Overclock.net › Forums › Graphics Cards › NVIDIA › Stock Phenom II 965=GTX 580 SLI bottleneck?
New Posts  All Forums:Forum Nav:

Stock Phenom II 965=GTX 580 SLI bottleneck? - Page 6

post #51 of 56
About what I mentioned. You could do better, but gotta test.
post #52 of 56
Thread Starter 
Quote:
Originally Posted by james8 View Post
i was making the point that you need a 3.8 GHz i7 or equivalent i5 to not be bottlenecked by GTX 580 SLI.
the link i put showed GTX 480 SLI being bottlenecked by anything less than a 3.8 Ghz i7.
so since the GTX 480 SLI is being bottlenecked, the cpu bottleneck on the GTX 580 SLI will be worse.

by the way i don't care if fps is over 60, if you overclock the CPU and fps go up, it means that you are cpu bottlenecked.
sure, if fps is over 60, you can arguably say that the bottleneck does not impact the gaming experience. but it is still a bottleneck
I dunno I don't think 15% is going to be that big of a deal
but then again it might be more than 15% with a stock 965 though
That guy is loaded anyway. I'll just tell him to upgrade to SB or BD later down line
post #53 of 56
Quote:
Originally Posted by th3illusiveman View Post
Thanks man!

Whats a Generally safe North-bridge Voltage for a 24/7 OC?
Up to 1.4v is fine on air cooling
    
CPUMotherboardGraphicsRAM
i7 2600k ASRock P67 Extreme4 MSI GTX 580 Lightning 2x4GB Corsair Vengeance DDR3 1866 9-10-9-27-1T 
Hard DriveOptical DriveCoolingOS
Mushkin 60GB SSD + 2xWD Black 500GB RAID0 Sony DVD-RW Noctua NH-D14 Windows 7 
MonitorKeyboardPowerCase
LG 22" 1080p + Acer 18" Das Keyboard - MX Blue Antec TruePower New TP-650 Cooler Master HAF 922 Black 
Mouse
Logitech 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 2600k ASRock P67 Extreme4 MSI GTX 580 Lightning 2x4GB Corsair Vengeance DDR3 1866 9-10-9-27-1T 
Hard DriveOptical DriveCoolingOS
Mushkin 60GB SSD + 2xWD Black 500GB RAID0 Sony DVD-RW Noctua NH-D14 Windows 7 
MonitorKeyboardPowerCase
LG 22" 1080p + Acer 18" Das Keyboard - MX Blue Antec TruePower New TP-650 Cooler Master HAF 922 Black 
Mouse
Logitech 
  hide details  
Reply
post #54 of 56
Quote:
Originally Posted by Rookie1337 View Post
LOL.
OK I can answer this: MOST GAMES ARE GPU DEPENDENT. Why people on here are willing to overlook that is beyond me. How many people are going to notice a 10 or less FPS difference without FRAPS telling them?

For Crysis, it is GPU dependent. Always has been and always will be.
If you fail to specify the settings involved, then you are completely wrong. Crysis (and many other games) can easily be bottlenecked by the CPU ... all you have to do is lower the gfx settings.

See this post:

http://www.overclock.net/11921967-post21.html

Here's a synopsis (my sig rig, CPU stock vs. OC'd):




Personally I wouldn't do 2x580's on a PHII at anything less than 30" monitor resolution ... unless there was room in the budget to upgrade the CPU/Platform at some time in the not-to-distant future.

Quote:
Originally Posted by james8 View Post
<snip> ... if you overclock the CPU and fps go up, it means that you are cpu bottlenecked.
sure, if fps is over 60, you can arguably say that the bottleneck does not impact the gaming experience. but it is still a bottleneck
This ^^^ exactly. People seem to like to make up their own definitions of bottleneck around OCN. There is no requirement in the definition of the term that says it must be 'noticeable' or dictates where in the FPS spectrum it must occur.

If increasing CPU clocks (or just having a more powerful cpu/platform combo) increases FPS AT ALL in a given testing scenario, then there's a bottleneck by the cpu/platform, on that particular test, at the particular settings you're running it at. That's what the term means.

Beyond that, you have to start talking about frequency/severity of BN, what games it occurs in, at what IQ settings, etc.
Edited by brettjv - 3/1/11 at 12:26pm
    
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 #55 of 56
This thread interests me. I wonder what my vantage scores would be..... I may check and post em up after work!
SuperMegaPwnsauce
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 4770k Asus Z87-PRO MSI GTX 1080 Gaming X Gskill Ripjaws X 19200 16GB 
Hard DriveCoolingOSMonitor
Samsung 850 EVO 500GB SSD BeQuiet! Dark Rock Pro 3 Windows 10 Home 64Bit  Dell S2716DG 2K 144hz G-Sync 
KeyboardPowerCaseMouse
Logitech G910 Orion Spectrum Seasonic Flagship PRIME TITANIUM 650W Fractal Design Define R5 W/ 3 Phanteks PH-F140S... Logitech G Pro 
Mouse PadAudioAudio
PC Gaming Master Race Glorious XXXL Pad & Razer... Creative Sound Blaster Z & Schiit M&M 2 Uber/Mu... HifiMan HE400i & Antlion ModMic 4.0 
  hide details  
Reply
SuperMegaPwnsauce
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 4770k Asus Z87-PRO MSI GTX 1080 Gaming X Gskill Ripjaws X 19200 16GB 
Hard DriveCoolingOSMonitor
Samsung 850 EVO 500GB SSD BeQuiet! Dark Rock Pro 3 Windows 10 Home 64Bit  Dell S2716DG 2K 144hz G-Sync 
KeyboardPowerCaseMouse
Logitech G910 Orion Spectrum Seasonic Flagship PRIME TITANIUM 650W Fractal Design Define R5 W/ 3 Phanteks PH-F140S... Logitech G Pro 
Mouse PadAudioAudio
PC Gaming Master Race Glorious XXXL Pad & Razer... Creative Sound Blaster Z & Schiit M&M 2 Uber/Mu... HifiMan HE400i & Antlion ModMic 4.0 
  hide details  
Reply
post #56 of 56
I am 6 pages late to the discussion but I will offer my analysis. It is a little long and there is a logic flow. So, follow the steps through to the end with some patience:

(I) First off, one fact
A stock Phenom II X6 1100T is sufficient to run GTX 570 SLI at full potential (at 2560x1600 with high IQ).
Link:http://benchmarkextreme.com/Articles...1100T/P17.html

(II) CPU consideration
a. stock Phenom II X4 965 performs similarly to X6 1100T because the 2 additional cores do not make a 'significant' difference in gaming

b. stock Phenom II X4 965 at 3.4GHz is comparable to stock Phenom II X6 1100T at stock 3.2GHz because the 0.2MHz is insignificant.

So, we can safely say that ''a stock Phenom II X4 965 is sufficient to run GTX570 SLI at full potential at 2560x1600 with high IQ.

(III) Resolution consideration
Now, it is easier to conclude if the OP's friend's resolution is 2560x1600.
But for the sake of investigation, assume OP's friend's resolution is 1920x1080/1200.

Then, the Q becomes a graphical comparison between:
A: GTX570 SLI at 2560x1600
B: GTX580 SLI at 1920x1080

A poses no problem for a stock Phenom II X4 965. How about B? B is obviously putting more load onto the CPU.

Using guru3d's framerates for 7 games tested, I averaged the framerate decrease when going from 1920x1200 to 2560x1600:-

1. for GTX570 SLI: (28+29+23+28+44+16+28)/7 = 28%
Link:http://www.guru3d.com/article/geforc...0-sli-review/5

2. for GTX580 SLI: (29+20+20+21+40+ 5+22)/7 = 22%
Link:http://www.guru3d.com/article/geforc...0-sli-review/5

Take a middle number of 25%

(IV) GPU consideration
At 1920x1200, GTX580 SLI is about 10% faster (more framerates) than GTX570 SLI
(averaging data of the 7 games here:http://www.guru3d.com/article/geforc...0-sli-review/5)


(V) Putting (III) and (IV) together

Things do not go linear but just as a rough estimate, let's say: B gives about 35% more framerates than A (25% coming from a lower resolution and 10% coming from a stronger pair of GPUs).

Then, the Q becomes: Is a stock Phenom II X4 965 sufficient to run GTX580 SLI at 1920x1200 (which is 35% graphically more than GTX570 SLI at 2560x1600 with high IQ) at full potential?

(VI) For this Q, we turn to this review

http://benchmarkextreme.com/Articles/GTX%20570/P21.html
This shows that, for the averave of 19 games, GTX570 3-way SLI is about 34% faster than GTX570 2-way SLI, at 2560x1600.
C: GTX570 3-way SLI at 2560x1600


So, C is graphically 34% higher than A.
And, B is also graphically 35% higher than A.

Then, the Q becomes: Is a stock Phenom II X4 965 sufficient to run GTX570 3-way SLI at 2560x1600 at full potential with high IQ.
(This Q is just the same as asking: Is a stock Phenom II X4 965 sufficient to run GTX580 SLI at 1920x1200 at full potential with high IQ)


(VI) Conclusion
The final answer is: Yes and No
Yes:
Quote:
'Everything below Cryostasis in the above graph is heavily GPU dependent and a stock Phenom II X6 1100T is sufficient even in TRI SLI mode.
This means, stock Phenom II X4 965 can run GTX580 SLI at 1920x1200 at full potential for games like Cryostasis, Battleforge, Stalker Cop, Napoleon Total War, Lost Planet, Metro 2033, Alien vs Predator.

No:
Quote:
In some GPU heavy titles TRI SLI exhibits significant CPU bottleneck even up to 4.2 GHZ CPU frequency. In the above graph all the games above the "average" bar fall in this category.
This means, Phenom II X4 965 experiences significant bottleneck for games like Resident Evil 5, Hawx2, Crysis Warhead, Mafia 2(no Physx), FarCry 2, Lost Planet 2, Necrovision

http://benchmarkextreme.com/Articles...1100T/P17.html

(I did this very late at night...excuse any mistake made)
Edited by windfire - 3/1/11 at 10:08pm
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: NVIDIA
Overclock.net › Forums › Graphics Cards › NVIDIA › Stock Phenom II 965=GTX 580 SLI bottleneck?