Welcome Guest ( Log In | Register )

44 Pages « < 13 14 15 16 17 > » Bottom

Outline · [ Standard ] · Linear+

 Intel Penryn Overclocking Thread, Toasting next gen 45nm babes ~

views
     
clawhammer
post Feb 14 2008, 11:00 PM

///M
Group Icon
VIP
8,788 posts

Joined: Jan 2003
From: Kuala Lumpur




QUOTE(allenultra @ Feb 14 2008, 12:00 PM)
I had experience with CS crashing on X2 3600+ 2.4ghz prime stable 10 hours.
The problem was ram timing. Trfc, too tight.

I guess a lot stressing needed to achieve a really stable overclocked rig.
*
You should use "Blend" and not only "Small FFT".
IcEMoCHa
post Feb 15 2008, 09:20 AM

~本田~
******
Senior Member
1,988 posts

Joined: Aug 2005
From: Puchong, Selangor



q9450 @ 3.6ghz or e8400 @ 4ghz ... i dunno which 1 to choose now...
kmarc
post Feb 15 2008, 09:31 AM

The future is here - Cryptocurrencies!
Group Icon
Elite
14,576 posts

Joined: May 2006
From: Sarawak



QUOTE(IcEMoCHa @ Feb 15 2008, 09:20 AM)
q9450 @ 3.6ghz or e8400 @ 4ghz ... i dunno which 1 to choose now...
*
If you are only using your rig for gaming, your E6600 is already enough la.... wink.gif
cstkl1
post Feb 15 2008, 11:45 AM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

QUOTE(IcEMoCHa @ Feb 15 2008, 09:20 AM)
q9450 @ 3.6ghz or e8400 @ 4ghz ... i dunno which 1 to choose now...
*
unless ure lucky

geee got somebody in lyn whos q6600 can do 3ghz at 1.1v.. kekeke

general_odin
post Feb 15 2008, 12:12 PM

no work, no money
*****
Senior Member
904 posts

Joined: Nov 2006
From: Selangor & Malacca, Malaysia



woot~
got my E8200 with GA-EP35-DS3 overclocked....

i just pump the FSB, no Vcore and Vdimm mod
OCed to 400Mhz FSB and 800MHz RAM

Attached Image

happy with it tongue.gif


Added on February 15, 2008, 12:15 pm
QUOTE(IcEMoCHa @ Feb 15 2008, 10:20 AM)
q9450 @ 3.6ghz or e8400 @ 4ghz ... i dunno which 1 to choose now...
*
OMG... u must be too rich to get a Q9450....
E8400 is more than enough for gaming purpose...
unless ur doing really lots of computing and takes in lots of processing power

i'm sure ur not a professor who work in a computer lab
calculating protein structure or wad so ever rite....

This post has been edited by general_odin: Feb 15 2008, 12:15 PM
flatfinger
post Feb 15 2008, 12:42 PM

Totally Retired
*****
Senior Member
925 posts

Joined: Jan 2005
From: depan lcd...dalam bilik..sekitar cyberjaya...

@general_odin

did u disable ur C1E function..if not the cpu will downclock the multiplier when no load...when on load it will back to the highest multiplier...
cstkl1
post Feb 15 2008, 01:44 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

off ure PAE..
no point..

IcEMoCHa
post Feb 15 2008, 07:12 PM

~本田~
******
Senior Member
1,988 posts

Joined: Aug 2005
From: Puchong, Selangor



QUOTE(kmarc @ Feb 15 2008, 09:31 AM)
If you are only using your rig for gaming, your E6600 is already enough la....  wink.gif
*
Wel... i still need another proc for building my other comp... lolz
clawhammer
post Feb 15 2008, 07:33 PM

///M
Group Icon
VIP
8,788 posts

Joined: Jan 2003
From: Kuala Lumpur




QUOTE(IcEMoCHa @ Feb 15 2008, 07:12 PM)
Wel... i still need another proc for building my other comp... lolz
*
I agree with kmarc and E6600 is more than enough but I guess your Quad and Pennryn are both superb so I suggest you sell me your E6600 laugh.gif
IcEMoCHa
post Feb 16 2008, 10:58 AM

~本田~
******
Senior Member
1,988 posts

Joined: Aug 2005
From: Puchong, Selangor



QUOTE(clawhammer @ Feb 15 2008, 07:33 PM)
I agree with kmarc and E6600 is more than enough but I guess your Quad and Pennryn are both superb so I suggest you sell me your E6600 laugh.gif
*
nono.... doh.gif i have a extra mobo(intel 975xbx2), graphic card(nvidia 7900gs extreme), PSU(420W), RAM(1GB corsair xms2 6400).... so i was thinking of building another rig from those... need another proc for my current rig and transfer the e6600 to the other...
Silverfire
post Feb 17 2008, 11:38 AM

Cruxiaer
*******
Senior Member
4,947 posts

Joined: Nov 2007



OK! I got a question: Which of those following OC settings run fastest?

Setting 1:
user posted image
Unstable with stock vcore. RAM is at 1066mhz @ 5-5-5-15.

Or Setting 2:
user posted image
Unstable with 1.225v. RAM is at 1000mhz @ 5-5-5-15.

Or Setting 3: 400mhz x8/3.2ghz(tested, stable) with RAM at 800mhz @ 4-4-4-12 @ 2.1v?


Added on for not 1:1.
Or Setting 4:400mhz x8/3.2ghz(tested, stable) with RAM at 1066mhz @ 5-5-5-15?

Or Setting 5:
user posted image
100% stable with stock vcore however I put 1.225v for it. RAM is at 960mhz @ 5-5-5-15.
Below is temperature for Setting 5:
user posted image

Please and Thank You!

This post has been edited by Silverfire: Feb 17 2008, 11:44 AM
OC4/3
post Feb 17 2008, 01:18 PM

.
Group Icon
Elite
4,746 posts

Joined: Dec 2007
From: Speed rule


First setting is the best and try to get it stable with 8x multi.


Added on February 17, 2008, 9:39 pmFirst setting is the best and try to get it stable with 8x multi.

This post has been edited by OC4/3: Feb 17 2008, 09:39 PM
flatfinger
post Feb 17 2008, 10:03 PM

Totally Retired
*****
Senior Member
925 posts

Joined: Jan 2005
From: depan lcd...dalam bilik..sekitar cyberjaya...

@Silverfire

Setting 3: 400mhz x8/3.2ghz(tested, stable) with RAM at 800mhz @ 4-4-4-12 @ 2.1v?

this one is the best...if u dont believe me...just run SP1M for all the setting (3x-5x persetting)...u can see the diff after that

800Mhz 4-4-4-12 is faster than 1000Mhz 5-5-5-15

p/s: ignore the temp for this time...all mobo doesn't seem given accurate reading...

This post has been edited by flatfinger: Feb 17 2008, 10:10 PM
clawhammer
post Feb 17 2008, 10:20 PM

///M
Group Icon
VIP
8,788 posts

Joined: Jan 2003
From: Kuala Lumpur




Silverfire, I'll go for the highest FSB followed by the RAM timings. Although in benchmark programs it'll show significant number difference for tighter timings, you can't really feel it in real life usage smile.gif
cstkl1
post Feb 17 2008, 11:01 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

dont u think ure overvolting that proc of ures..

and 533 bus.. the prob is not the vcore.. is the NB voltage

thats y ure 500x6 had problems

technically depends.. games i would go for 500x6

overall performance would be ure 400x8. at the lowest trd CL3.


but pretty sure u could have gone to 9x400 at less than 1.2v



This post has been edited by cstkl1: Feb 17 2008, 11:07 PM
Silverfire
post Feb 18 2008, 06:24 AM

Cruxiaer
*******
Senior Member
4,947 posts

Joined: Nov 2007



To flatfinger: Alright, will give it try~

To clawhammer: Ya... My brother also say like that.

To cstk1: How to increase NB voltage? When in 500x6 I can use normally for around 10 mins then suddenly BSOD-ed.


Added on for the line.
Thanks in advance!

This post has been edited by Silverfire: Feb 18 2008, 06:25 AM
cstkl1
post Feb 18 2008, 09:33 AM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

QUOTE(Silverfire @ Feb 18 2008, 06:24 AM)
To flatfinger: Alright, will give it try~

To clawhammer: Ya... My brother also say like that.

To cstk1: How to increase NB voltage? When in 500x6 I can use normally for around 10 mins then suddenly BSOD-ed.


Added on for the line.
Thanks in advance!
*
trial and error
definitely more than 1.5v on NB
and ure fsb termination voltage cannot go more than ure vcore.


clawhammer
post Feb 18 2008, 02:59 PM

///M
Group Icon
VIP
8,788 posts

Joined: Jan 2003
From: Kuala Lumpur




Let's not confuse the whole world (especially new learners) with regards to "FSB termination voltage". Many motherboards DOES NOT have this feature and it is either an ASUS or X38 standard feature (I don't know). Not having this voltage doesn't mean you can't overclock and this is one of the least to worry about.

Our focus are primarily on these few common voltages, i.e. Vcore, Vdimm, Vmch, etc. You can't overclock if your motherboard does not have settings for the mentioned voltages but you can still overclock if it doesn't have "FSB termination voltage".
cstkl1
post Feb 18 2008, 03:05 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

QUOTE(clawhammer @ Feb 18 2008, 02:59 PM)
Let's not confuse the whole world (especially new learners) with regards to "FSB termination voltage". Many motherboards DOES NOT have this feature and it is either an ASUS or X38 standard feature (I don't know). Not having this voltage doesn't mean you can't overclock and this is one of the least to worry about.

Our focus are primarily on these few common voltages, i.e. Vcore, Vdimm, Vmch, etc. You can't overclock if your motherboard does not have settings for the mentioned voltages but you can still overclock if it doesn't have "FSB termination voltage".
*
fsb termination voltage aka cpu vtt voltage aka ht-link voltage.

all boards has this

checking ure manual now.. find it weird that ure mobo doesnt has this.

P35 DS3L
http://america.giga-byte.com/FileList/Manu...d)s3l_2.0_e.pdf

Page 47
FSB Overvoltage

blink.gif

This post has been edited by cstkl1: Feb 18 2008, 03:29 PM
clawhammer
post Feb 18 2008, 03:54 PM

///M
Group Icon
VIP
8,788 posts

Joined: Jan 2003
From: Kuala Lumpur




The explanation should be more specific and for people that knows, we will understand FSB termination voltage is CPU VTT but for those that don't, they'll probably go lost somewhere and either spend their money on a new piece of hardware but in the end still don't get their desired results.

I did not specifically mention a particular motherboard but there are many boards out there without the word "FSB Termination Voltage" stated.

At the same time, increase of CPU VTT does not apply for overclocking of all CPU's. To some, it doesn't make a lot of difference.


Added on February 18, 2008, 4:03 pmSome further reading of what GTL, VTT is about:
http://www.edgeofstability.com/articles/df...5/gtl/gtl1.html

Excerpt from the article
"Simply raising VTT may or may not create voltage margins necessary to sufficiently skew signals as required to meet minimum sample and hold times for increased bus frequencies. Additionally, processors are particularly sensitive to even small increases in VTT as bus impedances and termination resistance values are quite low. In fact, increasing VTT will most likely just create unwanted device heating with little to no change in FSB stabilization.

http://www.thetechrepository.com/showthread.php?t=87

This post has been edited by clawhammer: Feb 18 2008, 04:03 PM

44 Pages « < 13 14 15 16 17 > » Top
 

Change to:
| Lo-Fi Version
0.0903sec    1.19    6 queries    GZIP Disabled
Time is now: 20th December 2025 - 12:24 PM