Phenom II 1100T, WTF throttling?

[MadDias]

If you look, the invividual core temps are reading 30c at the time I take the screenshot...and they were never over 47c themselves. I can't see the CPU being over 60-62c.



Obviously you aren't sure what you are talking about.

obviously u don´t know where the 57°C cpu max temp came from. look again, it´s in ur everest screenshot. this may be the temp where throttling kicks in.
u can´t see the cpu over 62°C cause u will never see it.
the DIE temp sensor is way off by about 10-15°.
no way my cpu can run cooler than the water cooling it, even when coretemp says so.
try to stay under 60°C and u´ll be fine.
to make it easier for u: when coretemp reaches 55°C u will expierence throttling.

[BeepBeep2]

obviously u don´t know where the 57°C cpu max temp came from. look again, it´s in ur everest screenshot. this may be the temp where throttling kicks in.
u can´t see the cpu over 62°C cause u will never see it.
the DIE temp sensor is way off by about 10-15°.
no way my cpu can run cooler than the water cooling it, even when coretemp says so.
try to stay under 60°C and u´ll be fine.
to make it easier for u: when coretemp reaches 55°C u will expierence throttling.

But it throttled again at only 51/52c when I lowered voltage to 1.44v and tested again.

[schoolslave]

So it only happens at a certain voltage?
It must be the mosfets then.
Anyone know if this happens with a CH IV as well?

[demonkevy666]

This CPU came with a 1.325v VID, I don't believe it's so bad tbh. I'd like to solve my throttling issue or I will end up running 3.8 @ 1.3v or something on my water loop.

certainly better then my chip for air.

Mines Vid is only 1.275 a whole 0.050 lower then yours

since I've already got a thread open...
Good news:
3.9 Ghz core/3010 NB is an hour stable at 1.39v (set 1.35v) core and 1.315v CPU_NB with a max CPU temp of 48c. I didn't see it throttle any.

Bad news:
PC Probe II just scared the $h^t out of me, telling me my 3.3v spiked to 4.81 and that my 12v was reading .11v with a nice loud alarm. Good one ASUS, but it's not april fools day yet

I've got a volt meter on my case and I've seen my 5.0 V drop to 4.0 -.-

Beep, are you running those Super Talent Hypers?

If you keep having problems with stabillity (like I have) you may want to swap them out for a different kit....

I put my old BBSE kit tonight because the problems always seemed ram related, and it's like a whole new animal!
Much more co-operative.

Priming right now @ 218x19x14 (4150/3050) @ 1.45 and 1.25 respectively.

mmmmm I love my 8Gbs Flares never any problems lol

lol...

Dave this isn't a ram problem...my CPU is throttling down to 800 Mhz under load with any voltage higher than 1.40v and 4 Ghz. I've proved 1845 6-6-6-18 stable on these sticks...

was that with the X4 ?

Temperatures! I'm sure you have a bad mount, my ambients here are mid 20's and at 3.9GHz with 1.495v (WCG puts a lot of stress on the chip, less voltage causes a crash after a few days) with a TRUE Black and two Gentle Typhoons at 1450RPM in push pull I'm reading 35'c at full load after... Current uptime is seven days. Bring to borked sensor into the equation and I'm sitting around 50'c.

I'd much rather see the limits of my cores before my NB adding the NB can add quite a bit of heat.

But it throttled again at only 51/52c when I lowered voltage to 1.44v and tested again.

off set core temp then.

[BeepBeep2]

Guys I'm currently taking a break from this thread and going outside to bench some. Results will be up on a new thread tonight.
I already got through 1M at 4800, Seems Core #0 and Core #2 are weak, they were good up until about 4700. Didn't test 3-5 yet. The CPU passed 3DMark06 at 4480, I was able to hit 26,200.

@ demonkevy666:
Yes, that was with X4, though I just benched 3D and SuperPi 1M cold with a 3250 NB and 1860 6-6-6-18. As far as the VID issue, I'm not sure which is actually better, yours is leakier which would theoretically end up in more OC headroom but also more heat...

[daytripper67mi]

I was pretty sure we aren't allowed to flash back to earlier bioses anymore(?)

Afudos allows flashing to older bioses,I use afudos229es if you need a copy let me know.

[demonkevy666]

I actually think this link to the PPL on the board.
Mines doing it too but it's no where near 62C.
I've gave 15C off set on core temp and I'm only hitting 54C right now.
the HTT fluctuation maybe also affecting it.
as it looks like a clock skip more so then a Throttling completely.

[BeepBeep2]

I actually think this link to the PPL on the board.
Mines doing it too but it's no where near 62C.
I've gave 15C off set on core temp and I'm only hitting 54C right now.
the HTT fluctuation maybe also affecting it.
as it looks like a clock skip more so then a Throttling completely.

I'm sure you saw the video...

If I push voltage up it will actually throttle for lengthened periods of time...

[mav2000]

Like I said before, this is a board issue with the mosfets over heating. I spoke to Asus and they CONFIRMED that it is a feature to make sure you dont blow the board. This does not happen on the CH IV or the CH IV extreme. It does happen to all M4A line up though. I have tried the M4A with a 965 BE and I did not have any problems what so ever. with a 1090T this problem did come up multiple times and I tried every bios setting and windows setting, but could not solve it. As this was a review sample I spoke to the guys at Asus and this is what I was told.

[BeepBeep2]

Like I said before, this is a board issue with the mosfets over heating. I spoke to Asus and they CONFIRMED that it is a feature to make sure you dont blow the board. This does not happen on the CH IV or the CH IV extreme. It does happen to all M4A line up though. I have tried the M4A with a 965 BE and I did not have any problems what so ever. with a 1090T this problem did come up multiple times and I tried every bios setting and windows setting, but could not solve it. As this was a review sample I spoke to the guys at Asus and this is what I was told.

Okay.

Looks like I'll wait out until ASUS releases it's 990FX board then make a move.

[Dirk Diggler]

I am seeing the same throttling also on my 1100T with the CIV Formula bios 1304

[VoodooProphetII]

I have not seen this issue running a 1090T on a Crosshair IV Formula with the 1304 bios update. I'm using a True Black and I've run Prime95 small FFT's for about an hour at 1.38V without any problems. When I try blend I'll randomly get two cores failing after 5 to 10 minutes. Never the same cores and it does not matter if I drop the memory speed and NB frequency to stock it will still have one or two cores fail after 5 to 10 minutes.

[BeepBeep2]

I have not seen this issue running a 1090T on a Crosshair IV Formula with the 1304 bios update. I'm using a True Black and I've run Prime95 small FFT's for about an hour at 1.38V without any problems. When I try blend I'll randomly get two cores failing after 5 to 10 minutes. Never the same cores and it does not matter if I drop the memory speed and NB frequency to stock it will still have one or two cores fail after 5 to 10 minutes.

Temps? That will help our cause, because I can run 1.375v-1.385v without throttling, 1.4v w/ LLC = 1.44v load and this causes the board/cpu to throttle down again.

If your temps are quite a bit higher than ours then maybe we are getting somewhere.

[chew*]

I think many of you are just victim to the hype of how fantastic windows 7 is.

Read

http://blogs.technet.com/b/askperf/a...oalescing.aspx

Fix it like this.

- Go to Regedit

- Find this key:- " 0cc5b647-c1df-4637-891a-dec35c318583 "

- Within this key, there is a value called: " ValueMax "

- This value represents the % number of cores the system will park - the default 100% ie: all Cores are potentially park-able

- Change the value from 64 to 0 so the " ValueMin " and " ValueMax " are both zero

- You will have to find the key a few times and repeat the process for each time it is found - the number of instances will depend on the number of power profiles in your system [ in my DAW it was only found twice ]

- Do a full shutdown and power-off and cold-re-start

[BeepBeep2]

But chew*, I run Win Vista SP2.

[demonkevy666]

But chew*, I run Win Vista SP2.

same here :/

[I.nfraR.ed]

My 3DMark scores suddenly dropped wuth 1304 while benching for the CountryCup, I guess because of throttling. I couldn't flashback to 0905, instead managed to flash 1102. Will try to put back 0905 with afudos, but I'm not sure if there's any version which will works.

[chew*]

hmmm,

read further up on it anyway......it seems it only applies to logical cores aka hyperthreading or multi socket. Might want to verify its not doing it whoever is using win7 and thuban however.

I dunno, guess I will drag the board out and see what I can make of it.

First I need to recover from this nasty,flu, virus, pneumonia I got while in CES though.

[Dirk Diggler]

I think many of you are just victim to the hype of how fantastic windows 7 is.

Read

http://blogs.technet.com/b/askperf/a...oalescing.aspx

Fix it like this.

- Go to Regedit

- Find this key:- " 0cc5b647-c1df-4637-891a-dec35c318583 "

- Within this key, there is a value called: " ValueMax "

- This value represents the % number of cores the system will park - the default 100% ie: all Cores are potentially park-able

- Change the value from 64 to 0 so the " ValueMin " and " ValueMax " are both zero

- You will have to find the key a few times and repeat the process for each time it is found - the number of instances will depend on the number of power profiles in your system [ in my DAW it was only found twice ]

- Do a full shutdown and power-off and cold-re-start

Even though I am running Windows 7 (64 bit) this wasnt happening with bios 1102 and 055. Started happening after the flash to 1304

[BeepBeep2]

Mats offered me the 64-bit iso he has from MSDNAA, I'll see if it does it on that after I reinstall. I'm betting it will, but I dunno.

The only time I had this 1100T in was after updated my M4A89TD to bios 1101. Unfortunately I don't know if it had this problem on earlier bioses.

[equinox1]

a bit of advise for what it's worth beepbeep based on my expierience with this board and a thuban is your temps. way too high . if your over 45c on cpu and running 2 5770 in xfire in a case with water a thuban is not going to like it for long. 4.1 or 4.2 in prime is not going to work. maybe if temps were at 20c but not in the 50's. nb is cooking because this board is not a high end board. poor cooling especially considering that you have poor air flow over the nb running water on the cpu. take the board out of a case and pump air at it to keep board temps at ambient and it might be better . i know , not practical or what you want but that's what it takes sometimes. whens the last time anyone saw chew overclock something in a case. dont try to push for something that defies principals of science, adapt to achieve them. just my perspective.

[chew*]

Even though I am running Windows 7 (64 bit) this wasnt happening with bios 1102 and 055. Started happening after the flash to 1304

well early on we had OCP issues, that got adressed.

anyone tried killing all temp and voltage monitoring?

Equinox is right to, if in a case I wouldn't even bother pushing over 1.475vcore or 1.25-1.3NB.

I know the limits of the cpu's and thats what I usually suggest for 24/7 useage and also make bios profiles for.

3.8 is a nice mission critical speed to shoot for least on air or equivalent beyond that is luck of cpu, and better than average or "annoying" fan cooling.

[Dirk Diggler]

well early on we had OCP issues, that got adressed.

anyone tried killing all temp and voltage monitoring?

Equinox is right to, if in a case I wouldn't even bother pushing over 1.475vcore or 1.25-1.3NB.

I know the limits of the cpu's and thats what I usually suggest for 24/7 useage and also make bios profiles for.

3.8 is a nice mission critical speed to shoot for least on air or equivalent beyond that is luck of cpu, and better than average or "annoying" fan cooling.

That was the next thing I was going to try. One benefit of 1304 is I actually was able to lower my voltages for the same clock with 1102 I was running higher voltages and higher temps 53-55 C and it wasnt throttling. I will continue testing to see what I come up with just thought it was weird the new bios started doing this outta the blue. Maybe it was something that was left unchecked in the previous versions of the bios

[equinox1]

That was the next thing I was going to try. One benefit of 1304 is I actually was able to lower my voltages for the same clock with 1102 I was running higher voltages and higher temps 53-55 C and it wasnt throttling. I will continue testing to see what I come up with just thought it was weird the new bios started doing this outta the blue. Maybe it was something that was left unchecked in the previous versions of the bios

or possibly it was a bios setting that the manufacturer implemented to prevent unwarrented damage to a product that that was run beyond spec. none of us do that , do we ?

[BeepBeep2]

a bit of advise for what it's worth beepbeep based on my expierience with this board and a thuban is your temps. way too high . if your over 45c on cpu and running 2 5770 in xfire in a case with water a thuban is not going to like it for long. 4.1 or 4.2 in prime is not going to work. maybe if temps were at 20c but not in the 50's. nb is cooking because this board is not a high end board. poor cooling especially considering that you have poor air flow over the nb running water on the cpu. take the board out of a case and pump air at it to keep board temps at ambient and it might be better . i know , not practical or what you want but that's what it takes sometimes. whens the last time anyone saw chew overclock something in a case. dont try to push for something that defies principals of science, adapt to achieve them. just my perspective.

Thanks for the advice. Btw, I offset core temp by 15c and got 56c load, I suppose that's what I'll go by. Still, I have a 105CFM Panaflo going right over the PWM/northbridge...

It is an 890FX mobo, so maybe while not considered high end like the ROG boards its still upper midrange...

well early on we had OCP issues, that got adressed.

anyone tried killing all temp and voltage monitoring?

Equinox is right to, if in a case I wouldn't even bother pushing over 1.475vcore or 1.25-1.3NB.

I know the limits of the cpu's and thats what I usually suggest for 24/7 useage and also make bios profiles for.

3.8 is a nice mission critical speed to shoot for least on air or equivalent beyond that is luck of cpu, and better than average or "annoying" fan cooling.

Well I'm running a 3x120 rad with 1650RPM fans...basically I'm comfortable up to 1.5-1.52v vcore and 1.30-1.35 NB. I know it's a little higher than what you suggest for air but then again I was running 1.60v/1.45v on my 965 for extended periods of time. (Load was still only about 52c on the 965 at those volts)

or possibly it was a bios setting that the manufacturer implemented to prevent unwarrented damage to a product that that was run beyond spec. none of us do that , do we ?

...and I believe this solves what was going on with MSI boards+Thuban which were fires and explosions correct?

I've been able to get 4 Ghz at 1.404v so far...but what I dont understand is the mobo temp sensor.

Constant 30c except when I took it out in the cold where it sunk all the way down to 5c.