Overclock.net › Forums › Industry News › Hardware News › [pclab] More Crysis 3 CPU benchmarks
New Posts  All Forums:Forum Nav:

[pclab] More Crysis 3 CPU benchmarks - Page 17

post #161 of 243
Quote:
Originally Posted by Flamingo View Post

If you mean checking in realtime, you can use the OSD from MSI's afterburner software.

Also I believe hwinfo has a logging feature. Aida64 too, but thats not free (14 day trial at most)

Allright thanks, the logging feature on HWINFO logs a ton of stuff, but if I reset the readings just before the level start and take a print as soon as I tab out it should give me a good idea of what's going on... Either I'm going about this completely wrong or my numbers are waaaay off from another hexacore bench I've seen.


Overall usage peaked at 34%, with an avg of 25%. It does show, however some cores going well up to the 90's at some points, and the other thing that makes me curious is that I can see usage across all the cpus/hts, completely different from:


Either I'm reading my stats wrong or I dunno, the lower usage per core can be explained with my overclock (comparing to that second image), but according to that graph, only 2 HT's are being (barely) used. Besides crysis I only had chrome / skypechat and hwinfo opened.
Warchief
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 3930k @4.5 ghz - 1.280v Asus Sabertooth X79 EVGA GTX 980 ACX 2.0 Corsair Vengeance 16gb 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 128GB Seagate 31000528AS 1tb LG DVDRW Corsair H80 
OSMonitorKeyboardPower
Windows 8.1 Pro Dell S2340L 23" Corsair K50 Sentey Golden Steel Power 850W 
CaseMouse
Corsair Carbide 500R Corsair M95 
  hide details  
Reply
Warchief
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 3930k @4.5 ghz - 1.280v Asus Sabertooth X79 EVGA GTX 980 ACX 2.0 Corsair Vengeance 16gb 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 128GB Seagate 31000528AS 1tb LG DVDRW Corsair H80 
OSMonitorKeyboardPower
Windows 8.1 Pro Dell S2340L 23" Corsair K50 Sentey Golden Steel Power 850W 
CaseMouse
Corsair Carbide 500R Corsair M95 
  hide details  
Reply
post #162 of 243
Quote:
Originally Posted by DMHernandez View Post


Overall usage peaked at 34%, with an avg of 25%. It does show, however some cores going well up to the 90's at some points, and the other thing that makes me curious is that I can see usage across all the cpus/hts, completely different from:

When I see stats like that ... 25% average, 34% max total usage ... if it's an 8 core system, I would tend to conclude that (unless you are simply never getting a CPU BN ... you can fix that by running the game at a super-low resolution wink.gif), that we're likely dealing with 2 core-coded game, with some usage on a 3rd ... assuming you have a quad with HT ... regardless of how many cores *appear* to be getting some action.

Honestly, unless every single core is pegged out (which would tell you something useful) and/or your overall usage is close to 100%, the % usage-per-core numbers that any of these programs provide actually mean next to nothing is the truth.

It's not a real-time, bare-metal measurement of the 'CPU' in it's entirety, it's a software measurement that shows you what Windows would like to be doing with the load. It doesn't mean the chips micro-code is going to 'agree' ... at all, nor does it 'account' for certain types of bottlenecks within the chip that can happen as it interacts with the particular code it's processing.

I've seen instances where the 'usage' in such graphs was LOWER when I knew for a fact that the game was being CPU bottlenecked (such as in Skyrim for the first 3 months of it's existence ... places like the stairs in Whiterun) via benchmarking at various different clocks (and seeing near a 1:1 ratio between FPS increase and clock speed increase ... which is the proper means of BN testing), and then HIGHER when the game wasn't being CPU BN'd.

Maybe I misread the complaints about Crysis 3, I was under the impression people weren't getting 99% GPU usage on a regular basis in this game, but maybe I misunderstood the complaints. Perhaps I shouldn't have included it as an example above when I wasn't entirely certain wink.gif

But BF3, for SURE, uses lots of cores ... and people complain constantly on the boards about NOT getting full GPU usage anyways.

Coding for multiples cores is very tricky for software like a game where the situation constantly changes due to user input, and 'outcomes' aren't known very far in advance. In contrast to, say, transcoding a video, where the entire task can be assessed at the start and farmed out to a bunch of cores without concern for dependencies and the timing of each threads completion.

On the whole, I remain unconvinced so far that there's actually much tangible benefit to games using lots of our cores, just because they're there wink.gif

But then again, I've also seen the problem of low gpu usage crop up when only 1 or 2 cores was being used, so ... who knows. One problem I'm convinced exists is that cpu's have only gotten like 40% faster since i7 came out 4 years ago ... and GPU's are like 400% faster with the Titan vs. GTX280 and HD4870 that were tops at that time.
Edited by brettjv - 2/25/13 at 11:04pm
    
CPUMotherboardGraphicsRAM
xeon X5675 6-core @ 4.1ghz (1.29v, 20x205 +ht ) rampage iii extreme msi rx470 gaming X (the $159 budget king) 3 x 2gb corsair xms3 pc12800 (9-9-9-24-1T@1600MHz) 
Hard DriveOptical DriveCoolingOS
hynix 250gb ssd (boot), 2tb deskstar (apps),1tb... plextor px-712sa - still the best optical drive... corsair h8o v2 aio W10 home 
MonitorPowerCaseAudio
asus vw266h 25.5" (1920x1200) abs sl (enermax revolution) * single 70A rail 850w silverstone rv-03 XFi Titanium 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
xeon X5675 6-core @ 4.1ghz (1.29v, 20x205 +ht ) rampage iii extreme msi rx470 gaming X (the $159 budget king) 3 x 2gb corsair xms3 pc12800 (9-9-9-24-1T@1600MHz) 
Hard DriveOptical DriveCoolingOS
hynix 250gb ssd (boot), 2tb deskstar (apps),1tb... plextor px-712sa - still the best optical drive... corsair h8o v2 aio W10 home 
MonitorPowerCaseAudio
asus vw266h 25.5" (1920x1200) abs sl (enermax revolution) * single 70A rail 850w silverstone rv-03 XFi Titanium 
  hide details  
Reply
post #163 of 243
Quote:
Originally Posted by Tsumi View Post

Thing is, the single decoder per module is such a massive bottleneck that it may as well be a single 128 bit FPU instead of two. And I'm pretty sure it doesn't help with the 256 bit throughput as well.

Steamroller will get two decoders that feed into the FPU.

http://en.wikipedia.org/wiki/AMD_Bulldozer

Two symmetrical 128-bit FMAC (fused multiply–add capability) floating-point pipelines per module that can be unified into one large 256-bit-wide unit if one of integer cores dispatch AVX instruction and two symmetrical x87/MMX/SSE capable FPPs for backward compatibility with SSE2 non-optimized software
    
CPUMotherboardGraphicsRAM
The Intel Core i3-4150 @ 3.5 GHz Asus B85M-G Rev 1.01, Bios: 2501 AMD Radeon HD 7850 ♥ 1GB 2x 4GB DDR3 1600 MHz CL9 
Hard DriveHard DriveHard DriveOptical Drive
Samsung 750 EVO 250GB Seagate Barracuda 1TB 7200.14 Seagate 500 GB 2.5" Samsung DVD/RW 
CoolingOSMonitorKeyboard
Corsair H70 Windows 10 64 bit Samsung A300N 20" 1600 x 900 60Hz 5ms 19Watt PS/2 Microsoft Wired Keyboard 500 
PowerCaseMouse
Corsair TX850 V2 CoolerMaster Elite 430 Black Logitech M170 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
The Intel Core i3-4150 @ 3.5 GHz Asus B85M-G Rev 1.01, Bios: 2501 AMD Radeon HD 7850 ♥ 1GB 2x 4GB DDR3 1600 MHz CL9 
Hard DriveHard DriveHard DriveOptical Drive
Samsung 750 EVO 250GB Seagate Barracuda 1TB 7200.14 Seagate 500 GB 2.5" Samsung DVD/RW 
CoolingOSMonitorKeyboard
Corsair H70 Windows 10 64 bit Samsung A300N 20" 1600 x 900 60Hz 5ms 19Watt PS/2 Microsoft Wired Keyboard 500 
PowerCaseMouse
Corsair TX850 V2 CoolerMaster Elite 430 Black Logitech M170 
  hide details  
Reply
post #164 of 243
Quote:
Originally Posted by Brutuz View Post

As graphics get harder and harder to noticably increase, what's one area that's easy to increase? Oh. Physics. Or AI. Both of which are heavily reliant on CPU Physics unless you try to give the bottlenecked part more work.

I feel like we've been waiting forever for the "improved AI" that is just around the corner. Same with physics. Even if we do get improved AI and physics, how well can that be multi-threaded? The AI might be able to run on a thread, but all of that is still going to have to wait on the thread getting user input, at least to some extent. I think there is a practical limit on how many cores a game can really use, and I doubt the number is really as high as 8. Even the OP tests show the 4 thread 3570 hanging with the 8 thread chips.
post #165 of 243
Can my system handle this game???? .....sig rig....
XELI
(13 items)
 
  
CPUMotherboardGraphicsRAM
i7 5820k ASUS X99-A GTX980 GeForce Gigabyte 4GB WindForce 3X OC 8GB G.SKILL 
Hard DriveOptical DriveOSMonitor
1x 4TB Seagate & 1x 3TB Seagate & 1x Samsung 85... 1x DVDRW & WIN 7 X64 ULTIMATE ENG LG W2261V-PF 22'' TFT BLACK 
KeyboardPowerCaseMouse
Logitech Ex110 (Wireless) ENERMAX 850W COOLERMASTER RC-832-KKN2-GP STACKER 832 SE EVO Logitech Ex110 (Wireless) 
Mouse Pad
No 
  hide details  
Reply
XELI
(13 items)
 
  
CPUMotherboardGraphicsRAM
i7 5820k ASUS X99-A GTX980 GeForce Gigabyte 4GB WindForce 3X OC 8GB G.SKILL 
Hard DriveOptical DriveOSMonitor
1x 4TB Seagate & 1x 3TB Seagate & 1x Samsung 85... 1x DVDRW & WIN 7 X64 ULTIMATE ENG LG W2261V-PF 22'' TFT BLACK 
KeyboardPowerCaseMouse
Logitech Ex110 (Wireless) ENERMAX 850W COOLERMASTER RC-832-KKN2-GP STACKER 832 SE EVO Logitech Ex110 (Wireless) 
Mouse Pad
No 
  hide details  
Reply
post #166 of 243
Quote:
Originally Posted by TRELOXELO View Post

Can my system handle this game???? .....sig rig....

Yes not all maxed though.
Intel Killer
(14 items)
 
  
CPUMotherboardGraphicsRAM
AMD 8320 4.4ghz Gigabyte 990fx XFX R9 290 (1100/1250) 8GB DDR3 1866 
Hard DriveCoolingOSMonitor
750GB Momentus XT Corsair H80i Windows 8 3x 23" 1080p 
KeyboardPower
Logitech Mk 710 OCZ 1000 watts 
  hide details  
Reply
Intel Killer
(14 items)
 
  
CPUMotherboardGraphicsRAM
AMD 8320 4.4ghz Gigabyte 990fx XFX R9 290 (1100/1250) 8GB DDR3 1866 
Hard DriveCoolingOSMonitor
750GB Momentus XT Corsair H80i Windows 8 3x 23" 1080p 
KeyboardPower
Logitech Mk 710 OCZ 1000 watts 
  hide details  
Reply
post #167 of 243
Quote:
Originally Posted by sumitlian View Post

http://en.wikipedia.org/wiki/AMD_Bulldozer

Two symmetrical 128-bit FMAC (fused multiply–add capability) floating-point pipelines per module that can be unified into one large 256-bit-wide unit if one of integer cores dispatch AVX instruction and two symmetrical x87/MMX/SSE capable FPPs for backward compatibility with SSE2 non-optimized software

How does that refute anything I said? I just said that the single front end decoder is a massive bottleneck.
Millenium Falcon
(24 items)
 
  
CPUMotherboardGraphicsRAM
Core i7 4930k MSI Big Bang Xpower II EVGA GTX 690 Patriot Viper II Sector 7 
Hard DriveHard DriveHard DriveHard Drive
OCZ Deneva 2 Corsair Force 3 Maxtor Western Digital Green 
Optical DriveCoolingCoolingCooling
Samsung BD/DVD-RW Swiftech MCP655 x2 Black Ice GTX 480 Black Ice GTX 280 
CoolingCoolingCoolingCooling
Alphacool Repack Dual D5 Watercool Heatkiller 3.0 Alphacool GTX 690 fullcover Bitspower Big Bang Xpower II fullcover 
OSMonitorKeyboardPower
Windows 8.1 64-bit Professional 3x Dell S2340 Max Keyboard Durandal CoolerMaster V1000 
CaseMouseMouse PadAudio
Azza Genesis 9000B Logitech G700 Roccat Alumic Onkyo HT-S9100THX 
  hide details  
Reply
Millenium Falcon
(24 items)
 
  
CPUMotherboardGraphicsRAM
Core i7 4930k MSI Big Bang Xpower II EVGA GTX 690 Patriot Viper II Sector 7 
Hard DriveHard DriveHard DriveHard Drive
OCZ Deneva 2 Corsair Force 3 Maxtor Western Digital Green 
Optical DriveCoolingCoolingCooling
Samsung BD/DVD-RW Swiftech MCP655 x2 Black Ice GTX 480 Black Ice GTX 280 
CoolingCoolingCoolingCooling
Alphacool Repack Dual D5 Watercool Heatkiller 3.0 Alphacool GTX 690 fullcover Bitspower Big Bang Xpower II fullcover 
OSMonitorKeyboardPower
Windows 8.1 64-bit Professional 3x Dell S2340 Max Keyboard Durandal CoolerMaster V1000 
CaseMouseMouse PadAudio
Azza Genesis 9000B Logitech G700 Roccat Alumic Onkyo HT-S9100THX 
  hide details  
Reply
post #168 of 243
Quote:
Originally Posted by Tsumi View Post

How does that refute anything I said? I just said that the single front end decoder is a massive bottleneck.

Oh sorry ! It was my bad doh.gif I reread the whole article in Tomshw, you were right. smile.gif Stemroller will fix these issues.
Edited by sumitlian - 2/26/13 at 1:52am
    
CPUMotherboardGraphicsRAM
The Intel Core i3-4150 @ 3.5 GHz Asus B85M-G Rev 1.01, Bios: 2501 AMD Radeon HD 7850 ♥ 1GB 2x 4GB DDR3 1600 MHz CL9 
Hard DriveHard DriveHard DriveOptical Drive
Samsung 750 EVO 250GB Seagate Barracuda 1TB 7200.14 Seagate 500 GB 2.5" Samsung DVD/RW 
CoolingOSMonitorKeyboard
Corsair H70 Windows 10 64 bit Samsung A300N 20" 1600 x 900 60Hz 5ms 19Watt PS/2 Microsoft Wired Keyboard 500 
PowerCaseMouse
Corsair TX850 V2 CoolerMaster Elite 430 Black Logitech M170 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
The Intel Core i3-4150 @ 3.5 GHz Asus B85M-G Rev 1.01, Bios: 2501 AMD Radeon HD 7850 ♥ 1GB 2x 4GB DDR3 1600 MHz CL9 
Hard DriveHard DriveHard DriveOptical Drive
Samsung 750 EVO 250GB Seagate Barracuda 1TB 7200.14 Seagate 500 GB 2.5" Samsung DVD/RW 
CoolingOSMonitorKeyboard
Corsair H70 Windows 10 64 bit Samsung A300N 20" 1600 x 900 60Hz 5ms 19Watt PS/2 Microsoft Wired Keyboard 500 
PowerCaseMouse
Corsair TX850 V2 CoolerMaster Elite 430 Black Logitech M170 
  hide details  
Reply
post #169 of 243
I did a little benching of my own and tried to use google docs to make a chart using Hwinfo's .csv file, I closed all visible apps and some processes but I do feel that I'd need a program that can monitor one specific application to be more precise, this is what I came up with (first time doing this kind of bench, so go easy tongue.gif)


(couldn't get googledoc's interactive chart to work on the forum so let's go with the .png)
Warchief
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 3930k @4.5 ghz - 1.280v Asus Sabertooth X79 EVGA GTX 980 ACX 2.0 Corsair Vengeance 16gb 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 128GB Seagate 31000528AS 1tb LG DVDRW Corsair H80 
OSMonitorKeyboardPower
Windows 8.1 Pro Dell S2340L 23" Corsair K50 Sentey Golden Steel Power 850W 
CaseMouse
Corsair Carbide 500R Corsair M95 
  hide details  
Reply
Warchief
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 3930k @4.5 ghz - 1.280v Asus Sabertooth X79 EVGA GTX 980 ACX 2.0 Corsair Vengeance 16gb 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 128GB Seagate 31000528AS 1tb LG DVDRW Corsair H80 
OSMonitorKeyboardPower
Windows 8.1 Pro Dell S2340L 23" Corsair K50 Sentey Golden Steel Power 850W 
CaseMouse
Corsair Carbide 500R Corsair M95 
  hide details  
Reply
post #170 of 243
Quote:
Originally Posted by Brutuz View Post

There are benchmarks that show sometimes running at low resolution gives you completely different results than high resolution (eg. That one before that showed an i3 beating an FX-6300 at 1366x768, but losing by far at 1080p which I now will see if I can find it and post when I get home) as well, you're yet to give me any good reason for it at all apart from bleating the same disproven "It shows CPU performance well!" crap over and over, running Starcraft II will show CPU performance better, running Sins of a Solar Empire Rebellion will show CPU performance better, running virtually any game with 3 GTX 680s will show CPU performance better...Why is it better? Because you're matching what the users will have it set at home.




From hardware cannucks. That's the one you are after i think. And that's the issue with benching while turning everything to low ,supposedly to illustrate cpu weakness.If HC had only tested at that resoution people would assume that since the i3 is handily beating the FX-6300 there, at 1080 it will be the same (since only the vga is affected) or at best GPU tied. What you get is indeed a GPU bottleneck with several processors but NOT the i3. It is obvious that something else is at play here besides IPC and low res/settings testing here is misleading. What could be the problem? I can only guess, it could be that at 720p sandy and ivy processors get a boost by their famous μοp cache feature and that at 1080p this makes no difference while graphic load becomes more threaded as we increase the quality, giving the edge to the FX-6300.
Mastodon Ryzen
(12 items)
 
HP Z220
(8 items)
 
 
CPUMotherboardGraphicsRAM
R7 1800X Asus Crosshair VI Hero Asus R9 Nano Gskill TridentZ 
Hard DriveOptical DriveCoolingOS
Pny SSD 240GB Crucial MX100 CM Nepton 280L Win 10 
MonitorPowerCaseMouse
Acer Predator XG270HU Freesync XFX 750W Pro HAF XM Logitech G502 
CPUMotherboardGraphicsCooling
i7 3770 HP Quadro K2000 HP 
OSPowerCaseMouse
Win 7  HP 400W HP CMT RAT 7 
  hide details  
Reply
Mastodon Ryzen
(12 items)
 
HP Z220
(8 items)
 
 
CPUMotherboardGraphicsRAM
R7 1800X Asus Crosshair VI Hero Asus R9 Nano Gskill TridentZ 
Hard DriveOptical DriveCoolingOS
Pny SSD 240GB Crucial MX100 CM Nepton 280L Win 10 
MonitorPowerCaseMouse
Acer Predator XG270HU Freesync XFX 750W Pro HAF XM Logitech G502 
CPUMotherboardGraphicsCooling
i7 3770 HP Quadro K2000 HP 
OSPowerCaseMouse
Win 7  HP 400W HP CMT RAT 7 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [pclab] More Crysis 3 CPU benchmarks