'...most of the 45nm Core 2 Quads don't overclock well...'

Soldato
Joined
31 Oct 2007
Posts
2,580
Location
Gods County(Allegedly)
The above is taken from Custom PC's cpu review in May '09. - normally I wouldn't even think twice about comments made in reviews but this one rattled my cage a little.

In their review they said that the max stable overclock they could get on a Q9550 was 3.3 ( motherboard Biostar TPower I45) which I agree is a low overclock.

I'm certainly no expert in the overclocking department but my Q9550 (CO stepping) is happily running at 3.4 undervolted , will run at 3.6 without any issue & I've had it up to 3.8 with a little boost to the vcore (less than 1.3v).- on an Asus P5K Premium P35 board.

Main reason I run it at 3.4 is that , imo, the extra performance I get running faster is not really worth the extra heat/wear BUT it would happily run at 3.8+ if I was so minded !!

So my thoughts were ' what's ' their problem :-

1) Chip - nothing I've heard recently saying that the 'new' issue (or come to that matter any 'old' issue) Q9550's are poor clockers ; or
2) Motherboard - all reviews I've read have been pretty positive; or
3) Poor ram or other component - surely not !! ; or
4) Some Error - human or otherwise ??

I've no intention of cancelling my sub. to the mag as I enjoy the read but I shall take their reviews with an even larger grain of salt in future ...


UNLESS I'm wrong and the 45nm quads no longer clock well or the Biostar board is poor AND if this should be the case then please enlighten me. :o

I also found a little worrying there - uncommented - use of 1.45v to overclock the 45nm dual/quad cores . I'm not saying that such v cannot be used just that all the informed user reviews I've read of on this subject say that using that level of v's may damage the chip pretty quickly and should be thought of as a max.

Sorry - Rant over :D
 
happily running at 3.4 undervolted

Main reason I run it at 3.4 is that , imo, the extra performance I get running faster is not really worth the extra heat/wear


Exactly my situation too, and no I've not heard anything about poor performance from new batches of chips or boards either :)
 
you can get 3.4 at stock....and the Voltages they put through there CPUs seems to be a bit suicidal sometimes.
I have a crappy Q9550 but others have happily hit 4Ghz on under 1.35V
 
I suspect this is a "problem exists between chair and keyboard" incident.

Like many of the early Phenom reviews, testers either simply didn't know what they were doing or didn't take the time to do it properly.
 
Hmmm, whilst I like Custom PC, they're not really thorough enough sometimes and it irks me that they form opinions without ever considering other findings in the industry.

I had a C0 Q9450 that did 3.6ghz pretty easily - would have done 3.8-4ghz if I could have been bothered to fiddle with the GTLs.

I also found a little worrying there - uncommented - use of 1.45v to overclock the 45nm dual/quad cores . I'm not saying that such v cannot be used just that all the informed user reviews I've read of on this subject say that using that level of v's may damage the chip pretty quickly and should be thought of as a max.

Someone wrote in a few months back saying this - "I've heard over 1.4v can damage the chips - stop recommending such high voltages" and they basically told him to shut up.
 
This is one of the reason's that i cancelled my subscription to Custom PC a year or so ago. Poor quality reviews and half baked opinions. Plus by the time the magazine has arrived it's old news anyway.

1.45v should be the max given to a 45nm core.
 
testers either simply didn't know what they were doing or didn't take the time to do it properly

Am of the opinion that this is , sadly , probably true - A lesson in how not to 'win friends & influence people' :o
 
9550 chip

No trouble at 3.8 at 1.17v have had it at 4.4 but cooling is a bit lacking and temps are a bit high, have got to windows at 4.7 but bsod so went back to 3.8 and runs great there even encoding 3hr films only gets to 60-62C so quite happy thank you intel
 
I think they are just incompetent at overclocking. They also probably get extra money based on the bias of their reviews and such like. I read an article over on another forum, which was saying about how retail shops are posting reviews quicker than the review sites themselves, and so they should since the shops seem to do a better job at it as well.
 
The biostar is known to be a poor clocker of quads

Was not aware - so thanks for that. The 3/4 reviews I'd read on the board didn't mention this ; as I wasn't interested in buying did not research any further (knowing me I'd probably have found out after buying :().

This still does not excuse the mag. - they should know this and mention it !!

Got to say that even taking into account the above their 'results' are still somewhat contradictory as , using the same board, they were able to get the following 'best' overclock results :-

Q6600 + 41%
Q8200 + 20%
Q8300 + 32%
Q9300 + 32%
Q9450 + 17%
Q9550 + 17%
Q9650 + 32%

Disregarding the Q6600 (simply because it's on the 65nm build) there's still a lot of difference between the chips & if it were the board you'd think that the results would be consistent, which they aren't - :confused:
 
customPC is a magazine.



There isnt a magazine on earth, be it pc, audio, visual hardware.....whatever..... thats isnt out there for one thing and one thing only - to sell hardware. CustomPC hasnt been acurate for years and they wont be because they arent there to give you a throughly thought out and executed, honest review. they are the pc equivilent of what HIFI, giving a product a 5 star review and then having an advert for said product on the very next page. Coincidence much?

Custom is an amusing mag but never take it as ghospel. if you want the truth stick to the forums (not theirs i might add).
 
Was not aware - so thanks for that. The 3/4 reviews I'd read on the board didn't mention this ; as I wasn't interested in buying did not research any further (knowing me I'd probably have found out after buying :().

This still does not excuse the mag. - they should know this and mention it !!

Got to say that even taking into account the above their 'results' are still somewhat contradictory as , using the same board, they were able to get the following 'best' overclock results :-

Q6600 + 41%
Q8200 + 20%
Q8300 + 32%
Q9300 + 32%
Q9450 + 17%
Q9550 + 17%
Q9650 + 32%

Disregarding the Q6600 (simply because it's on the 65nm build) there's still a lot of difference between the chips & if it were the board you'd think that the results would be consistent, which they aren't - :confused:

To say its poor was probably inaccurate, but there are better clockers for quads out there. Particularly the P5Q boards.

The best resource for the TP I45 is here:

http://www.rebelshavenforum.com/sis-bin/ultimatebb.cgi?category=2
 
Back
Top Bottom