How to Overclock the Alienware 18 and Haswell CPU (or actually have it run full stock Turbo Speed)

So I've been trying to OC my 4810mq I did all the settings the above poster did but for some reason I can't change the Turbo Boost Power Max setting, anyone know why?

 

Looks like it runs quite a lot cooler than your Clevo from what I can tell... certainly more than 3.9GHz. DumbDumb's new Alienware 18 with 4940MX can run 4.3GHz on factory thermal paste without any problems. I also just posted something here that may help answer your question... http://forum.notebookreview.com/ali...a-any-tips-before-i-start-45.html#post9723407

 

How irritating that the 18 can run high CPU but has the PSU limitation which screws the GPUs... I was thinking about buying the 1800 dollar 18 off Amazon and swapping the parts from my Clevo into the Alienware but there's no point if I can't overvolt the GPUs later on.

Seems like you can't win.

 

Depends on how far you want to overvolt the GPU. Have a look at what I posted here: http://forum.notebookreview.com/ali...enchmark-thread-part-3-a-139.html#post9723671

 

Yeah my Clevo doesn't overload the A/C adapter until 1.1v at which point 3dmark 11 will overload with the turbo turned off on the CPU. 1.085 or whatever it is right before that will overload with turbo on.

I'm just extremely annoyed with the stupid design Clevo went with for the slave GPU that causes it to be 10C higher than the master under load which causes it to throttle quickly with the stock vbios.

We will see what happens when I get the Liquid Ultra. I'm going to try the GC Extreme on it first though, almost out of IC Diamond. I'm a bit concerned there may not be sufficient contact to even use CLU on the slave GPU...

Sent from my HTC One_M8 using Tapatalk

 

Actually, poor contact is where CLU can potentially do better as long as it bridges the gap fully. It transfers heat more effectively due to it being metal, whereas non-metallic pastes perform very poorly without good contact.

Look at the poor contact on my primary GPU heat sink on the 18. CLU dropped the GPU Core Max about 12°C on the Alienware 18 (about 19°C lower Core Max on the 4930MX @ 4.3GHz).

Excellent contact with 4930MX die and heat sink...	Less effective contact with 780M die and heat sink...

 

Well maybe I can turn this around then and make my Clevo great. That would make me ecstatic!

Sent from my HTC One_M8 using Tapatalk

 

I bet you can, Brother Ethrem. It has taken a lot of time, tweaking, trial and error, and fussing to get this Alienware 18 where it is today. It performs a LOT better than when it was first released. Of course, A08 and A09 are to thank for some of that. The earlier system BIOS were lousy in more ways than one. If it could take advantage of the dual AC adapter like the M18xR2 I think it is finally to the point where it would spank that old beast, but since it can't it never will unless we can fix that with some soldering. I need to reach out to svl7 about that before long and see if we can come up with something that will work to remove the power limit.

 

I think the Clevo has a lot of potential. It kind of seems like Clevo machines are capable of much more but out of the box they're middle of the pack.

I really think that I need to get in contact with Xotic about getting a spare bottom cover so I can do some mods to get the airflow up. And then I get to learn how to dremel haha. I swear by the time I'm done, I'm going to be very familiar with tools I never touched before. Never been one for hard mods but this machine just begs for it.

Sent from my HTC One_M8 using Tapatalk

 

Hello guy's. I'm new to this forum and yesterday my 4810qm got replaced by a 4930mx

My cinebench score jumped from 650 to 835 at 4.3ghz. Also I used Mr.Fox guide to get 4,3 ghz stable which worked fantastic.

But I have a little problem. I can't set flex ratio to 32 in xtu. Every time after a reboot xtu says that it couldn't set it. In The Bios it's set to disable... Can anyone help me?

At 4,3ghz my cpu reaches 93 degrees during cinebench. Is this ok so far for Liquid metal?

 

Hello all,

I have gotten my pc to run at 4.5GHZ, and it is pretty stable no BSOD, while gaming or surfing web or benching. Of course now that I am at 4.5GHZ, want to go to 4.7 or even 4.9. However, when I try for 4.7 I get an immediate freeze followed by BOSD. I have had the following settings

-Core current limit 142A,
-Processor current limit 300A,
-Turbo boost power time window to 3145728
-Core voltage 1.320v

These are the biggest changes I made, but I still got the BSOD. When I imputed these numbers XTU did not request a reboot just click on apply, so I thought I was good to good-no dice. I'm not trying to fry my MB, but should I up the numbers? If so, how high would be safe. Turbo boost power time window is maxed, so are Turbo boost long/short power max. If, I get this worked out how high can I overclock to safely 4.7ghz, 4.9, 5.0???

Thanks for any input. Oh, just to be safe I turned down Core limit to 130A, and Processor Limit to 185A.

 

Hi guys, when running the benchmarking tool built into the XTU, my processor speed downclocks from 4.2GHz to 3.6 GHz. Otherwise during idle and playing CPU intensive games (CSGO), it remains at 4.2 GHz from what I can tell. Any idea why this occurs? I understand this tool stresses the CPU in different ways than other programs, but my 3920xm always maintained level speeds during this test.

Mr. Fox-- I am using your XTU settings from the first post in this thread, except at 42 multiplier instead of 43.

In XTU, my processor frequency shows up as 0 GHz even though it is sensed just fine in ThrottleStop and OpenHardwareMonitor. I also get CPU Throttling ranging up to 50% during the XTU benchmark, but 0% otherwise. Overall I get around 900marks @ 4.2ghz even though the frequency varies between 3.6 and 4.2 during the test. I also keep EIST checked in ThrottleStop since it needs to be enabled in XTU (even though I've seen you post before that EIST is unchecked in your ThrottleStop screenshots.)



Program versions used:
XTU 4.4.04
ThrottleStop 6
Alienware 18 BIOS A09

 

Can you post a screen shot of your XTU settings?

 

Here it is:
https://www.dropbox.com/s/pnnlgm2ji16kr1t/Screenshot 2014-08-02 14.45.54.png

 

Try the settings shown on this screen and see if they work better. You can set the multis at 42x4 if you want to, but set everything else the same.

 

Thanks for the settings. I gave them a whirl, but I still get downclocking during the XTU benchmark test. I took a screenshot mid-test: https://www.dropbox.com/s/5gu4bw2lis75u5t/Screenshot 2014-08-02 15.30.51.png

 

Finally broke through the 1000 mark in XTU. It was weird, I turned down a few things, turned up others and BOOM:

Here's a screen shot:







Still testing to see how high and stable I can go.

 

Fluctuating clock speeds in XTU benchmark is normal. Try some other benchmark instead. wPrime 32M, ThrottleStop benchmark, 3DMark11 or 3DMark Vantage and see how it goes.

Excellent job!

 

On a slightly related note, my laptop has been freezing up after I come back after being AFK. I have all the power options for turning off display, hard disk, etc disabled, but I have noticed that whenever I come back after being gone for a while (>30min), my laptop freezes up and I have to do a hard reset.

My feeling is it's probably a RAM or GFX card issue. My RAM is not OC'ed. I am using the unlocked vBios from TI for these 780m SLI (I run with SLI disabled though), in addition to the XTU settings you shared earlier. This was happening even before I used the new XTU settings though, so I doubt it's that. Any ideas?

 

Potentially using the wrong CPU power settings is causing the lock-up. Try removing the CPU overclock (set BIOS defaults) and see if the freezing stops. If it does, then you'll know for sure that you need to use different settings.

 

I'll try that. I was looking around in my Event Viewer and noticed these "Warning" messages: "The speed of processor X in group 0 is being limited by system firmware. The processor has been in this reduced performance state for 71 seconds since the last report." for all of my processors. Is this normal?

FYI: I ran the memory diagnostic tool at boot and it found no errors.

I am also using a very outdated driver set for the 780m (v314.22 with a modded INF file) since no other driver set allows my BenQ monitor (XL2420z) to run at all resolutions and refresh rates-- should run at 144Hz on 1920x1080 and any other res. I posted about this issue elsewhere, but here are screenshots of the issue:

• 
  With v337.88, 327.33 (and almost all other drivers I have tried): https://www.dropbox.com/s/ovzeqieaj3tfqha/Screenshot 2014-08-02 04.36.19.png

• 
  With 314.22 (the only driver that works with my 780m setup): https://www.dropbox.com/s/orqu5vw40o1ycxf/Screenshot 2014-08-02 04.47.50.png

 

You must be running the "Balanced" or "Power saver" profile instead of "High performance" power profile. That's what most commonly generates that error. Unless you have a special reason to (need to run on battery) always use "High performance" and that error will most likely go away.

If it does not, or you are already using the "High performance" power profile then you may be experiencing core parking. See bolded text and links below...

 

It's in high performance and I also unparked my CPUs when I first got the laptop (re-checked just now and they are all still unparked.)

Some people say having EIST enabled could cause this, but there's no way around that if I want to use XTU to overclock right?

 

If your CPU is running hot it can cause this error message if thermal throttling is occurring. What kind of load temps are you seeing? Here again, if the CPU power settings are not correct it can generate this message the same as thermal throttling can.

Having EIST or c-states enabled can also generate this message, and so can having virtualization enabled in the BIOS or running on battery power, but this message is not necessarily an indication that your system is malfunctioning. It's merely logging that the CPU is not running at maximum clock speeds. If the system is functioning correctly there is probably no reason to worry about the message in Event Viewer.

 

I finally repasted my Alienware 18 with liquid ultra and I'm trying to figure out how to overclock my cpu. The problem that I'm having is that sometimes when I play a game or watch a video I get a blue screen and a whea_uncorrectable_error. What settings do I need to change to be able to get a stable 4.3Ghz overclock? Thank you in advance for any help.

 

No offense, but what you need is some reading glasses.

http://forum.notebookreview.com/ali...actually-have-run-full-stock-turbo-speed.html

 

The error you got is when there is not enough voltage. At least that is my experience, and that's what you glean from reading through the link post by J.Dre. I suggest looking through this thread:

http://forum.notebookreview.com/ali...eeded-oc-cpu-bios-oc-isn-t-enough-4940mx.html


Its only 3 pages long, and it talks about voltage settings for stable systems at 4.3 and above. The core voltage being discussed is the core voltage setting in XTU

 

I actually copied these settings from one of Mr. Fox's posts on another thread. Would you mind sharing your xtu settings for 4.3Ghz?

 

I also started with Mr. Fox's XTU 4.3 settings, but they did not work for me. It becomes really clear quickly (BSOD) that one setting does not work for all. Through trial and error these settings work for me:






At this point the only thing I mess with is the core voltage. I make it higher when I want to increase my multipliers. I'm using the 4940mx CPU.

 

Thanks, I will give these settings a try when I get home from work.

 

LOL... it's even a sticky thread. I merged it.

Scanner is right... trial and error is definitely required to determine what degree of overclocking weirdness or incompetence exists with each Haswell CPU. It's probably the most inconsistent product Intel has ever released.

 

Today I opened xtu and loaded the same settings I posted previously. For some reason they are now working and I haven't had a reboot or blue screen. Is that normal or should I still try to tweak my settings?

 

First try to set the Core Voltage mode to Adaptive then decrease by - 25 mV and see if it helps

if that doesn't help, then set the Core Voltage mode back to Static and lower the core voltage mode a bit until you find a sweet spot

 

Thanks, I will try using the settings you suggested.

 

So, attempt number 2 went alittle better after reading through this whole thread again. While I have built several rigs, I am new to OC'ing so please bear with me.
It seems that under idle I sit around 4.1 or so on but seem to be one core, under load using 4 cores I drop as low as 3.1 but tuning utility shows no thermal throttling. I am missing something, but not a clue as to what.
so the rig is Alienware 18, I7 4940mx, 32G ram, dual Nvidia 880's, windows 7 ult. am using intel extreme tuning utility.



so, things are great, start playing BF4 and about 5 min. in and the system shuts down. Got a message about watchdog maybe being the reasion on reboot. left the tuning utility off, and no problems with game, though the system got hot. at Idle my gpu temps are 52c and 56c, looked right after exiting game and was at high 80's. currently have SLI off. Is this normal? Havent screwed with the gpu's at all. Thanks in advance for any advice.

 

Following many private messages regarding overclocking a 4900MQ i have put up the following post in the hope that this will help those of you that are not familiar with XTU particularly new members. Seems like the 4900MQ is becoming very popular these days. Please note these settings may or may not work with your particular CPU as Mr Fox has stated many times the variance with Haswell CPU's is massive,i have made this template as an average overclock and is not the exact settings i use for my CPU (mine is above average and only needs Vcore 1.15 for 4.2Ghz) yours may require more or if your lucky maybe less Vcore than this template so rough guide only but the basic settings are correct.....happy overclocking guys/girls:thumbsup:

 

Increase turbo boost power max and short time power max to around 200Watts....that should help,the reason it shuts down is probably excessive heat or the system is just not stable also your Vcore seems a little high for 4.3Ghz drop your multipliers to 42,42,42,42.....

 

I am running XTU version 5.1.0.23 for some reasion the turbo boost power max and short time power max wont allow me to alter those setting. Turbo boost short power max is maxed at 85w and will not allow me to adjust any higher, the short time power max at 71w max wont allow me to set any higher as well. I am assuming that by Vcore you mean core voltage? I know I must be over looking something or just not understanding being the noob that I am. THanks for any and all adive. I think the heat issue may have largely been due to running SLI and when running that the system gets how!

 

The version of XTU you are using is for Haswell-E and may not allow you to change the settings i mentioned try version 4.2.0.8 :thumbsup: Vcore is core voltage....

 

Given the direction Intel is taking it might be wise to buy the best non-MX CPU from this point onwards, unless Intel really makes a good comeback with Skylake. Otherwise these 1k Xtreme CPUs are a complete insult and waste of money.

Btw any reason you're not pushing the chip harder on 1 and 2 active cores? Not that 200 MHz would make a big difference, just curious. I'm still awaiting the day AMD releases a 10GHz chip so I can finally play Flight Sim X with all settings maxed out LOL. (FSX was coded in an era where only single core CPUs existed, so it's all about single core speed)

P.S. Having recently watched many, many episodes of Border Security Australia, I feel the itch to visit Oz again LOL (did a whirlwind tour of the east coast with my parents way back in 1999 when I was still a kid)

 

You can try penultimate version.Intel® Extreme Tuning Utility version 4.4.0.4

 

Disable the "Package Turbo Power Lock" and reboot. That will unlock your TDP sliders. Your CPU will never perform well with that stupid thing enabled. I don't understand why it even exists. It never should have been enabled in the BIOS by Alienware. After you reboot it will be gone.

 

Alienware 17 does not have this (Package Turbo Power Lock) adjustment in xtu. '-)

 

The Alienware 18 doesn't have it once you get rid of it.

But, if you flash a new BIOS you have to disable it again.

 

Why is it different adjustment in the xtu on Alienware 17 versus 18? Is Dell playing with the firmware ? :laugh:

 

With my cpu doing 44,43,42,42 requires a fair bit more Vcore 1.225 which given the extra heat just isn't worth it. I didn't realize that XTU 4.4 is the best version will have to check it out when I get home..what's so special about it? In Australia we all have kangaroos in our back yards at least that's what most Americans think lol they are actually really large mice...:hi2:

 

Ah Ha! Thank you Mr. Fox...


Ok. so next couple of questions. Again forgive me for being such a noob. I went in to this thinking that this setting would be for all four cores, yet it only sets one. So, I see when benching that when all four cores are running it will overall drop to as low as 3.1. Is this normal, or is there something else that I am missing? I am looking at the repasting of CPU and both GPUs. Im not sure about liqued ultra, but looking at ICD. Before I take a dive with this was wondering if any of you folks have used or have experiance with something along the lines of Artic Silver 5? Also would past or pads be better for GPUs based on the experance you folks have had. Dear god why do I feel like I just opened a can of worms.lol. I have read through many topics on here but havent seen what I am looking for yet, so thought I would ask. Thanks again folks.

 

You can pretty much forget about overclocking your Haswell with AS5... it is not a good enough paste to manage it. IC Diamond will do a little better than AS5, but it will still run way too hot. Haswell is INSANELY HOT. Liquid Ultra is the only viable solution if you plan to run 4.3GHz or higher. Otherwise, forget about overclocking your 18. Manually set all 4 cores to the same ratio in the BIOS before setting it with XTU. Single-core overclocking is totally worthless.

Have a look at the thread for the quote below. There are some posts prior to my quote relating to AS5.

 

Thank you Mr. Fox.
So in short to stay warm in the winter months over clock and sit behind the laptop. lol.
Ok, so to insure that I have this correct. Clocking all 4 cores for 4.3 is pretty much pointless unless I want to set up a perment cooling station as well as repaste. That said, if I repaste and clock at 4.0 to 4.1 maybe 4.3 if my chip can handle it, permanently for everyday and gaming shouldnt have that big of an issue? I have seen both alot for and aganist liquid ultra, mostly being having to repaste about every three months (or six months) or so if I remember correctly. Have you been running the liquid ultra for awhile and if so how much of a pain is it? I dont mind the tear down and putting back together, I did that the day it arrived just to peak under the hood. lol. It's not something that I want to do every other week though. Thanks again.

 

Liquid Ultra should last more than 6 months. I have been using it about that long now. I took things apart to experiment at month 5 and at that point it showed no evidence of graded performance. It's easy to remove, easy to apply, extremely effective and very durable. Perhaps those against it did not apply it correctly. The kind of overclocking I do, IC Diamond is the next best thing and it is usually good for about 3 months, but it cannot come close to cooling my Haswell or Ivy Bridge XM processors nearly as well as Liquid Ultra. I cannot say enough good about it and at this point I plan on using it exclusively on all systems that I overclock the CPU... it's that good. I have a number of posts in this thread and you will find the evidence is well documented.

http://forum.notebookreview.com/ali...ra-any-tips-before-i-start-5.html#post9520969

I still use and recommend IC Diamond for average systems because it is effective and durable. I have had no problems with it either. It does a fine job on GPUs as well.