I'm not sure if this is happening to anyone else, but I've recently been experiencing some serious issues with my Dell Latitude E6400 (with latest BIOS, and all other relevant updates) while surfing, playing flash videos full screen, etc. After extended periods, Windows Vista slows to a halt to the point I can barely minimize a window without severely delayed lag. I've also noticed that the typical games I play have also had its frame rates severely slowed down as well.
I have been trying out Mozilla Firefox 3.1 Beta 2 and also the latest ESET Smart Security 4 Beta, but I've doubt these are the culprit. Just for good measure, I've uninstalled these but the issue still seems to persist.
As I use this mainly for graphic design (using latest Adobe CS4), it's quite critical to be able to handle most of these tasks. For that matter, as a $2,000+ laptop, I'd at least assume to be able to surf and play flash videos without issues. I've been told my Dell tech support to run Diagnostics, but everything seems to complete successfully. They haven't gotten back to me again yet - but I've purchased a 3 year Complete Cover warranty - what am I to do?
I'm a Canadian consumer and have heard nothing but the best from them - but as I am currently lodged in Hong Kong, I purchased this through Dell Hong Kong and am worried that my warranty (even though purchased), will not be rewarded as well as the Canadian counterpart without a severe delay - which means my work will be compromised.
Any ideas would be great!
Thanks for all the great info folks. I have an E-6400 and have noticed the same issue over the past few months. I wanted to add a little to the conversation in hopes that it might spur a brilliant idea or perhpas help Dell with a fix.
I installed a fan control utility (http://www.diefer.de/i8kfan/) and told it to monitor the temps of the CPU, GPU, memory, system board, HDD as well as the CPU utilization. The results were very interesting and showed that if you reproduce the symptom by placing the CPU under about a 50% load for an extended period of time (hours) I can watch the temps of all the monitored components stay stable in the (60s or so with the fan forced to high) with the exception of the system board. The system board goes from 60 degrees up to 96 or 97 degrees and then falls over 10 (+ or -) minutes with workaround 1-3 implameted. The interesting thing is that as it peaks and then falls, the CPU utilization stays fairly constent util the system board temp gets near its valley and then the CPU maxes out for a few minutes. After the CPU returns to normal, the whole cycle begins again. I am happy to share log files if anyone is interested.
I have worked around this by doing 4 things (in the order I implemented them):
I have not tried all the permutations to see which steps are absolutely necessary but I do know that if you only do items 1-3, it is not enough to eliminate the problem completely. With all four work arounds implamented I do not get the system board temp roller coaster or CPU utilization spikes I was. I don't have the laptop we paid for but it is tolerable for the moment.
I had the same thing happen to me. I found a solution. add a laptop lap fan. " http://www.bitsinconline.com/blog/?p=121 " BITS Orlando Computer Repair Blog had solution. Sounds simple but it cooled down my GPU enough to keep my laptop with e-dock and dual monitors running perfect. My GPU stays at aroung 70 to 72 degrees.
Jamal Hartage
Hi all,
I'm a new user of the E6400, and of course i fell into this issue. Here is a short account of what happened and how i solved the issue, it might help some of you:
- i use the PC for mixed work/home usage, and got the 800Mhz symptom primarly when using 3D games at home. (i got the Intel GMA)
- very predictable behavior:
1) launch the heavy app, it will be running fine for 10/20 minutes, CPU temp stable around 45-50, frequency =2800 (T9600), fan at full speed.
2) Then it goes down the drain to 800 Mhz in 4 steps of 30 seconds: 2800 => 2100 (76%), after 30 seconds => 1500 (58%), after 30 seconds => 800 (26%), after 30 seconds. It stays there until i leave the app. The temperature does not look like rising when this triggers. And of course in the meantime the game completely frozen to 2fps instead of 40fps, making it unusable.
- i carefully read this thread and started by upgrading the BIOS from A17 to A19
1) launch the heavy app, it will be running fine for 10/20 minutes, CPU temp stable around 45-50, frequency =2800 (T9600)
2) Then it goes in the same cycle: (2800 => 2100 => 1500 => 800). This happens even as the temperature is going down, which i agree is a complete shame to be put on DELL . After 30 seconds, the same process happens again in reverse, coming back to 2800 Mhz in 4 steps of 30 seconds. Overall, the cycle is very precisely 4 minutes.
So this shows the new BIOS was indeed modified in the thermal tables so that it recovers much more quickly from the underclocking, but the initial downclocking still happens as before.
From The Intel website Core 2 technical docs, you can see what happens to us is ultimately linked to the CPU thermal protection, but in our case it is triggered by the BIOS too early. This feature is part of SpeedStep, and i had read in another thread that disabling speedstep would put the CPU at the lowest setting (800Mhz). Still i decided to give it a try and went into the BIOS, disabled Speedstep and rebooted .....
The CPU was now running at 2800 !!!, so it looks like disabling speedstep actually puts the CPU at its normal speed. I went thru my test plan several times and my system is now always at 2800, no more clockdowns. i can at least use my heavy apps. The temp stays reasonable at 50-55, fan at full speed.
So for me the solution was easy : turn off SpeedStep in the BIOS
Now turning off SpeedStep has a few bad side effects, but nothing critical:
- loose EIST, that is the cpu cannot downclock to 800 Mhz when there is no activity, this will definitelly hurt autonomy when running on battery.
- loose thermal protection, but there is another one in the CPU if you reach absolute critical Tj temperature of 105. Then the CPU would shutdown, you would get a BSOD in windows. This has a very low probability of happening in our case, as we are having the symptoms at a much lower temperature anyway.
If one day Dell comes up with a final fix in the BIOS, i will re-activate SpeedStep ...
I got a motherboard and heatsink replacement to fix the dock/undock problem and my problems related to this thread seem to have vanished (without the use of any software).
Please let me know, since I have just baught E6400 (did not get it yet) and I run into this thread:
Did the new BIOS update A20 for E6400 solved this problem or not ? (on only partially)
?
Thank you
An update from me.
I've had 3 new heat sinks / motherboards put in, and I think I finally have the problem resolved (at least for the time being).
The reason for 3 motherboards was:
1. First one sent to replace the issue caused a separate graphics issue.
2. Second one sent to replace the one above was completely dead and the system would not boot
3. Third one finally seemed to correct the issue.
Interestingly the third one they send me (which is not displaying this issue) has BIOS A12 on it. I am not sure the BIOS version is the full story though.
no luck,
updated to A20, disabled SpeedStep in BIOS, none of those worked for me.
to test it, i disabled rmclock (put it in current profile - no managment) and after a few minutes watching flash, it starts breezeng. while breezing, the only way to get to normal funtioning is to enable rmclock again.
Jorge Riva Melendez
Been more than 6 months, I'm guessing DELL chose to drop this product issue and wont bother anymore. Not even a single word of update since few months ago.
I've been using so many DELLs products since 2004, and loved it, but this stupid and silly unresolved problem PLUS ignorance brings total disappointment to me.
I won't recommend this product nor DELL itself to anybody anymore.
yes, and thats why im going to write a review of this product on the main page, with a link to this thread. fell free to do the same.
i usually dont write reviews, unless im very happy / upset with the product, wich in this case is the second one. i also bought many DELL laptops before, and even recommend this brand as the best.
i think the had too much time to fix this, and its the moment to stop buyers to keep thinking in this machine as an alternative, like haimke that just bought this laptop without knowing this very annoying issue.
First of all: thanks to all the people who spent the time to document the problem and suggest solutions. It is not a good situation, but it's good to know that I am not the only one and that I am not crazy -- which i started to suspect after 3 months of dealing with this issue :)
Quick recap of my situation: E6400, 2.8 CPU, 4GB RAM, 256GB Patriot SSD. For about 8 mo ran Win XP with the dock with 2 monitors and no issues. In October 09 decided to upgrade to Win 7 and started experiencing issues while in the dock. No issues when undocked. Because it started happening after I installed Win 7 I started blaming either OS or drivers.Contacted Dell, had them replace the dock but that didn't help.
Ran w/o dual-monitors for 4 months until 2 weeks ago I started experiencing the same problems outside of the dock. Still being convinced that the issue is Win 7 - opened a ticket with Microsoft but that didn't help (they had me uninstall Nvidia drivers b/c memory dump showed that it was taking up most CPU). Not very relevant in the light of "downclocking". Over the last weekend i completely reinstalled Win7 with all the latest drivers and A20 BIOS. Was looking forward to getting into the office and finally using my two 22" monitors.... but 10 minutes into being docked, the same issue happened.
I asked my IT guy look into it by contacting dell -- he did one better and found this forum (http://www.sevenforums.com/general-discussion/47799-dell-latitude-e6400-windows-7-docking-problems.html#post557499) which led me here.
To summarize: latest A20 BIOS doesn't seem to address the issue...
I agree - A20 does not fix this issue. I don't even have a dock and still experience significant, system-crippling throttling. Windows 7 takes ages to boot, predominantly because my 2.5GHz processor has been throttled to 700Mhz. without any indication/warning/ability to override.. it's just not acceptable! I think there's more to this that shonky programming by Dell though - thinking back, I may have only experienced this issue since installing Windows 7 (I was an early adopter as we got it pre-release at work).
What I find most disgusting about this issue is a total lack of communication from Dell. No wonder they continue to lose market share. I think I'm going to upgrade to a typewriter.
Though this isn't the "real" solution, here is a solution that should solve the problem to some extent. Someone found a blog posting on this, and shared it with me.
You can see the real fan speeds and even change when it will go into kick-down mode. Take a look!
to enable this mode (from in Windows or anyplace)
hold down (fn+shift) then type (while still holding down fn+shift) 1,5,3,2,4 - This will enable the special mode.
After that fn+R will bring you to a text based screen (again, even in Windows) where you can get to all the temps, fan speeds, etc, and even set the throttling percentages.
You can then use fn+R anytime to get into this page until reboot, at which time you need to turn it on again.
FYI, once you manually mess with them, the fans will not change speed until you reboot.
Hope this helps!
Kudos to the one that shared the info about the "secret menu" "thermal control panel". I've heard about that earlier and wasn't sure about if it's true.
That's helpful. Any description about the temp sensors available? Well, I try to figure it out.
Strange is, that I never faced the throttling since the T-something (temporary) BIOS and A17/19 BIOS. I can run several 3D and CPU consuming tasks (not to name it games) without downclocking anymore.(Intel GMA graphics)
I agree to the DELL guy. That's a nice bad cooling design with the current E Latitudes...
Hi, i compared the thermal tables from A19 to A20. There were no changes at all. Same values as to the debug mode table.
--johannes