Real Temp - New temp program for Intel Core processors

[unclewebb]

No, im serious there is something wrong with the middle cores in your program. Its not how long the system is up. Mine has been up for Weeks and it still says 16C.

I think what Jason was asking was if you are using a calibration factor in RealTemp. This is where there seems to be errors in the data coming from the on chip digital thermal sensors. Without any calibration factor, temps below ambient or below your water temp are typical. If you can, post a screen shot of RealTemp with your system at idle. Data from someone that knows their system temps inside out is just what I need.

bryanw1995: I don't think that tjunction is 100c on a Q6600. I don't know if realtemp has it right at 85c, either, unfortunately...

There are two different Q6600 processors. RealTemp uses 85C for TjMax for the B3 stepping and 95C for the G0 stepping. The best thing about RealTemp is that there is no reason to be sad. If you have done your own testing and you know what TjMax is for your processor then you can set TjMax to that value. If RealTemp is assuming TjMax=85C for your processor and you know it is 95C then go into the RealTemp.ini file and do this:

TjMax0=2
TjMax1=2
TjMax2=2
TjMax3=2

This bumps TjMax by two 5 degree steps for each of your cores. If you believe TjMax=100C then bump it three steps. Problem solved.

Check the Thermal Specification for the SL9UM Stepping B3:
http://processorfinder.intel.com/det...px?sSpec=SL9UM

and here is the same Q6600 SLACR Stepping G0:
http://processorfinder.intel.com/det...px?sSpec=SLACR

The B3 seems to have more in common with the other B2 stepping dual cores which are all TjMax=85C. Here is an E6600 B2:
http://processorfinder.intel.com/det...px?sSpec=SL9S8

I also seem to remember when the original Quad came out that people were complaining that with the retail heatsink and fan that they could get up to the throttling point without too much trouble. Show us some idle testing at low core voltage and low MHz so I can get this corrected if I have made a mistake.

[IanB]

maybe you need to evaluate how you deliver the message about your objections or recommendations...

putting it to pen is a whole lot different than speaking it out loud to the individual(s)

i'm not trying to flame you, just give you a nudge in the right direction

1. I made it clear I was impressed with the app and unclewebb's work and I've actually recommended this app previously in other threads here, so I've got nothing to prove about my intentions in posting what I did. I'm not responsible for people misinterpreting my words, whether accidentally or wilfully.

2. The comment I made was about a specific change to the interface that was going to make it much harder for people to use, and I have justified my opinion with some general comments about UI design, something I know a little more about than most here, I'd guess. I build Windows apps and PHP websites that rely on clear interfaces and good design to make them easily accessible as well as functional.

3. My only interest is in helping improve the app so that it requires less support from unclewebb and is easier to use, which obviously benefits all users, not just me. Anyone who suggests I have an ego problem in making a constructive comment like that is simply an idiot.

4. unclewebb has already accepted my constructive criticism in the spirit in which it was intended, as a helpful comment from one app author to another. Anyone else's opinion on the matter is frankly irrelevant.

5. You've gotta laugh. The "ego" poster is the one with 1000s of posts to his name merely celebrating new releases of app software with better features and/or design - yet he flames (lamely) someone who is trying to help improve THIS software, implying improvements aren't important. What a du... well, I don't have to say it.

Any further comments or suggestions I make will be in private. Unlike some people here, I can actually contribute practically to this project in ways that would benefit everyone, as unclewebb is aware, but funnily enough my ego doesn't need the credit for that.

[lowtcell]

unclewebb
Thanks for the new version, great work. Some results -

1600 idle


3400 idle


3400 load

[HeinzS]

Hello,

here are my Idle temps(picture1) on a Asus P5B with Artic Freezer 7 Pro, Room temp are 19°C. CPU+Cores temps rises about 10°C with BIOS 1803 on on my old E6600, so Speedfan has modifed temps for CPU and Cores. Hoping Asus will correct the BIOS shortly.

Picture 2 shows a anormaly CPU temp which i get sometimes.

[axis]

Hello,

here are my Idle temps(picture1) on a Asus P5B with Artic Freezer 7 Pro, Room temp are 19°C. CPU+Cores temps rises about 10°C with BIOS 1803 on on my old E6600, so Speedfan has modifed temps for CPU and Cores. Hoping Asus will correct the BIOS shortly.

Picture 2 shows a anormaly CPU temp which i get sometimes.

Why are you running a 3.0ghz E8400 at 1.2ghz? Would you happen to have speed stepping on?

Axis

[ChrisZ]

Dayam i wish you had a yorkfield, you'd probably understand what i mean.

I have and I can confirm (QX9650, C0).

The system is pretty much like yours, H2O with a *huge* Reserator 2 radiator.

Left the system unpowered overnight and the whole thing including 1 litre of coolant (+ 7kg radiator) was 20 C at the morning. Booted to idle: core 2 and 3 read 10 C in RealTemp. That is 10 grades below ambient.

Now I'll PM my log to uncle :-)

[jason4207]

Why are you running a 3.0ghz E8400 at 1.2ghz? Would you happen to have speed stepping on?

Axis

Read the first post (and my post just a few posts above yours). This is done to properly calibrate idle temps.

[unclewebb]

lowtcell: Your idle temps look appropriate. Can you re-run your load temps, preferably with Prime small FFTs? The link for that data is missing in your above post.

HeinzS: I also have a P5B Dlx and I can confirm that the CPU reported temps for that board are sometimes out to lunch. Your E8400 idle core temps look a little too high which is why my E8400 needs to run a Idle0=-2 correction factor on both cores for better accuracy. You also might have the sticking sensor issue where they simply won't go any lower than a fixed value even if you move to the North pole.

[axis]

Read the first post (and my post just a few posts above yours). This is done to properly calibrate idle temps.

Sorry, missed your post completely. I semi-sorta understand now.

Axis

[fng77]

Got pretty high results. E2160 rules wins over the E8400 with about 30000 points.

[HeinzS]

@unclewebb
my P5B is nondeluxe and with Idle1=-2 there are -3°C to core0. The sensors go lower, Realtemplog: "07:17:30 37 37" this morning directly after turn my PC on with Core speed 1200 MHz.

With BIOS 1803 there are no more the PECI feature enable/disable and in the past I get the same wrong temps with old BIOS with my E6600.

[unclewebb]

axis: Just to put things into English. What my low MHz / low voltage test is designed to do is to lower your core temperatures as low as they can possibly go. There are 101 variables that will effect your load temperatures but at Idle, if you can control these two variables, you should find that your reported Idle temperatures are a little above your room temperature. The digital thermal sensors are calibrated by Intel to trigger thermal throttling at TjMax and generally speaking they are very accurate but as you move farther and farther away from TjMax, the sensors no longer move at the same rate as changes in your core temperature. Read the explanation in post#1 and let me know anything that doesn't make sense.

ChrisZ: Which brings me to your data ChrisZ. Your sensors are doing a great job of proving the point that they shouldn't be fully trusted at idle.

Looking through your data the first thing I notice is that core1 is stuck at 22C and can't go lower. The next thing I look for is two cores that are running mates. By that I mean that when you apply a Prime load and then go back to idle the two cores will track each other very closely. This happens with core0 and core1 once the temperature goes above 22C and core1 becomes unstuck.

I think with a +2 Idle correction for both of these cores that you should get some reasonably believable and accurate data. Core0 will be very accurate from idle to TjMax and core1 will be just as accurate as soon as it gets above the sticking point. It is presently sticking at DTS=73 so whenever the DTS goes lower than this number on core1, it should be fine.

Core2 and core 3 are a bigger mess. Here's some data from the log you sent:

11:27:52 25 25 13 17
11:27:53 29 29 18 22
11:27:54 30 30 19 23
11:27:55 31 31 20 23

I have some faith in core0. In the above data, core2 is reporting 11C to 12C lower than that.

11:27:24 16 22 10 10
11:27:25 15 22 10 10
11:27:26 15 22 10 10
11:27:27 15 22 10 10

In this data core0 is still moving freely but I think your other 3 sensors are stuck. If core2 wasn't stuck at 10 I think it would be showing readings about 11C less than core0 which is about 4C. I've seen an E8400 with similar issues where one core wasn't just too low at idle but way too low and was even reporting negative numbers though the user was only air cooled.

I usually don't have 100% faith in these sensors until core temps are over 60C and because you are water cooled you are a long way from that. Below this point, some sensors can be slightly manipulated to produce some reasonably accurate temperature data but the sensors on core2 and core3 of your Quad don't fit that category.

For core2 and core3 I would bump TjMax to help make the numbers a little more accurate even if they're not ever going to be 100% accurate.

TjMax0=0
TjMax1=0
TjMax2=1
TjMax3=1
Idle0=2
Idle1=2
Idle2=2
Idle3=2

If you were air cooled and had some data up near 60C then there might be a better set of correction factors for your cpu. For your situation, I think the above factors is your best compromise for accuracy in the range you operate.

Core3 will track core0 and core1 much closer and core2 won't be too far behind.

fng77: Congratulations on your super E2160! The problem with my original benchmark is the timer I used is totally proportional to boot MHz so if a user used SetFSB after booting up, it wouldn't properly show any improvements in the benchmark score. My replacement timer may have been good in theory but obviously I need to find something better. Thanks for the heads up. Does it always report that score or is it normal most of the time?

Here's three rounds of XS Bench run simultaneously on the Quad. It works most the time.

[lowtcell]

lowtcell: Your idle temps look appropriate. Can you re-run your load temps, preferably with Prime small FFTs? The link for that data is missing in your above post.

I don't know what happened to the link there Here it is again

[fng77]

axis: .



fng77: Congratulations on your super E2160! The problem with my original benchmark is the timer I used is totally proportional to boot MHz so if a user used SetFSB after booting up, it wouldn't properly show any improvements in the benchmark score. My replacement timer may have been good in theory but obviously I need to find something better. Thanks for the heads up. Does it always report that score or is it normal most of the time?

I did another test shortly after and then I got 10000. I´m gonna do one right now...

Now I got: 1060 witch should be normal for my system.

[unclewebb]

I can't remember if I changed anything in the timing calculation between version 2.3 and 2.4 so try downloading that and see how things go. The benchmark is small and not effected by size of cache or memory timings so your E2160 should run equal to much more expensive processors but it shouldn't be running 20X faster.

[aoch88]

I know my chip's sensor is broken. Changing from air to water would still get me 44C idle on Core Temp but with this Real Temp however, it's 10C lesser (34C)

I guess the main improvement which Real Temp could do is to make it sit on the tray icon like MBM, Core Temp, etc.

[unclewebb]

aoch88: RealTemp and the competition assume a different TjMax for some processors. All programs probably get it right once in a while! Tray icon for RealTemp is still on the things to do list. Accuracy is still the most important thing for most users.

[ChrisZ]

ChrisZ: Which brings me to your data ChrisZ.
If you (...) had some data up near 60C then there might be a better set of correction factors for your cpu.

Everything for you
This shouldn't be impossible to achieve, it's just a matter of putting a large pillow on the radiator

I'll do it later today.

Edit: Done. Plz check your PM.

[fng77]

I can't remember if I changed anything in the timing calculation between version 2.3 and 2.4 so try downloading that and see how things go. The benchmark is small and not effected by size of cache or memory timings so your E2160 should run equal to much more expensive processors but it shouldn't be running 20X faster.

In version 2.3 I always got realistic results, this problem started to occur since I upgraded to 2.4

[davejohn]

In version 2.3 I always got realistic results, this problem started to occur since I upgraded to 2.4

Same here:

[lowtcell]

I just ran it and got 1147 Seems to work ok.

[unclewebb]

I'll have a look at this and go back to the old timing system if I have to. It used to be very accurate as long as you did all your overclocking from the bios but the stop watch I was using also got overclocked when using SetFSB from within Windows. I'm sure I'll find a way to get it working consistently no matter how you like to overclock.

ChrisZ: Here's some of the data you sent me so others can follow along.

00:36:42 54 54 45 48
00:36:43 54 53 44 48
00:36:44 53 53 45 48
00:36:45 54 53 45 48

As before, core0 and core1 are tracking each other very well but there is something strange going on with the other two. During the last couple of pages a few of us have noticed the exact same thing with our Quads. In theory, 4 cores all running the same MHz and core voltage and doing the exact same thing running Prime should all be very close to the same temperature. We started thinking that maybe Intel is setting TjMax 5C higher on core2 and core3 but not just on your Quad.

I pushed my Quad much hotter and this is what I got:



An exact 5C difference as I was approaching TjMax. Idle calibration has no effect at this temp.

Then there is Ace-a-Rue.
http://www.xtremesystems.org/forums/...&postcount=632

An exact 5C difference on his QX9650 at full load.

and then there is Nevrsadie. He re-did his heatsink and he started getting the same thing. His cores are presently the same in this screen shot but look at his maximum temps.

http://www.xtremesystems.org/forums/...&postcount=655

There is definitely a pattern here but why?

Just when I thought Intel might be messing with our minds by using two different TjMax values in the same Quad, I went back and had a look at some of my original testing which I posted here:

http://www.xtremesystems.org/forums/...&postcount=568

At idle and without a heatsink attached to my Q6600, core0 and core1 are tracking each other exactly and core3 is now only 1C less which is very normal.

The act of tightening the heatsink or water block might be causing uneven pressure on the IHS and the two dual cores that lay beneath it. Does excessive or uneven pressure on the IHS effect the data coming from the on chip DTS sensors? Definitely an moment. I'm generally the type to find a big pipe wrench when it's time to snug down a heatsink so maybe I need to take it easy next time. Next test; run Prime with a loose heatsink and snug until the temps go a little wonky on core2 and core3 then back off half a turn on each screw and it should be perfect.

[ChrisZ]

The act of tightening the heatsink or water block might be causing uneven pressure on the IHS and the two dual cores that lay beneath it. Does excessive or uneven pressure on the IHS effect the data coming from the on chip DTS sensors? Definitely an moment.

You are definitely onto something here.

When reading Intel white papers for my CPU, I have noticed that they are very explicit on the pressure in IHS, it is exactly defined there in Newtons.

Now to my CPU water block (Zalman ZM-WB4 Gold). It comes with several mounting accessories, one of them is for my socket 775. When I initially tried to assemble it I discovered that the springs on the water block would exercise the pressure on CPU of several orders of magnitude higher that stock cooler ever did. It would literally press the CPU right through the mainboard to the other side I guess.

My suspicion is also further reinforced by the fact that all mounting screws seemed to be way too short to reach the base of the mounting socket unless springs were loaded really hard.

So I have adjusted the springs (by bending them with tongs) and the pressure on the IHS became much, much lighter.

Another observation I made was in the course of reseating of the water block (because initially when I saw my core temperatures I thought they must been result of a seating mistake on my side) I noted also that the goop (Arctic Silver) was distributed unevenly, which indicates that IHS is not exactly planar.

However I hesitate to lap a $1200 chip with sandpaper just to see what possibly this could change

All in all, the physical pressure applied on IHS may be the contributing factor, especially when only stock coolers seem to exercise the correct force right out of the box...

[unclewebb]

I'm using a Tuniq Tower which uses 4 thumb screws into a base plate on the bottom side of the motherboard. I generally tighten until all 4 screws have bottomed out at the end of their threads but now I have serious doubts if this is too much pressure and causing some of the issues we're seeing.

The H bracket that rests on the top side of the heatsink base plate gets bent after a good snugging. I've also read the Intel docs and know how specific they are about contact pressure so there is likely a good reason for that. I might need to try and find some hard, non-compressible rubber to mount between the bracket and the motherboard to prevent excessive force.

While running Prime, I'm going to monitor the temps and then back off the screws one turn at a time to reduce the load on the IHS and wait to see what happens.

[ChrisZ]

While running Prime, I'm going to monitor the temps and then back off the screws one turn at a time to reduce the load on the IHS and wait to see what happens.

Are you using a goop (aka thermal compound) flexible enough to accomodate the change? IHS will probably undeform slightly I guess...