I wonder what Dell changed in your system and what might have been defective. Do you have an idea?
I think, the real difference came with new heat pipe and new hdd (hitachi instead seagate - hitachi is 4C (celsius) cooler). My "old" system (old MB and HDD) with closed lid and docked was so warm, that you couldn't keep your hand on the keyboard for long. Current system is not so much warmer, than older 610 system under same conditions (dock, closed lid, ext. monitors).
i wonder if there is any visual clues to this new heatpipe. could you maybe post a good pic of thsi new heatpipe here?
I got no pictures unfortunately, however, there are two heatpipes for the E6400. One for the Intel GPU and one for the Nvidia GPU. The latter one is a bit thicker. When DELL replaced my heatpipe, they have installed the bigger one.
More news about the BIOS thermal table modifications:
Today, I received a phone call and e-mail from DELL Global Commercial Support Services department. The issue has been narrowed down. A new BIOS is currently in test. They expect publishing it till end of this month. I've asked for a beta for testing and hope to receive one. It is also quite likely that it will be for all E-series Latitudes.
Hi. First of all, please excuse my rusty english. And let me say "thank you" for this thread. I nearly went crazy with this problem. Thought it was a Windows problem. Good to see I'm not the only one.
I found a solution / workaround for this problem. A special thanks to "tinkerdude" and his great pdf.file showing what happens at what time. It lead me to the fact that the problem has to be the temperature limits in BIOS and / or WinDrivers. They are to narrow. What happens here is all right but should happen at ca. 60/70 °C and not at 45°C. In my oppinion a BIOS and / or driver update should fix this very easy. DELL just have to set higher (much higher?) limits. The CPU used in the Dell Latitudes should survive much higher temp. without any damage. I really wonder who set limits to 45°C.
Today I've spoken with the DELL service. They were friendly and keen to help me. But finally they say the problem is well known, and an update will be released. In near future - perhaps September or later.
For the time till the release of such an update I want to present you my solution / workaround to this problem. I own a Dell Latitude E6400 with a 2.26 GhZ Dual Core and the following worked fine for me :
WARNING ! You should know / understand what you are going to do ! If you don't - better wait for a bios update. You could cause damage to your Latitude !
How to use ATSIV and RMClock together : read this Thread : http://forum.rightmark.org/topic.cgi?id=6:600-2 . -> (Rember you have to start the created vbs file with Administrator Privileges.)
Once you started RMClock you can manually change a lot of things.
AGAIN : BE CAREFUL : DO YOU UNDERSTAND WHAT ALL THE OPTIONS AND VALUES MEAN ? If not - better wait for an BIOS update. It's your own risk.
I disabled the Dynamic FSB Frequency Switching, chose a Profile, enabled it and configured the chosen Profile with 100% throttle, 8x VID and 1.1250 V (fits for my 2.26 GHz Dual Core). Applied.
Now RMClock has controll of all the CPU powermanagement including speedstep, throttle etc. It doesn't matter what BIOS or Windows want to do - RMClock decides.
Voila - now I have a 2.26 GHz CPU and not a 700 MHz one all the time. Temperature is fine. (RMClocl shows you the current temperature).
Just remember : From now on Windows has less / no controll of your CPU untill you quit RMClock. There will be no powersaving etc. unless you do it your own - by creating a Powersaving Profile inside of RMClock.
So, I hope I could help some of you.
Neither the less I hope DELL will bring out an update soon... The above is just an unaesthetic workaround with losing comfort and not a final patch. But at least I can use my Latitude now.
now that you told me I found your old posting concerning RMClock. This thread grew big ... hard to read everything :)
In your old post you wrote you lower the core voltage, right ? That's an option, too. Lowered core voltage ist the best way to lower the temperature. But in our case I think it's not neccessary. There is no (damage causing) overheating but wrongly set limits. So what I do is setting everything fix to its standard values. Means, I choose the standard multiplicator, standard voltage and turn throttling and DFFS off. From now on CPU and FSB aren't under- or even overclocked. They just work at their standard settings.
The deactiviation of the DFFS Function should solve your remaining Problem. The RMClock CPUsettings just avoid throttling oder speedstepping the CPU, but the FSB will reduce its speed neithertheless. Once you stopp the slowing of FSB you should have full power in all states. When having tried this setting, can you give me your feedback ? I'm interested if it helped you out.
Still I have another question to everybody here : Is someone using Windows XP and having the "temperature problem",too ? Just used Windows XP for some time, and it seems to me didn't have this problem, at least not that bad... I really ask myself, whether DELL has to update the BIOS or a simple Vista / Windows 7 driver could be the solution. As RMClock shows you, all this can be controlled from inside windows... Or even worse : When it runs fine without any modifications using Windows XP but really doesn't using Vista / 7 is Windows Vista superior to BIOS in this case ??? And if thats the case : Will a BIOS update change anything ?! Or will a new, special windows driver be neccessary ?
Yes, what you described is what I'm doing. Except the dynamic FSB switching (DFFS in Advanced CPU settings of RM Clock). Thanks for this hint. I'll try it.
So far I can "heat up" (it's not really hot as you all know) the system but when overdoing it with the 2nd monitor etc. still it clocked down half, but not fully as without RM Clock. I'll see if that DFFS will stop that, too.
I'll keep you all posted if I ever receive a BIOS beta for testing. (In the end I hope that DELL will learn something from it. Imagine all the hundreds of hours all of us have invested in troubleshooting and (this is more painful) in convincing DELL that there is a problem and that it is not the heatpipe or whatsoever.
Still I have another question to everybody here : Is someone using Windows XP and having the "temperature problem",too ?
Well, yes, absolutely. That's all I run. Any report you saw from me so far was based strictly on XP. I'm certainly not the only one, of course. I've really not even touched Vista or Win7 yet, but may play with either or both to see what differences, if any, there are in the symptoms. You'll be interested to know that Linux folks also report the problem. In fact, whatever Dell runs their advanced diagnostics on (the pre-boot assessment diagnostics you get by booting from power-down while holding the Fn key or, seperately, through the F12 key at boot-time) also seems to be affected.
Hi tinkerdude. Thanks for your reply. So it is definetely a not operating system related "bug". Let's wait for a BIOS update with higher temperature limits for downclocking / throttling - this should solve the problem. Still I wonder how this construction failure could happen. DELL produces such a variety of Computers and most of them don't clock down at that low temperatures ... Also wonder why it took about half a year to react. It's a pitty. Aside from this error I'm so far a happy DELL owner.
For all of you using Windows XP, RMClock will be a temporary solution, too. It's easy to start inside Windows XP. No additional software or drivers needed. For more iformation see my first posting concernin RMClock above.
Yes, I can override all CPU/FSB throttling with RMClock, but it is an ugly kludge that grossly violates standards. And for less technical folks, it is difficult to understand, difficult to configure correctly and, as you appropriately warn, can even be dangerous. I have to use it to keep my system from becoming useless under load, though, so I do (on occasion, things got so bad it bluescreened).
I'm not sure the "enable Dynamic FSB Frequency Switching (DFFS)" setting needs to be unchecked, though. It won't hurt to do so, but as long as you're using a profile that avoids the lowest-performing P-state, DFFS should never be engaged anyway. I definitely don't need it on my system to avoid all CPU/FSB downclocking.
And on a related note....
So far I can "heat up" (it's not really hot as you all know) the system but when overdoing it with the 2nd monitor etc. still it clocked down half...
JoeB7, when you say "clocked down half", do you mean your CPU clocked down by half? your FSB? How did you measure this? Or are you saying the system *feels* like it's clocked down by half?
Two reasons I'm curious are:
1. RMClock should prevent all CPU/FSB downclocking, if configured appropriately, even without unchecking "enable Dynamic FSB Frequency Switching (DFFS)". It works that way on my system, at least. If it doesn't on your system, there's something about this problem I'm not understanding (and I'd want to investigate that if I could).
2. This morning, I confirmed that yet another type of throttling occurs. If temps rise fast and strong (even though we're still talking 60's and 70's Celsius, which is still well within operating range, of course), then sometimes, at one of the 30-second intervals where CPU throttling occurs, the NVIDIA GPU gets throttled (by more than half - see attached capture from RivaTuner). This was docked, dual monitors. Now, that wouldn't apply to your Intel graphics system, but it highlights the fact that there may be other throttling going on that we haven't yet discovered. I wonder whether your remaining slowdowns might be due to some as yet undiscovered throttling? Looking at the GM45 Express Chipset datasheets, for instance, there are several additional types of throttling that are technically possible, however unlikely their use may seem.
1. RMClock should prevent all CPU/FSB downclocking, if configured appropriately, even without unchecking "enable Dynamic FSB Frequency Switching (DFFS)". It works that way on my system, at least. If it doesn't on your system, there's something about this problem I'm not understanding (and I'd want to investigate that if I could).
I own a Latitude E6400 with a P8400 CPU (2.26 GHz). When configuring RMClock "at maximum" concerning CPU, which in my case means 100% throttle, 8x VID and 1.1250 V CPU Speed remains at about 1.8 GHz. The only way for me to get it to 2.26 GHz is unchecking "enable Dynamic FSB Frequency Switching (DFFS). Honestly I can't give you a theoratical reason for this behaviour. But practical it's the only way for me to reach the 2.26 GHz. I think there is something lowering the FSB Frequency by using DFFS ...
Yes, I can override all CPU/FSB throttling with RMClock, but it is an ugly kludge that grossly violates standards. And for less technical folks, it is difficult to understand, difficult to configure correctly and, as you appropriately warn, can even be dangerous. I have to use it to keep my system from becoming useless under load,
Same in my oppinion. It's an ugly workaround. But at least it offers a working system till the time DELL hopefully brings a real solution.
JoeB7, when you say "clocked down half", do you mean your CPU clocked down by half? your FSB? How did you measure this? Or are you saying the system *feels* like it's clocked down by half?
I saw when I hovered over the RMClock icon in the system tray where the Ghz is shown. I used a 3d application with dual monitor this time. (3d on one monitor - 2D application on the 2nd monitor). So I managed that some sort of downclocking happened even with RMClock. It showed half of the max processor speed after "CPU throttle" in the tool tip (hovering the sys tray icon of RM Clock.)
JoeB7, when you say "clocked down half", do you mean your CPU clocked down by half? your FSB? How did you measure this? Or are you saying the system *feels* like it's clocked down by half?
I saw when I hovered over the RMClock icon in the system tray where the Ghz is shown. I used a 3d application with dual monitor this time. (3d on one monitor - 2D application on the 2nd monitor). So I managed that some sort of downclocking happened even with RMClock. It showed half of the max processor speed after "CPU throttle" in the tool tip (hovering the sys tray icon of RM Clock.)
[/quote]
Mighty curious. I guess it might be explained somehow in the differences between our systems (my E6500/NVIDIA/XP vs your E6400/Intel/Vista), but it seems impractical, unfortunately, to try and troubleshoot the source of the difference. I hope you'll keep us posted if you get wind of any new information on the remaining throttling (or if that curious reading is explained some other way).
I tried just now to make headway once again with Dell Technical Support. After explaining the problem to the Hardware Department, they forwarded me to the Software Department . I explained the problem again to the Software Department, and then they tried to forward me back to the Hardware Department 8-P. I was not too excited about that. I wound up speaking to a “second-level technical support supervisor” who said he would submit the issue to the “Escalation Team”, who would contact me directly to follow up in a day or two. He described the “Escalation Team” as very capable with a wide spectrum of expertise. Time will tell on that. Of course, I never even got the last call I was promised.
I never received a call back. I probably spent a couple hours today talking to 6 different Dell people trying to just get an answer to the question "am I going to get this callback or not?". I got nowhere. It was a Kafkaesque journey taking me from El Salvador to Tampa, then to India, the Phillipines and ending somewhere in Central Florida (the last person I talked to wouldn't reveal the city, but it wasn't Tampa according to him). Total dead end. I was polite, but persistent. It didn't matter. So this is the second time I've been led to believe something's happening, been promised a call back and then was blown off.
I've started one last track. I did reach someone in Tampa who was engaged and conscientious enough to record the details in my no-doubt extensive log for this service request, including details on how to reproduce the problem and I gave him the link to my 59 page report. So they have all they need to follow up on this. I was told that at this coming Monday's regular meeting of the Tampa lab technician(s) and Tampa second-level tech support staff and Dell Engineering (presumably calling in from somewhere else), that this issue would be brought up. But I'm at the end of my rope with Dell Tech Support. If they call back, fine, but I'm not calling them anymore. I've done my part ( 8 separate calls talking to 12 different people). I still encourage others to bang on them politely.
JoeB7, your recent report regarding your contact with Dell was promising, but today I re-read your post from over 3 months ago (April 22) where you said:
Please note, they have replicated the issue in their laboratory and reported it to the DELL Latitude product management. I hope there will be some more info soon.
I'm starting to see a pattern. Lots of promises, but no results. It seems that this could be a deliberate corporate response to try and keep what appears to be a serious, but often invisible design flaw or defect hushed up as much as possible to avoid having to fix it and/or to avoid bad publicity. Or it could just be a stellar case of collective incompetence.
I still hope to continue investigating and publicizing this problem and encourage others to do the same.
JoeB7, your recent report regarding your contact with Dell was promising, but today I re-read your post from over 3 months ago (April 22) where you said:
[quote user="JoeB7"]
Please note, they have replicated the issue in their laboratory and reported it to the DELL Latitude product management. I hope there will be some more info soon.
I'm starting to see a pattern. Lots of promises, but no results. It seems that this could be a deliberate corporate response to try and keep what appears to be a serious, but often invisible design flaw or defect hushed up as much as possible to avoid having to fix it and/or to avoid bad publicity. Or it could just be a stellar case of collective incompetence.
I still hope to continue investigating and publicizing this problem and encourage others to do the same.
[/quote]
I absolutely agree that everyone should do what they can. I tried it via a well recognized magazine. Fantomex had the idea of putting a video on youtube: How to "overheat" / downclock the Dell E6400 within some minutes...
I'm still a bit confident that DELL will solve the issue with an BIOS update, as I receive bi-weekly calls and I'm now connected to a different department. However, the contact there said, they are very happy, that someone from a company with the buying power we have has raised the issue. It sounded that those technicians and support specialists have been aware of the problem already, but internally they didn't get heard. So they really said they are glad that I went through our EMEA sourcing department to our US HQ global sourcing department that made the deal with DELL once. That brought a lot of attention to it. The attention that was necessary that finally the product management started testing. DELL is also a big customer from us and so the voice was heard finally.
It'll be helpful if you guys contact your DELL account manager if you have one. I'm not sure at what company size this is an available option.
Anyhow, the way this issue has been treated so far, the time it took and the many proofs we all delivered, sheds a bad light on DELL. They should better take care of what their customers are reporting. But maybe we are not enough and many E-series users never found out about the issue they have.
Hi. Perhaps I have an idea how to fix "bios" without bios update or even changing bios itself.
As we / you found out, the reason for all problems are the low temperature limits in a bios table. This BIOS Table exists. As far as I found out it's the DSDT. DSDT is part of the ACPI and provides all the values for Temperature, Fan, Speedstep, throttling and so on. When booting any operating system ACPI gives all its tables and values to the operating system. From this point on BIOS has no controll but gave all controll to your oerpating system which works with the BIOS values it received. You can read this tables out. They are a combination of clear text and hexadecimal. There is an deassembler and a compiler. So you can create a dump of acpi, change the values for the temperature limits, compile the new ACPI and give it to the operating system. Then you can tell Windows / Linux not to use the ACPI the BIOS provides at startup but to use the one you created.
So this is not a real BIOS update. But it is nearly the same. You don't change values in BIOS but you change the values inside the operating system which normalwise receives these values from BIOS. From the moment on Windows or Linux are loaded its just the same as if there were new values in the BIOS.
Just created an ACPI dump, will give it a try tomorrow.
Anybody who followed this way before or gained some experience with it ?
ivka57
28 Posts
0
August 4th, 2009 02:00
I think, the real difference came with new heat pipe and new hdd (hitachi instead seagate - hitachi is 4C (celsius) cooler). My "old" system (old MB and HDD) with closed lid and docked was so warm, that you couldn't keep your hand on the keyboard for long. Current system is not so much warmer, than older 610 system under same conditions (dock, closed lid, ext. monitors).
mk
concentriq
1 Message
0
August 4th, 2009 13:00
i wonder if there is any visual clues to this new heatpipe. could you maybe post a good pic of thsi new heatpipe here?
JoeB7
55 Posts
0
August 4th, 2009 14:00
I got no pictures unfortunately, however, there are two heatpipes for the E6400. One for the Intel GPU and one for the Nvidia GPU. The latter one is a bit thicker. When DELL replaced my heatpipe, they have installed the bigger one.
More news about the BIOS thermal table modifications:
Today, I received a phone call and e-mail from DELL Global Commercial Support Services department. The issue has been narrowed down. A new BIOS is currently in test. They expect publishing it till end of this month. I've asked for a beta for testing and hope to receive one. It is also quite likely that it will be for all E-series Latitudes.
ankaro
10 Posts
0
August 5th, 2009 15:00
Hi. First of all, please excuse my rusty english. And let me say "thank you" for this thread. I nearly went crazy with this problem. Thought it was a Windows problem. Good to see I'm not the only one.
I found a solution / workaround for this problem. A special thanks to "tinkerdude" and his great pdf.file showing what happens at what time. It lead me to the fact that the problem has to be the temperature limits in BIOS and / or WinDrivers. They are to narrow. What happens here is all right but should happen at ca. 60/70 °C and not at 45°C. In my oppinion a BIOS and / or driver update should fix this very easy. DELL just have to set higher (much higher?) limits. The CPU used in the Dell Latitudes should survive much higher temp. without any damage. I really wonder who set limits to 45°C.
Today I've spoken with the DELL service. They were friendly and keen to help me. But finally they say the problem is well known, and an update will be released. In near future - perhaps September or later.
For the time till the release of such an update I want to present you my solution / workaround to this problem. I own a Dell Latitude E6400 with a 2.26 GhZ Dual Core and the following worked fine for me :
WARNING ! You should know / understand what you are going to do ! If you don't - better wait for a bios update. You could cause damage to your Latitude !
If you use Windows XP/ VISTA you only need this program : RMClock ( its Freeware : http://cpu.rightmark.org/products/rmclock.shtml ) It need Administrator Privileges.
If you use Windows Vista 64 or Windows 7 - 64 you need two programs : RMClock as above and ATSIV. I found ATSIV here : http://www.computerbase.de/forum/showpost.php?p=3465614&postcount=6 You only need the ATSIV file out of this zip package.
How to use ATSIV and RMClock together : read this Thread : http://forum.rightmark.org/topic.cgi?id=6:600-2 . -> (Rember you have to start the created vbs file with Administrator Privileges.)
Once you started RMClock you can manually change a lot of things.
AGAIN : BE CAREFUL : DO YOU UNDERSTAND WHAT ALL THE OPTIONS AND VALUES MEAN ? If not - better wait for an BIOS update. It's your own risk.
I disabled the Dynamic FSB Frequency Switching, chose a Profile, enabled it and configured the chosen Profile with 100% throttle, 8x VID and 1.1250 V (fits for my 2.26 GHz Dual Core). Applied.
Now RMClock has controll of all the CPU powermanagement including speedstep, throttle etc. It doesn't matter what BIOS or Windows want to do - RMClock decides.
Voila - now I have a 2.26 GHz CPU and not a 700 MHz one all the time. Temperature is fine. (RMClocl shows you the current temperature).
Just remember : From now on Windows has less / no controll of your CPU untill you quit RMClock. There will be no powersaving etc. unless you do it your own - by creating a Powersaving Profile inside of RMClock.
So, I hope I could help some of you.
Neither the less I hope DELL will bring out an update soon... The above is just an unaesthetic workaround with losing comfort and not a final patch. But at least I can use my Latitude now.
ankaro
10 Posts
0
August 6th, 2009 01:00
Hi JB,
now that you told me I found your old posting concerning RMClock. This thread grew big ... hard to read everything :)
In your old post you wrote you lower the core voltage, right ? That's an option, too. Lowered core voltage ist the best way to lower the temperature. But in our case I think it's not neccessary. There is no (damage causing) overheating but wrongly set limits. So what I do is setting everything fix to its standard values. Means, I choose the standard multiplicator, standard voltage and turn throttling and DFFS off. From now on CPU and FSB aren't under- or even overclocked. They just work at their standard settings.
The deactiviation of the DFFS Function should solve your remaining Problem. The RMClock CPUsettings just avoid throttling oder speedstepping the CPU, but the FSB will reduce its speed neithertheless. Once you stopp the slowing of FSB you should have full power in all states. When having tried this setting, can you give me your feedback ? I'm interested if it helped you out.
Still I have another question to everybody here : Is someone using Windows XP and having the "temperature problem",too ? Just used Windows XP for some time, and it seems to me didn't have this problem, at least not that bad... I really ask myself, whether DELL has to update the BIOS or a simple Vista / Windows 7 driver could be the solution. As RMClock shows you, all this can be controlled from inside windows... Or even worse : When it runs fine without any modifications using Windows XP but really doesn't using Vista / 7 is Windows Vista superior to BIOS in this case ??? And if thats the case : Will a BIOS update change anything ?! Or will a new, special windows driver be neccessary ?
Regards
JoeB7
55 Posts
0
August 6th, 2009 01:00
Hey ankaro, welcome to the club. ;-)
Yes, what you described is what I'm doing. Except the dynamic FSB switching (DFFS in Advanced CPU settings of RM Clock). Thanks for this hint. I'll try it.
So far I can "heat up" (it's not really hot as you all know) the system but when overdoing it with the 2nd monitor etc. still it clocked down half, but not fully as without RM Clock. I'll see if that DFFS will stop that, too.
I'll keep you all posted if I ever receive a BIOS beta for testing. (In the end I hope that DELL will learn something from it. Imagine all the hundreds of hours all of us have invested in troubleshooting and (this is more painful) in convincing DELL that there is a problem and that it is not the heatpipe or whatsoever.
Regards, JB
tinkerdude
22 Posts
0
August 6th, 2009 04:00
Well, yes, absolutely. That's all I run. Any report you saw from me so far was based strictly on XP. I'm certainly not the only one, of course. I've really not even touched Vista or Win7 yet, but may play with either or both to see what differences, if any, there are in the symptoms. You'll be interested to know that Linux folks also report the problem. In fact, whatever Dell runs their advanced diagnostics on (the pre-boot assessment diagnostics you get by booting from power-down while holding the Fn key or, seperately, through the F12 key at boot-time) also seems to be affected.
ankaro
10 Posts
0
August 6th, 2009 05:00
Hi tinkerdude. Thanks for your reply. So it is definetely a not operating system related "bug". Let's wait for a BIOS update with higher temperature limits for downclocking / throttling - this should solve the problem. Still I wonder how this construction failure could happen. DELL produces such a variety of Computers and most of them don't clock down at that low temperatures ... Also wonder why it took about half a year to react. It's a pitty. Aside from this error I'm so far a happy DELL owner.
For all of you using Windows XP, RMClock will be a temporary solution, too. It's easy to start inside Windows XP. No additional software or drivers needed. For more iformation see my first posting concernin RMClock above.
Did you give RMclock a try tinkerdude ?
Regards
tinkerdude
22 Posts
0
August 6th, 2009 10:00
Yes, I can override all CPU/FSB throttling with RMClock, but it is an ugly kludge that grossly violates standards. And for less technical folks, it is difficult to understand, difficult to configure correctly and, as you appropriately warn, can even be dangerous. I have to use it to keep my system from becoming useless under load, though, so I do (on occasion, things got so bad it bluescreened).
I'm not sure the "enable Dynamic FSB Frequency Switching (DFFS)" setting needs to be unchecked, though. It won't hurt to do so, but as long as you're using a profile that avoids the lowest-performing P-state, DFFS should never be engaged anyway. I definitely don't need it on my system to avoid all CPU/FSB downclocking.
And on a related note....
JoeB7, when you say "clocked down half", do you mean your CPU clocked down by half? your FSB? How did you measure this? Or are you saying the system *feels* like it's clocked down by half?
Two reasons I'm curious are:
1. RMClock should prevent all CPU/FSB downclocking, if configured appropriately, even without unchecking "enable Dynamic FSB Frequency Switching (DFFS)". It works that way on my system, at least. If it doesn't on your system, there's something about this problem I'm not understanding (and I'd want to investigate that if I could).
2. This morning, I confirmed that yet another type of throttling occurs. If temps rise fast and strong (even though we're still talking 60's and 70's Celsius, which is still well within operating range, of course), then sometimes, at one of the 30-second intervals where CPU throttling occurs, the NVIDIA GPU gets throttled (by more than half - see attached capture from RivaTuner). This was docked, dual monitors. Now, that wouldn't apply to your Intel graphics system, but it highlights the fact that there may be other throttling going on that we haven't yet discovered. I wonder whether your remaining slowdowns might be due to some as yet undiscovered throttling? Looking at the GM45 Express Chipset datasheets, for instance, there are several additional types of throttling that are technically possible, however unlikely their use may seem.
ankaro
10 Posts
0
August 6th, 2009 11:00
I own a Latitude E6400 with a P8400 CPU (2.26 GHz). When configuring RMClock "at maximum" concerning CPU, which in my case means 100% throttle, 8x VID and 1.1250 V CPU Speed remains at about 1.8 GHz. The only way for me to get it to 2.26 GHz is unchecking "enable Dynamic FSB Frequency Switching (DFFS). Honestly I can't give you a theoratical reason for this behaviour. But practical it's the only way for me to reach the 2.26 GHz. I think there is something lowering the FSB Frequency by using DFFS ...
Same in my oppinion. It's an ugly workaround. But at least it offers a working system till the time DELL hopefully brings a real solution.
JoeB7
55 Posts
0
August 6th, 2009 13:00
I saw when I hovered over the RMClock icon in the system tray where the Ghz is shown. I used a 3d application with dual monitor this time. (3d on one monitor - 2D application on the 2nd monitor). So I managed that some sort of downclocking happened even with RMClock. It showed half of the max processor speed after "CPU throttle" in the tool tip (hovering the sys tray icon of RM Clock.)
tinkerdude
22 Posts
0
August 6th, 2009 17:00
I saw when I hovered over the RMClock icon in the system tray where the Ghz is shown. I used a 3d application with dual monitor this time. (3d on one monitor - 2D application on the 2nd monitor). So I managed that some sort of downclocking happened even with RMClock. It showed half of the max processor speed after "CPU throttle" in the tool tip (hovering the sys tray icon of RM Clock.)
[/quote]
Mighty curious. I guess it might be explained somehow in the differences between our systems (my E6500/NVIDIA/XP vs your E6400/Intel/Vista), but it seems impractical, unfortunately, to try and troubleshoot the source of the difference. I hope you'll keep us posted if you get wind of any new information on the remaining throttling (or if that curious reading is explained some other way).
tinkerdude
22 Posts
0
August 6th, 2009 17:00
Last Friday, July 31st, I wrote:
I never received a call back. I probably spent a couple hours today talking to 6 different Dell people trying to just get an answer to the question "am I going to get this callback or not?". I got nowhere. It was a Kafkaesque journey taking me from El Salvador to Tampa, then to India, the Phillipines and ending somewhere in Central Florida (the last person I talked to wouldn't reveal the city, but it wasn't Tampa according to him). Total dead end. I was polite, but persistent. It didn't matter. So this is the second time I've been led to believe something's happening, been promised a call back and then was blown off.
I've started one last track. I did reach someone in Tampa who was engaged and conscientious enough to record the details in my no-doubt extensive log for this service request, including details on how to reproduce the problem and I gave him the link to my 59 page report. So they have all they need to follow up on this. I was told that at this coming Monday's regular meeting of the Tampa lab technician(s) and Tampa second-level tech support staff and Dell Engineering (presumably calling in from somewhere else), that this issue would be brought up. But I'm at the end of my rope with Dell Tech Support. If they call back, fine, but I'm not calling them anymore. I've done my part ( 8 separate calls talking to 12 different people). I still encourage others to bang on them politely.
JoeB7, your recent report regarding your contact with Dell was promising, but today I re-read your post from over 3 months ago (April 22) where you said:
I'm starting to see a pattern. Lots of promises, but no results. It seems that this could be a deliberate corporate response to try and keep what appears to be a serious, but often invisible design flaw or defect hushed up as much as possible to avoid having to fix it and/or to avoid bad publicity. Or it could just be a stellar case of collective incompetence.
I still hope to continue investigating and publicizing this problem and encourage others to do the same.
JoeB7
55 Posts
0
August 7th, 2009 02:00
I'm starting to see a pattern. Lots of promises, but no results. It seems that this could be a deliberate corporate response to try and keep what appears to be a serious, but often invisible design flaw or defect hushed up as much as possible to avoid having to fix it and/or to avoid bad publicity. Or it could just be a stellar case of collective incompetence.
I still hope to continue investigating and publicizing this problem and encourage others to do the same.
[/quote]
I absolutely agree that everyone should do what they can. I tried it via a well recognized magazine. Fantomex had the idea of putting a video on youtube: How to "overheat" / downclock the Dell E6400 within some minutes...
I'm still a bit confident that DELL will solve the issue with an BIOS update, as I receive bi-weekly calls and I'm now connected to a different department. However, the contact there said, they are very happy, that someone from a company with the buying power we have has raised the issue. It sounded that those technicians and support specialists have been aware of the problem already, but internally they didn't get heard. So they really said they are glad that I went through our EMEA sourcing department to our US HQ global sourcing department that made the deal with DELL once. That brought a lot of attention to it. The attention that was necessary that finally the product management started testing. DELL is also a big customer from us and so the voice was heard finally.
It'll be helpful if you guys contact your DELL account manager if you have one. I'm not sure at what company size this is an available option.
Anyhow, the way this issue has been treated so far, the time it took and the many proofs we all delivered, sheds a bad light on DELL. They should better take care of what their customers are reporting. But maybe we are not enough and many E-series users never found out about the issue they have.
ankaro
10 Posts
0
August 7th, 2009 09:00
Hi. Perhaps I have an idea how to fix "bios" without bios update or even changing bios itself.
As we / you found out, the reason for all problems are the low temperature limits in a bios table. This BIOS Table exists. As far as I found out it's the DSDT. DSDT is part of the ACPI and provides all the values for Temperature, Fan, Speedstep, throttling and so on. When booting any operating system ACPI gives all its tables and values to the operating system. From this point on BIOS has no controll but gave all controll to your oerpating system which works with the BIOS values it received. You can read this tables out. They are a combination of clear text and hexadecimal. There is an deassembler and a compiler. So you can create a dump of acpi, change the values for the temperature limits, compile the new ACPI and give it to the operating system. Then you can tell Windows / Linux not to use the ACPI the BIOS provides at startup but to use the one you created.
So this is not a real BIOS update. But it is nearly the same. You don't change values in BIOS but you change the values inside the operating system which normalwise receives these values from BIOS. From the moment on Windows or Linux are loaded its just the same as if there were new values in the BIOS.
Just created an ACPI dump, will give it a try tomorrow.
Anybody who followed this way before or gained some experience with it ?