X3350 / Q9450 Overclocking Thread

[bryanW1995]

this is very frustrating. I've looked all over, also without success

 

[Foxery]

The Thermal specs sheet is actually here:
121 page PDF
But I still don't see any obvious reference to TJunction in it. Perhaps someone can determine it based on the wealth of other figures and calculations they show?

 

[bryanW1995]

foxery, did you ever install a q9300 in your ip35e?

 

[bryanW1995]

ok, running prime95 at 1.4v in bios, 1.28 real, 450x8. temps are a little high at 77,70,70,69 right now. Using antec 900 with the 2x120 fans in front on medium, the rear 120mm on medium, and top fan on high. I have the tuniq exhausting towards the top fan because of the case location. Tuniq is at 2100 rpm. Top fan and tuniq are too loud for everyday operation imho. Of course, I won't be here, so the real question is if I can get wife to deal with it...:evil: I'm going to run another couple minutes and see how high I can get without increasing vcore any more.

Is anyone out there familiar with gtl settings on the ip 35 pro? I've heard here and there that they can help with stability, but right now I'm just guessing on whether to move them up or down, etc. The only gtl adjustments available are for core 0/2 and core 1/3, there aren't any for fsb,nb,etc.

 

[runawayprisoner]

Mmmm... may I ask everyone's VID? That might be a clue... to something. Mine was 1.125v.

 

[tenax]

Originally posted by: runawayprisoner
Mmmm... may I ask everyone's VID? That might be a clue... to something. Mine was 1.125v.

q9450
B1 stepping
1.125

 

[bryanW1995]

1.20 vid for me.

3640 was fine at 1.40, but I pushed to 3704. I couldn't get that stable more than a few minutes, even "pushing" vcore to 1.41. I'll claim success on this chip, I think that I have a legit chance of running it 24/7 at 3.6 without any fan higher than medium after I lap the tuniq and hsf.

 

[runawayprisoner]

B1 stepping? That's weird? I thought Q9450 was C1 or C0?

And... thanks for your input, guys. Might I ask for your max stable OC so far... as well? Thank you, and sorry for the trouble. I just want to know if there's any secret to getting over 475 FSB here.

 

[bryanW1995]

I'm scared to go over the 1.41 v that wasn't stable for me at 463x8. I might be able to do it with additional cooling on the mobo but I'm really very very pleased with this cpu.

tenax has an ES iirc.

mine is c1 stepping, sorry.

 

[runawayprisoner]

I thought the ES was C0? Oh well.

And... try increasing your NB volt and FSBT (VTT) volt. Take them to over 1.5v on each and see if you can get higher FSB and lower your vCore.

 

[afreak]

Originally posted by: runawayprisoner
I thought the ES was C0? Oh well.

And... try increasing your NB volt and FSBT (VTT) volt. Take them to over 1.5v on each and see if you can get higher FSB and lower your vCore.

I agree. From my experience, Vcore doesn't play a very important role in OCing to 4ghz. It's all about NB voltage, FSB voltage, and NB GTL reference voltage.

you can see my post about it here
http://forums.anandtech.com/me...=2174531&enterthread=y

 

[bryanW1995]

ok, right now I'm 8x450, running prime 95 at 1.37vcore, 1.52 mch, 1.2 cpu vtt, 1.09 ich, and 1.65 ichio. oh, and cpu gtlref is at 69% on all 4 cores.

standard voltages are: vcore 1.2, mch 125, cpu vtt 1.2, ich 1.05, ichio 1.5, and cpu gtlref 67%.

Many of these settings are either different or at least appear to be different than asus and/or gigabyte settings. the mch, ich, and ichio are the same settings I had on e6750 @ 3.6 so I just cranked them up to the same. the vtt was 1.31 and gtlref was at the standard 67%, obviously vcore was much higher.

I really think that I just want to leave this guy at 3.6 and crank down everything as low as possible. My only problem is that I'll probably end up running the stupid ballistix at 1:1.2 to get to the 266 strap...damn you 2.2v, why can't I quit you???

 

[Foxery]

Originally posted by: bryanW1995
foxery, did you ever install a q9300 in your ip35e?

It's due in Monday. Stupid FedEx.

Originally posted by: bryanW1995
Is anyone out there familiar with gtl settings on the ip 35 pro? I've heard here and there that they can help with stability, but right now I'm just guessing on whether to move them up or down, etc. The only gtl adjustments available are for core 0/2 and core 1/3, there aren't any for fsb,nb,etc.

Raising these very slightly may help stability at high FSB settings. By slightly, I mean like 3-5%. There was a technical writeup linked last week about exactly what this does... The key points I got from it, before it went over my head, was not to push this setting very far, and I got the impression that the writer's mobo had more options than our Abits do.

 

[bryanW1995]

does the ip35 have that setting, too? I have my 35-e in the box, waiting for another deal on G0 quads (or to see you get your Q9300 working )

 

[imported_BigJim]

Thinking out loud here. Maybe I speak for the "silent majority" on this thread. I just want a simple OC. No water, no lapping. Willing to bump a few voltages, but want to stick to air. As we know, power (and therefore temperature) go up exponentially with voltage. Even water, will take away the temperature, but not any potentially damaging high voltage. My theory is that most intel CPUs have a sweet spot at about 20% overclock. For example the e8400 with its 9x multiplier seems to go along easily up to 3.6GHz but voltage/temp go up exponentially after that. And I do want a quad, and 2 or 4 2GB memory sticks. If I'm honest with myself, I can't find a way to use 8GB. I think I can get away with 4 1GB sticks of 1066 AxeRAM (already own 2).

Interestingly, the q6700 is out. This is a 65nm part with a 10x multiplier (its not the qx6700 with the unlocked multiplier). The q9450 is a 45nm part with an 8x multiplier. Both of these are 2.66 parts. And with my "intel designs in 20% headroom" theory/urban-legend, the easy OC runs to about 3.2Ghz on both chips.

I don't want to p*ss on the campfire here, but the q6700 is a lot cheaper. It gets an Intel price cut on April 20, and there have already been a few sales. It also goes easier on your FSB and your NB, due to its 10x multiplier. Indeed that multiplier would be welcome news for those who want to push for 3.6 or 4.0Ghz. Yes, it doesn't take advantage of the 1066 and 1333 memory, but that stuff sells a a huge premium, I mean 100% extra price for 33% extra (maybe) performance. Someone talk me out of it, but it just seems like this is the way to get through the hot summer until Nehalem hits at the end of the year and makes all our motherboards and CPUs obsolete.

 

[bryanW1995]

the 9x multi of the Q6600 is fine, especially for your ddr2 1066 ram. 3.6 is pretty much the limit on most quads (other than the QX9650) on air for 24/7 use. A Q6600 g0, a Q6700, a Q9450, all of them are going to max in most cases around 3.6. I got lucky on mine and could conceivably push 3.7 or maybe even 3.8, but I don't want to run 1.42+ vcore 24/7 on a 45nm quad. The 65nm quads aren't fsb-limited like the 45nm's, but they are HEAT-limited on air. If you're worried about a hot summer then you might want to rethink things and maybe even shoot for a Q9300 on a ds3L or ip35-e.

 

[imported_BigJim]

Thanks man. Sounds like any of those could work for me. For some reason I couldn't get my q6600 SLACR over 2.88GHz, so I sold it off. Feel like I should try something new maybe the Q6700. Is this exactly the same as the QX6700 but now they locked the multiplier to 10?

Saw this on Fudzilla "Intel plans it usual price cut that will affect mostly the old 65nm parts. On April 20th Intel will cut Core 2 Quad Q6700 from the current $527 suggested retail price to $270. The super popular Q6600 at 2.4GHz will drop from $270 to $229. At the same time, Intel plans to introduce its Core 2 Duo E8300 45nm based CPU with 6MB cache and 2.83GHz clock and it will start selling it for $169."

 

[aigomorla]

I will say this b4 people start having dead chips in a few months.

DO NOT FEED YOUR 45nm QUAD ANY VOLTAGE OVER 1.45V FOR LONG PERIODS OF TIME.

You have been warned.

Also thanks to ruby:

If you compare real temp and core temp, you'll see that its 10C off. Also the TDP is 10C. I wonder whose TDP is the 10 higher. :T

So you can use either program. If you do use realtemp, the TDP would be 95. If you used coretemp it would be 105. DOE i gave the answer away to the top.

 

[bryanW1995]

I brought this up the other day to mark. Neither of us could find the answer. The problem is, I'm at 70/65/67/63 on coretemp .97.1 right now and 60/55/57/53 on real temp. core temp shows 105 tjunction and real temp shows 95 tjunction. Based upon what everyone else has said I'm operating under the assumption that coretemp is correct, however, I would like to know for sure.

 

[TankGuys]

Try this for your temps:

http://www.xtremesystems.org/f...?p=2809778#post2809778

 

[bryanW1995]

wow, that's great info! at least we now know WHY we can't find tjunction max anywhere with certainty... and it's good to note that realtemp does appear to be more accurate with their 95c tjunction max, too.

 

[aigomorla]

Originally posted by: bryanW1995
wow, that's great info! at least we now know WHY we can't find tjunction max anywhere with certainty... and it's good to note that realtemp does appear to be more accurate with their 95c tjunction max, too.

You cant ask what is more reliable because there both the same.

The results are identical in the aspect that one is 10C lower. Coretemp on every aspect is the same as realtemp, but coretemp is only 10C higher on everything nimus the VID.

 

[runawayprisoner]

Okay... new Q9450 may come in anytime this week. It's time to see if it's the chip or the motherboard.

 

[themattman]

Unfortunately it turned out that my motherboard doesn't like certain DDR2-1066 sticks, so I decided to order a set of mushkin DDR2-800 sticks to make sure it isn't a problem with the motherboard. I can only stare at my Q9450 and the TRUE attached to it

 

[Foxery]

Ok, Q9300 (Xeon equiv) came in today, and so far my thoughts are... WTF?

Temperatures make no sense

Perfectly stable at 3.2ghz (428 FSB, 1.25volts) but as you can see, it should have burst into flames if RealTemp is to be believed. CoreTemp, of course, reads even higher. And yet I'm sitting here writing this while Prime95 cranks away, and listening to MP3s to boot...

edit:
Results @ stock settings a few posts down. Still stupidly hot.