Wolfdale/Yorkfield overclocking thread...

PangingJr · Jul 10, 2008, 10:49 PM

hmm, i never have the kind of CPU so i would have to try booting up the system with a few core voltage settings through the BIOS,
can you boot the system and get in to Windows at all with 1.4-1.50V vcore?

zaim · Jul 11, 2008, 12:25 AM

haven't tried =P

swimtech · Jul 11, 2008, 07:45 PM

Cool PangingJr. - nice stable 3.6! A great overclock!

PangingJr · Jul 11, 2008, 08:50 PM

3.46GHz aren't too far apart in overall CPU and memory performance and it will use less voltages all around.

PangingJr · Jul 27, 2008, 02:29 PM

PangingJr · Jul 29, 2008, 12:13 AM

my best Q9300 475x7.5 overclock, so far.
as far as i know, all the voltage values that i use in this overclock are well within limit on all hardwares.
and this may be my final overclock on this CPU.

PangingJr · Jul 30, 2008, 01:24 AM

just another overclock attempt.

is it possible to Overclock core 2 Quad 9300 using (mostly) only Vcore and Vdimm.?

PangingJr · Jul 30, 2008, 01:38 AM

PangingJr · Jul 31, 2008, 02:24 AM

just another overclock attempt.

will using more Vcore cause system overheating and a crash?

edit the line "4x1GB DDR2 1080 MHz @ 2.1V",
the system memory is actually runing "4x1GB DDR2 1110 MHz @ 2.1V"

PangingJr · Jul 31, 2008, 02:56 AM

wow this is a hot day today (room's air con is under cleaning).
2hrs...

PangingJr · Jul 31, 2008, 04:13 AM

well, the room temp is getting even high and it's like a room with 35C or more right now.

this overclock setting has been testing by myself for sometime now, i can either use a more lower vcore setting with an increase on the other voltage setting, or i can use a high vcore setting like this one.

anyways, let me talk a little about what i think on the CPU's core temp (as i've been seeing some were having some questions about it),
when running a test on your system don't relay on the core temp that are reading by any programs (likes core temp, real temp, speedfan, etc), instead, check and monitor the temp before the TJ Max.

the CPU's core temp reading on the programs is all upto you. if you are using a cool room for your computer system, then you may want to set a lower TJ Max value like 85-90C, or if your computer room is a very warm room, then you might want to set a more higher TJ Max value, likes 95C-105C. so when the system is in idle you may get a more accurate temp reading.

for an example, if i believe this correct, here's the core temp value during the system (with the 462 FSB overclocked setting above) is in idle.

real temp, core temp and everest shows the same CPU's core temps (when the same TJMax value is in use on all 3 programs).

swimtech · Jul 31, 2008, 09:38 PM

I think lowering the TJ Max setting might be dangerous in that the program would be badly underreporting temps at load - which is what most folks will think to read if they are not remembering to pay close attention. I don't understand why the program would report idle temps better though either, because the TJ Max setting doesn't actually change the lack of sensor accuracy at low temps. The difference in idle temp figures between cores is because of that lack of accuracy, and calibration is just supposed to equalize the inaccuracy across the differing core sensors at idle, isn't it?

It would take some retraining for most folks, I think, to ignore the temp figures and look instead at the Difference to TJ Max figure.

Do you think this is more accurate than going through the calibration procedure? I did that recently btw...

PangingJr · Jul 31, 2008, 10:55 PM

oh i didn't try to stop people to monitor their core temp reading.
what is your Intel officially TJMax value for your CPU anyway?

swimtech · Aug 2, 2008, 09:53 AM

Oh, of course - I know you're just doing your thing - it's neat to see too.

About the TJMax for the Q6600 - no one knows for sure and Intel hasn't yet published that figure for my processor. I use the TJMax figure that the author of RealTemp suggests because it seems he's done a decent analysis of what it likely is - he's tried very hard to be accurate. I use the 95C figure.

I've gone through the calibration procedure he outlines with a thermometer, and I have the advantage of having a stable room temperature of 23C (give or take just 1C). Note below that applying the calibration factors changes the Distance to TJ Max figures when, by my understanding, it really shouldn't. However, the idle temps shown are now evened out, and stay within one or two degrees of one another when the cores are evenly loaded during stress testing.

And, at all times, I just look to keep the max temp below 71C - even during stress testing - as outlined by Intel. Actually haven't ever exceeded 64C or so at any time so far as reported by RealTemp.

PangingJr · Aug 2, 2008, 10:26 AM

after i've already finished all the overclocking i don't really care about the core temp thing since i don't what CPU's core temp value i should be using either.

i think, this particular Q9300 system will be set at 462x7.5 since it doesn't use much voltage or gives much heat to the room at all.
when i actually use the system duing the days it's far more cooler, i never have any problem with system over heating, i'll probably reduce the case fans to a minimum.

here's some of abit more info about the TJMax and DTS...
TSK's Crazy Theory on TJmax and the DTS - Overclock.net - Overclocking.net

swimtech · Aug 2, 2008, 12:09 PM

Yeah, crazy - not. Nice read...

It really doesn't make sense from all I've learned and seen for the past 35 years that there would be much difference in Tcase and Tcore at idle. The IHS is designed to conduct heat very well and does its job, no doubt, otherwise Intel would be living in recall city!

About the descrepancy in Distance to TJmax in my screenshot, well, looks like unclewebb just hasn't gotten around to adjusting that displayed figure in his software with repsect to the entered calibration factors is all. TJmax should not change, it is a practical absolute for each processor stepping (well, withing manufacturing tolerances, which are extremely tight I would think...).

That was why I was wondering about why you were moving the TJmax figure around so much - I thought you might have found something but just didn't explain it yet...

Edit: And with the cooling you have available in that case of yours now your 3.465Ghz overclock looks like where I would settle too, and your volts - that's a fast system for sure!

PangingJr · Aug 3, 2008, 01:48 AM

so far the 462x7.5 is my favorite overclocking on this system, it give a good performance on both CPU/RAM.
475-480x7.5 (3.56-3.6GHz) didn’t seem to give much of a more CPU/RAM performance and the overclocking will require me having to use quite a bit more volts on the system, as well as, this can also means more heat that i have to control them.

right now the motherboard's DIMM slots are right beneath CPU cooler and i don't like that i'm not able to go into the DIMM slots without openning the whole front or top section of my computer case.
so i plan to replace the CPU cooler to something that is much smaller and i will do this when i decide to put in a new RAM kit in this system (these 2 RAM kits is needed for my P35-DS4 based board computer). i can't remove the whole board and rebulid the system with a new CPU cooler and RAM at this time around, i haven't gotten out to check on some new computer parts yet.

about the core temp, i just noticed that whether you like to believe and set any TJMax values on the software program, the Distance to TJmax were moving the same. as far as i know, there are users out there don't understand why the core temp figure on their CPU during idle are much higher than (or in some cases, lower than) the room temperature, i mean they already checked all the things, case fans, CPU's heatsinkfans, no overclock, and all.
so it's like the OP's post in that forum (on the very first post), he seems to believe the TJMax value for his CPU was more like 75C (not 85C or 95C). the lower TJMax values you set the lower core temp is reported by the software programs.

swimtech · Aug 3, 2008, 11:00 AM

Oh sure, I see you can set the TJmax and read the program's displayed Distance to TJmax that way - it's definitely consistent and relative.

I think, from that thread and unclewebb's input and all, what I got from it was that TJMax is really Tcase + 20 or 25c (I need to read it again, and the Intel specs, to be more accurate though...) for the desktop processors for which Intel doesn't publish the figure for.

You've found your sweet spot along the same criteria that I found mine. It was about real performance balanced against power usage/stress on the system for both of us I think. After a point, performance reaches a point of diminishing returns versus power usage. I'm thinking it would look like a so-called J curve when plotted on a graph. The key is finding the spot just before the plot turns upward exponentially.

We've found ours, and I noticed your voltage is almost .05v less than mine - which is almost exactly the same as the difference between Intel's own rated maximum voltages for our respective processors. Yours is 1.45v, mine is 1.50. How 'bout that!

PangingJr · Aug 3, 2008, 11:51 AM

i think just like the real temp owner has mentiond, that the Dts should not be use to report idle temperature,
and since i don't know for sure about the TJmax value on my CPU's i will have to use program like core temp and real temp as an overclocking tool only.

PangingJr · Aug 6, 2008, 05:04 AM

just succeeded in overclock Q9300 to 450 FSB using 4x1GB D9 based RAM modules and at a low vdimm voltage.

PangingJr · Aug 16, 2008, 04:18 PM

Intel Will Disclose DTS Specs At IDF - Tom's Hardware

Quote:

At IDF this year, Intel is expected to unveil (PDF) information that will lead to improved internal processor temperature monitoring. Intel has never previously disclosed information about the Tjunction value before of its 45nm desktop processors, a value that is needed to accurately measure the internal CPU temperature. Previously, these Tjunction values were estimated by software, but as each model and stepping of processor can potentially have a different Tjunction value, not to mention varying temperature response curves, no current method can be accurately trusted.

Intel is appealing to the enthusiast community by fully disclosing the information needed to now properly determine this information on day 3 of IDF. If you are at IDF this year, the course will be #TMTS001, and will start at 1:40pm. Regular computer users will still benefit from this released knowledge by soon being able to monitor accurately the temperature of each core in their current Core 2, when previously unable to do so. Serious overclockers may find this particularly useful.

Previous Intel chips, prior to 45nm, were capable of measuring internal CPU temperature through an on-board diode and was measured externally. With the 45nm processors though, a new method was needed and so Intel engineers came up with the Digital Thermal Sensor, or DTS. Instead of relying on external measurements of the temperature, Intel had integrated that aspect into the processor. By using a thermal threshold value, the Tjunction, along with the data provided by the DTS, the internal temperature of the processor can be determined. Tjunction values in the past have only been provided for mobile processor versions, as they lacked an integrated heat spreader.

Look for our complete coverage of IDF next week.

Intel Developer Forum (IDF) 2008 to Include Presentation on 45nm Core 2 DTS Specification (Link: AnandTech )

Quote:

We met with Intel today and have some interesting IDF (Intel Developer Forum) 2008 news. IDF is scheduled this year for August 19 - 21 in San Francisco. It is shaping up to be an exciting event for those tech geeks like us as X58/i7, Larrabee, and System on Chip (SoC) technologies will be discussed in detail among other things. We learned this afternoon that full disclosure of Intel's existing 45nm processor Digital Thermal Sensor (DTS) specification will be presented on Day 3. This 50-minute technical presentation on DTS (course #TMTS001) will start at 1:40 pm. A complete schedule of events can be found by visiting Intel’s official IDF website.

Benson Inkley, a senior power/thermal engineer with Intel, is prepared to address nearly every aspect of DTS functionality for the attendees. However, perhaps the biggest surprise to come out of his presentation will be the first-ever public disclosure of the maximum Tjunction value for all Core 2 Duo/Quad/Extreme desktop processors built on current 45nm-process technology.

Armed with this information, seasoned application developers and amateur coders alike will finally have everything they need to implement the most accurate, real-time core temperature display tool possible. We discussed this topic in our Intel Core 2 Duo E8500 review back in March 2008, except our discussion left quite a few unanswered questions flapping in the breeze. Come next Thursday, anyone walking out of Mr. Inkley's technical session will have all the knowledge needed to lay any longstanding DTS questions to rest.

We applaud Intel for recognizing the enthusiast community’s interest in this subject. Overclockers, performance enthusiasts, and everyday users will finally be able to monitor their CPU’s individual core temperatures without wondering if the reported results are accurate or not. Check back here on Thursday, August 21, for a detailed update on DTS.

PangingJr · Aug 16, 2008, 04:34 PM

The Wolfdale Vtt ( FSB Termination ) game.Place your bet! - XtremeSystems Forums

Quote:

So I'm back in action with some CPUs designed to... die
We'll start with a Core 2 Duo E8500 "Wolfdale" core 45nm processor and the Asus Striker II Extreme.
This system will be my...temporary Call Of Duty 4: Modern Warfare, Crysis, and Rainbow Six 3: Raven Shield online gaming rig for this... wee... wait, 'till it dies

I'm about to set the system up now, and I'll start with 1.6V FSB Termination Voltage ( VTT ) and a "reasonable" overclock with a ThermalRight Ultra-120.

According to the... VTT Myth anything over 1.4V is deadly.
How long do you think it'll last ?

I'm going to leave the PC running Folding@Home GPU ( and CPU maybe ? I have to check the CPU utilization from the 2 F@Hgpu instances first ) 24/7.
When I can/want to game I'll pause the F@H's and game on.
When I finish gaming the F@H's will be up and running again.
So this system and the CPU particularly will be fully utilized ( 100% ) 24/7.

If the CPU survives more than 7 days at this FSB Termination voltage I'll move on to 1.7V on the 8th day.