable but does confirm the conclusion. i think we both know it isn't recommended to make a purchase based on ONE review. but it is better than i have seen from what sites you mentioned among others. as far as using FCAT for a metric; it has been tested and proven to show stuttering issues with GPUs in a multi card circumstance BUT until that gets ironed out; how can one know if any apparent stuttering is caused solely by the CPU?

well if cpu A doesn't stutter but cpu B does, yes you think thats enough but really has that been tested? has the results been shared among colleagues and been replicated by them? and has the original findings been accepted? just because a metric has been shown useful in one instance doesn't mean it applies to all. it would be reasonable but if we demand high standards then lets not take any short cuts. set anal retentiveness on snare drum! (that was a joke - i crack myself up) look how long before fraps was shown to be flawed with a multi gpu setup before it was accepted; alien babel did a series of testing if RAM or SSDs helped prevent stuttering.

but i'll make this suggestion; you can if you want to look at the glass as half empty but i think its half full and thats cool. but i sorta of think it's unfair to an individual when we criticize them for failing to uphold a standard that we set for a group.
Quote:
Originally Posted by Michalius View Post

Also, my 5600K was purposefully picked to serve as an HTPC, so it does great there. I just decided to run some benches before I put the HTPC together to see how it'd perform with some proper GPUs.

It ran like crap.

I'm not some spurned lover tongue.gif

crap sounds like subjective terminology . . .lachen.gif

yeah i did it again . . . . biggrin.gif
loon 3.2
(18 items)
 
  
CPUMotherboardGraphicsRAM
i7-3770K Asus P8Z77-V Pro EVGA 980TI SC+ 16Gb PNY ddr3 1866 
Hard DriveHard DriveHard DriveOptical Drive
PNY 1311 240Gb 1 TB Seagate 3 TB WD Blue DVD DVDRW+/- 
CoolingCoolingOSMonitor
EKWB P280 kit EK-VGA supremacy Win X LG 24MC57HQ-P 
KeyboardPowerCaseMouse
Ducky Zero [blues] EVGA SuperNova 750 G2 Stryker M [hammered and drilled] corsair M65 
AudioAudio
SB Recon3D Klipsch ProMedia 2.1  
  hide details  
Reply
loon 3.2
(18 items)
 
  
CPUMotherboardGraphicsRAM
i7-3770K Asus P8Z77-V Pro EVGA 980TI SC+ 16Gb PNY ddr3 1866 
Hard DriveHard DriveHard DriveOptical Drive
PNY 1311 240Gb 1 TB Seagate 3 TB WD Blue DVD DVDRW+/- 
CoolingCoolingOSMonitor
EKWB P280 kit EK-VGA supremacy Win X LG 24MC57HQ-P 
KeyboardPowerCaseMouse
Ducky Zero [blues] EVGA SuperNova 750 G2 Stryker M [hammered and drilled] corsair M65 
AudioAudio
SB Recon3D Klipsch ProMedia 2.1  
  hide details  
Reply
post #55 of 90
Not offended at all!

I get what you are saying, but, I think when someone says 'frame times', and the initial response is, 'but stuttering...' displays a common misconception about what frame time testing is.

When you measure FPS, the computer is determining how many frames were generated in the last second. That's good and all to get a rough estimate of performance, but when you are talking about giving a single number to represent 30-150+ points of data that were generated in that second, you immediately have to think closely about whether or not this is a good way to measure performance.

Frame time testing doesn't exist to just find stutter, frame time testing exists because it is simply a more accurate way of looking at your data. This is objectively true in every sense, as FPS is nothing more than an average of the very same data over a second. The reason why frame time testing then became associated with stutter, is because since it was a more accurate measurement, for the first time, stutter could be verified and plotted.

That being said, there's a lot of different things you can do with frame time testing. 99th percentile more or less represents a typical 'average' frame rate. You cut off the top 1% of outliers, and look at what the minimum time is to render a frame. If the rest of the benchmark runs smoothly, then often times, 99th percentile matches up really well with average frame rate if you simply covert it. That isn't designed to find stutter, it's essentially a more accurate version of Avg FPS, because they both attempt to convey 'general performance you can expect'.

Ian knows all of this, it's clear from his introduction that he knows it is the most accurate way to measure performance. The issue is that it is incredibly time intensive. I did frame time testing with 6 different games, seeing what the effect of increase core frequency would be on gaming performance, doing steps of 200mhz from 3600 to 4600. It literally took me nearly a month to complete in my free time. This has been a big change, and it's been hard for a lot of people to accept because it's difficult to get all of the data sorted, and it's difficult to understand when the entire community has been entrenched in it's less accurate sibling for so many years.

When I see an article like this that has mostly been done with the better testing (albeit at 1080p and only with a single GPU), and the article at hand has wildly different results than the more accurate test, something seems afoul. This is doubly true when the writer of the article even makes note that this data isn't as accurate, he just couldn't afford the time to do it properly. That's precisely what Ian says at the beginning.

As I said, that's fine. I'm glad he did the tests and presented information. My issue isn't with that. My issue is that with this very same information, which he accepts as being not quite in depth enough, is the basis for consumer advice on which CPU to buy.

I do not know what the internal structure of Anand is, and I will not say that this is the case for certain. But, I can promise you an article with a title and a conclusion about 'Investigating Average Frame Rates at 1440p' will get far fewer clicks than what they put up.
zfz.Professional
(16 items)
 
zfz.Prodigy
(9 items)
 
zfz.Bench
(15 items)
 
CPUMotherboardGraphicsRAM
2500K ASUS P8Z77-I Deluxe EVGA GTX 670 8GB Samsung 30nm 
Hard DriveCoolingOSCase
Samsung 830 128GB H60 Win 7 64 Bit BitFenix Spectre Pro 
Audio
NuForce uDAC 2 
CPUMotherboardGraphicsRAM
3570K Gigabyte Z77X-UP7 ASUS RoG Matrix 7970 8GB Samsung DDR3 1600 
Hard DriveCoolingOSMonitor
128GB Crucial M4 Prolimatech Megahalems Black Win 7 64bit BenQ XL2420T 
PowerCase
Corsair AX650 TripT Custom Bench 
  hide details  
Reply
zfz.Professional
(16 items)
 
zfz.Prodigy
(9 items)
 
zfz.Bench
(15 items)
 
CPUMotherboardGraphicsRAM
2500K ASUS P8Z77-I Deluxe EVGA GTX 670 8GB Samsung 30nm 
Hard DriveCoolingOSCase
Samsung 830 128GB H60 Win 7 64 Bit BitFenix Spectre Pro 
Audio
NuForce uDAC 2 
CPUMotherboardGraphicsRAM
3570K Gigabyte Z77X-UP7 ASUS RoG Matrix 7970 8GB Samsung DDR3 1600 
Hard DriveCoolingOSMonitor
128GB Crucial M4 Prolimatech Megahalems Black Win 7 64bit BenQ XL2420T 
PowerCase
Corsair AX650 TripT Custom Bench 
  hide details  
Reply
post #56 of 90
Quote:
Originally Posted by mav451 View Post

To be fair, how many noobs asking for advice are pairing their build with a 1440P (or higher) res monitor? tongue.gif
I'd wager they are on a low resolution (1080P or 1440x900).

It's the same at 1080p, I didn't really notice the performance improvement from going to my i5 (From a FX-4170) and I game at 1080p.

As for 1440x900 and below...Well, you're not likely to see the difference in performance unless you've managed to have a refresh rate of like 240Hz or something as FPS will already be plenty high even if you have a $50 CPU.
Quote:
Originally Posted by Michalius View Post

You're right, and that's definitely the upshot of this article.

However, with caveats like this:
Followed by this gross exaggeration:
Anyone who has ever done benching knows you should run each test at least 5 times, and then either average the findings, or use *all* of it.

For the lack of proper performance measuring (as in, Frame Time) we are presented with this:
You say FCAT is too hard, and then discount Fraps, which has been proven to be plenty accurate in terms of reporting frame times on single cards. Even when it's off, it's still far more accurate than any FPS measurement, which is simply an average of frame times over a second.

So we have the writer admitting that this article is missing the most accurate measurements of performance. But then we get conclusions like this:
That can be literally translated as, "do not mind performing horribly at video game that is using two or fewer threads, and even BF3 multiplayer".

How can this sort of conclusion be drawn when the most important gaming data is completely missing?

Here's how the recommended 5600K performs with a single GPU in a typical CPU demanding game.

This is my point. You can't base conclusions on poor testing. I get the benefits of this article, especially comparing such a wide range of CPUs. But to acknowledge that the test is severely lacking due to time constraints, exaggerate on the difficulty of doing multiple runs to gather accurate data, and then finish the article with a conclusion on which processor to buy in a single GPU configuration that performs like absolute garbage (trust me, I know, I have a 5600K in my HTPC) is borderline offensive to me.

This should be relegated to a community amateur's benchmarks for some interesting data. It is precisely that. It's not an article that is worthy of actually basing consumer advice on. Doing so is disingenuous to the reader, and at odds with the whole point of journalism and consumer advice.

Anand, how did you guys okay this?

FCAT shows the differences, but it also grossly exaggerates them...I play the hell out of Skyrim, I did on my FX-4170 and I do on my i5, there's a slight difference between a 4.2Ghz FX and a 4.5Ghz i5 in Whiterun or other busy towns, and in extremely large battles (ie. Me playing with the console and spawning dragons everywhere) but for easily 95% of the time I'm playing? No difference.

The fact of the matter is virtually every game is GPU limited all of the time, some are CPU limited at times (eg. Skyrim, Civ V as I bet those low FPS rates come from when it's calculating turns, when you're doing more than waiting for it to finish it's a smooth 60fps all around in DX11 mode for me on my rig, even when I had the FX) and a handful are CPU limited most of the time (eg. Sins of a Solar Empire, SC2)

Also, how does the 5600K perform like garbage? It's faster than a lot of CPUs that most people would consider perfectly fine for gaming, maybe it's a bit slower than Intel (~55% so if you're talking equal clocks using only one, and precisely one thread, less so as you add more threads and compare at stock clocks, etc) but it's certainly not garbage.
Quote:
Originally Posted by Artikbot View Post

Well, at 1440p with a HD7970, there simply is no difference between Intel and AMD, so... Yeah.

Even with two HD7970s the difference seems pretty small, and I can vouch for 1080p being enough to cause a GPU bottleneck quite often, although not as much as 1440p.
Quote:
Originally Posted by ZealotKi11er View Post

CPU is still important and we need faster CPU for games. The thing about CPU is that you cant test them suing GPU limitation games. Grab 2 x Titan OC @ 1080p no AA and test CPUs. I know in dome games CPU matter a lot. In BF3 MP for example. Yeah it might be 80fps vs 100fps but still one CPU is faster then the other. Also for those that want 120Hz the CPU need to be top of the time to get close to 120 fps.

If you're buying a 120Hz monitor I have no idea why you're even considering AMD.

As for the rest..I noticed a significant performance increase in BF3 MP going from a stock Core 2 Duo E6700 to an Phenom II x3 720 @ 3.2Ghz, not so much from the Phenom II to a FX-4170 clocked at 4.2Ghz and another tiny boost from going to my i5 when I did a quick test not too long ago...It is to a point but honestly, as long as your CPU was launched in or after 2010 and is a Quad Core or dual with HT then you're fine for CPU upgrades for now.

And it doesn't matter when it's 80fps vs 100fps, plus, the FX is faster elsewhere and starting to show that speed in games now...Crysis 3 is slightly (Still unnoticeable) faster on an FX-83*0 than the i5 and faster still (By 20fps, definitely noticeable) on an i7 for example.
Quote:
Originally Posted by Michalius View Post

Pretty much.

You basically have to clock your processor at 4.5 and have something with really great single threaded performance to maintain low frame times in just about any UE3 game.

I still wonder just how noticeable the frame-times are though, since I can't tell the difference with my i5 OCed and at stock in any game, even the CPU limited ones; they still end up lagging around the same time, probably with higher FPS but it's still a stuttery mess.
That's including UE3 games which have been all GPU limited in my experience.
Quote:
Originally Posted by Dimaggio1103 View Post

So I wonder how all these people with FX CPU's are not killing themselves with frustration, because their CPU can't play any games. I mean from what i'm gathering from all these debates, the FX CPU is so weak when you boot up a game it just craps itself, right? rolleyes.gif

/Sarcasm

This is why I think frame-time testing should be done alongside FPS testing, because the fact of the matter is that some people don't notice it...Look at the CFX FCAT stuff, if you believe a lot of those it says CFX is worse than single-cards despite giving higher FPS, yet there's a lot of people who didn't notice it.
    
CPUMotherboardGraphicsRAM
Intel Core i5 3570k @ 4.5Ghz ASRock Z77 Pro3 Powercolor Radeon HD7950 3GB @ 1150/1350 4x4GB G.Skill Ares 2000Mhz CL9 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 250GB Western Digital Black 1TB WD1002FAEX Seagate Barracuda 3TB ST3000DM001 Samsung Spinpoint EcoGreen 2TB 
Optical DriveCoolingCoolingCooling
Pioneer DVR-220LBKS Noctua NH-D14 Scythe Gentle Typhoon 1850rpm Corsair AF140 Quiet Edition 
CoolingOSMonitorMonitor
Arcitc Cooling Acclero Twin Turbo II Arch Linux x86-64, amdgpu BenQ G2220HD BenQ G2020HD 
KeyboardPowerCaseMouse
Ducky Shine III Year of the Snake, Cherry Blue Silverstone Strider Plus 600w CoolerMaster CM690 II Black and White SteelSeries Sensei Professional 
Mouse PadAudioOther
Artisan Hien Mid Japan Black Large ASUS Xonar DX NZXT Sentry Mesh 30w Fan Controller 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel Core i5 3570k @ 4.5Ghz ASRock Z77 Pro3 Powercolor Radeon HD7950 3GB @ 1150/1350 4x4GB G.Skill Ares 2000Mhz CL9 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 250GB Western Digital Black 1TB WD1002FAEX Seagate Barracuda 3TB ST3000DM001 Samsung Spinpoint EcoGreen 2TB 
Optical DriveCoolingCoolingCooling
Pioneer DVR-220LBKS Noctua NH-D14 Scythe Gentle Typhoon 1850rpm Corsair AF140 Quiet Edition 
CoolingOSMonitorMonitor
Arcitc Cooling Acclero Twin Turbo II Arch Linux x86-64, amdgpu BenQ G2220HD BenQ G2020HD 
KeyboardPowerCaseMouse
Ducky Shine III Year of the Snake, Cherry Blue Silverstone Strider Plus 600w CoolerMaster CM690 II Black and White SteelSeries Sensei Professional 
Mouse PadAudioOther
Artisan Hien Mid Japan Black Large ASUS Xonar DX NZXT Sentry Mesh 30w Fan Controller 
  hide details  
Reply
post #57 of 90
Quote:
Originally Posted by Artikbot View Post

Hm, there is a small benefit, but nothing too radical... Seems to be the clockspeed difference more than anything.


Also sidenote, lol at Piledriver roflstomping Phenom.

Dang my system got old...

Piledriver got better memory and its stock for stock (way higher clocks)
Cougartown
(19 items)
 
  
CPUMotherboardGraphicsRAM
i7 2600k asrock extreme 4 z77 powercooler r9 290 Crucial Technology  
RAMHard DriveHard DriveOptical Drive
Crucial Technology  Ocz agility 3 240gb Samsung Spinpoint F4EG 2 TB SATA2 5400rpm 32 MB... Light-on drive 
CoolingOSMonitorKeyboard
Coolermaster Hyper 212+ Windows 8.1 Seiki 39inch 4k television rosewill rk-9000 with cherry blues 
PowerCaseMouseMouse Pad
CORSAIR Gaming Series GS700 700W ATX12V v2.3 80... Cougar Evolution Goldensunksy gaming mouse 3 dollar mouse pad from target 
Audio
TV audio 
  hide details  
Reply
Cougartown
(19 items)
 
  
CPUMotherboardGraphicsRAM
i7 2600k asrock extreme 4 z77 powercooler r9 290 Crucial Technology  
RAMHard DriveHard DriveOptical Drive
Crucial Technology  Ocz agility 3 240gb Samsung Spinpoint F4EG 2 TB SATA2 5400rpm 32 MB... Light-on drive 
CoolingOSMonitorKeyboard
Coolermaster Hyper 212+ Windows 8.1 Seiki 39inch 4k television rosewill rk-9000 with cherry blues 
PowerCaseMouseMouse Pad
CORSAIR Gaming Series GS700 700W ATX12V v2.3 80... Cougar Evolution Goldensunksy gaming mouse 3 dollar mouse pad from target 
Audio
TV audio 
  hide details  
Reply
post #58 of 90
Min fps would be more interesting then average frankly.
post #59 of 90
Quote:
Originally Posted by Brutuz View Post

This is why I think frame-time testing should be done alongside FPS testing, because the fact of the matter is that some people don't notice it...Look at the CFX FCAT stuff, if you believe a lot of those it says CFX is worse than single-cards despite giving higher FPS, yet there's a lot of people who didn't notice it.

Yeah I think I would prefer to see the full kitchen sink tbh haha, but maybe I'm really too demanding.
FCAT and 99th percentile frame times, median (important) and minimum FPS, then mean FPS as a last resort. I like how TR has separate graphs for 50ms/33/16...but frankly the only one that matters is 16ms or faster (say 8ms) as that represents 60fps and 120fps respectively.
Main
(20 items)
 
  
CPUMotherboardGraphicsRAM
i5 4670K @ 4.7Ghz [1.284v] Z87X-UD4H [F7] MSI GTX 1070 Gaming X 2x4GB Samsung MV-3V4G3; 10-10-10-28 @ 2133Mhz [... 
Hard DriveHard DriveHard DriveHard Drive
1x Samsung 850 Pro 256GB 1x Crucial M500 960GB 1x WD4003FZEX 1x WD30EFRX 
CoolingCoolingOSMonitor
NH-D14 3x A15s @ 600RPM 2x Phanteks F140SP BBK (front), SFF21E (bottom) Win 10 Pro x64 Catleap 2B @ 119hz +1 
MonitorPowerCaseMouse
U3014 eVGA 750G2 Fractal R5 - Blackout Edition MS WMO 1.1a 
Mouse PadAudio
fUnc 1030 Creative Sound Blaster Z 
  hide details  
Reply
Main
(20 items)
 
  
CPUMotherboardGraphicsRAM
i5 4670K @ 4.7Ghz [1.284v] Z87X-UD4H [F7] MSI GTX 1070 Gaming X 2x4GB Samsung MV-3V4G3; 10-10-10-28 @ 2133Mhz [... 
Hard DriveHard DriveHard DriveHard Drive
1x Samsung 850 Pro 256GB 1x Crucial M500 960GB 1x WD4003FZEX 1x WD30EFRX 
CoolingCoolingOSMonitor
NH-D14 3x A15s @ 600RPM 2x Phanteks F140SP BBK (front), SFF21E (bottom) Win 10 Pro x64 Catleap 2B @ 119hz +1 
MonitorPowerCaseMouse
U3014 eVGA 750G2 Fractal R5 - Blackout Edition MS WMO 1.1a 
Mouse PadAudio
fUnc 1030 Creative Sound Blaster Z 
  hide details  
Reply
post #60 of 90
Thread Starter 
i had a thought and (before it died of loneliness) remembered reading about what and where fraps takes measurements from. so i'm going to toss this into using FRAPS for measuring the CPU performance:
AMD Comments on GPU Stuttering, Offers Driver Roadmap & Perspective on Benchmarking
Quote:
To get to that problem, we’re going to have to take a look at how FRAPS measures framerates. Going back to our diagram of the rendering pipeline, FRAPS hooks into the pipeline very early, at the application stage.

By injecting its DLL into the application, FRAPS then serves to intercept the Direct3D Present call as it’s being made to Direct3D. From here FRAPS can then delay the call for a split second to insert the draw commands to draw its overlay, or FRAPS can simply move on. When it comes to measuring framerates and frametimes what FRAPS is doing is to measure the Present calls. Every time it sees a new present call get pushed out, it counts that as a new frame, does any necessary logging, and then passes that Present call on to Direct3D.
This method is easy to accomplish and works with almost any application, which is what makes FRAPS so versatile. When it comes to measuring the average FPS over a benchmark run for example, FRAPS is great because every Present call it sees will eventually end up triggering a frame to be displayed. The average framerate is merely the number of Present calls FRAPS sees, divided by how long FRAPS was running for.

The problem here is not in using FRAPS to measure average framerates over the run of a benchmark, but rather when it comes to using FRAPS to measure individual frames. FRAPS is at the very start of the rendering pipeline; it’s before the GPU, it’s before the drivers, it’s even before Direct3D and the context queue. As such FRAPS can tell you all about what goes into the rendering pipeline, but FRAPS cannot tell you what comes out of the rendering pipeline.
and a bit more concise:
FCAT: The Evolution of Frame Interval Benchmarking, Part 1
Quote:
Direct3D (and OpenGL) uses a complex rendering pipeline that spans several different mechanisms and stages. When a frame is generated by an application, it must travel through the pipeline to Direct3D, the video drivers, a frame queue (the context queue), a GPU scheduler, the video drivers again, the GPU, and finally after that a frame can be displayed. The pipeline analogy is used here because that’s exactly what it is, with the added complexity of the context queue sitting in the middle of that pipeline.

FRAPS for its part exists at almost the very beginning of this pipeline. It interfaces with individual applications and intercepts the Present calls made to Direct3D that mark the end of each frame. By counting Present calls FRAPS can easily tell how many frames have gone into the pipeline, making it a simple and effective tool for measuring average framerates.

so where is the best point to measure the performance of a CPU? after the application and before it goes through the GPU pipeline? or after it is displayed on the screen? where are there less potential variables involved? though the same author, who is different than this article, wrote the same two articles i am referring to, it does point out FRAPS is absolutely acceptable to measure frame rates. it appears the cpu may have little or no control over how the gpu performs once the present call is passed down through the graphics pipeline; it goes around a carousel with the 3d runtime and driver, a context queue and then a scheduler to the driver.

but hey, i could be wrong . . . .
loon 3.2
(18 items)
 
  
CPUMotherboardGraphicsRAM
i7-3770K Asus P8Z77-V Pro EVGA 980TI SC+ 16Gb PNY ddr3 1866 
Hard DriveHard DriveHard DriveOptical Drive
PNY 1311 240Gb 1 TB Seagate 3 TB WD Blue DVD DVDRW+/- 
CoolingCoolingOSMonitor
EKWB P280 kit EK-VGA supremacy Win X LG 24MC57HQ-P 
KeyboardPowerCaseMouse
Ducky Zero [blues] EVGA SuperNova 750 G2 Stryker M [hammered and drilled] corsair M65 
AudioAudio
SB Recon3D Klipsch ProMedia 2.1  
  hide details  
Reply
loon 3.2
(18 items)
 
  
CPUMotherboardGraphicsRAM
i7-3770K Asus P8Z77-V Pro EVGA 980TI SC+ 16Gb PNY ddr3 1866 
Hard DriveHard DriveHard DriveOptical Drive
PNY 1311 240Gb 1 TB Seagate 3 TB WD Blue DVD DVDRW+/- 
CoolingCoolingOSMonitor
EKWB P280 kit EK-VGA supremacy Win X LG 24MC57HQ-P 
KeyboardPowerCaseMouse
Ducky Zero [blues] EVGA SuperNova 750 G2 Stryker M [hammered and drilled] corsair M65 
AudioAudio
SB Recon3D Klipsch ProMedia 2.1  
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [AnandTech] Choosing a Gaming CPU at 1440p: Adding in Haswell