![]() |
|
|
#1 |
![]() |
9600GT: Raise the PCI-E bus? or the core, or both.
Hey, just had a question and hoped some of you guys could give some advice or thoughts about it. I've been tweeking with my 9600GT, OCing and such, as well as my cpu and PCI-e bus. Now as im sure most of you know, Nvidia has that link trick thingie that links the Pci-e bus to the core clock of the GPU, thus raising the pci-e bus by 10% will increase the core of the card by 10%. With that said, I've read and seen people -not- touching their pci-e bus because of that and just OC the card it's self. Currently I have the pci-e bus running higher than stock, as well as the core of my card. Would I see better results if I left the core of the card untouched and tweaked the pci-e bus or vise versa, left the Bus untouched and just OC'd the card. Or should I find a reasonable medium with both. Temps seem to be fine OC'd with load at around 67 degrees, idling around 46-50. Thanks
![]() Edit: I realized after I posted this could probably go in the Overclock forums, Mods if you see fit, feel free to move it
|
|
|
|
|
|
#2 |
|
Eligible for custom title
Join Date: Oct 2006
Location: Missoula, Montana, USA
Posts: 6,558 (2.72/day)
Thanks: 727
Thanked 1,349 Times in 1,118 Posts
|
I just got my 9600GT and I'm considering messing with it. I currently have mine OC'd to 760core/1900shader, and memory at stock 1000 (Palit Sonic 9600GT, it's somewhat pre-oc'd, 700 core, 1750 shader, 1000 (2000 effective) memory). I've been using Rivatuner 2.08 for now, I am also going to try Palit's provided OC program also.
I'd say you should run some game benchmarks and even some synthetic benchmarks and see what your findings are! I'm definately interested myself, I suppose OC-ing the PCI-e bus could also give the card some more "theoretical" bandwidth...and maybe give an extra couple of FPS at the same OC in comparison to 100 locked PCI-e and OC'd via program.
|
|
|
|
|
|
#3 |
![]() |
Cool
, let's see if we get an answer
|
|
|
|
|
|
#4 |
|
Creator Solaris Utility DVD
Join Date: Aug 2005
Location: Reinacting scenes from platoon with Charlie Sheen
Posts: 13,708 (4.84/day)
Thanks: 4,365
Thanked 3,295 Times in 2,311 Posts
|
i wonder...maybe ill try but you see upping the pci-e freq OC's it...so if your already maxing it via RT you may end up with an unstable card.
__________________
I Made the Millionth post! | "Please come to WI now so I can beat you over the head with a bratwurst."-Kreij
PS3 mod 8500/8600GT Mod Guide Rebuild a Copperhead Heat Ware NF4 Ultra SLI Mod Solaris Utility DVD 4.0 Broken CPU pin guide Vista Mark |
|
|
|
|
|
#5 |
|
Maximum Overclocker
Join Date: Nov 2006
Location: Ohio
Posts: 12,236 (5.16/day)
Thanks: 2,054
Thanked 2,039 Times in 1,549 Posts
|
leave the PCI-E bus locked at no more than 110. Mine is 101.
|
|
|
|
|
|
#6 | |
|
The Mad Moderator
Join Date: Feb 2007
Location: Wherever I can find the iron.
Posts: 11,448 (5.01/day)
Thanks: 2,519
Thanked 3,054 Times in 2,098 Posts
|
Quote:
|
|
|
|
|
|
|
#7 |
![]() |
Could you elaborate?
Yeah, you don't -actually- see the difference, as in, visable higher core clocks. but W1zz was able to prove that upping the PCI-E bus increased the performance of the card. |
|
|
|
![]() |
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | |
|
|
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Sapphire PURE Innovation PI-A9RX480 | W1zzard | Reviews | 11 | May 8, 2009 08:46 PM |
| Core 2 Duo Temperature Guide Updated March 24th, 2007 | KATANAUK | Overclocking & Cooling | 34 | Jan 20, 2009 12:01 AM |
| Core 2 Clubhouse | tkpenalty | techPowerUp! Club Forum | 329 | Feb 9, 2008 10:17 PM |
| IDF Prague: Quad-Core | W1zzard | Reviews | 1 | Oct 18, 2006 04:49 PM |
| Matrox to make low-profile PCI cards | W1zzard | News | 0 | Sep 18, 2004 11:46 AM |