If I were working tomorrow I would definitely be sleeping. I'd have to wake up in 4 hours. My local hardware store should have a power meter.
-
-
cookinwitdiesel Retired Bencher
I meant I maxed out the ac adapter's rating
Will check the chat log on my desktop, is where I was chatting with Don in July when I got into all this benchmarking rubbish with my whitebook
And my lowly desktop only pushed up to an even kilowattlol
-
Mandrake i have noticeds the cpu slowdown it seem like from time to time my systeme just work slowmotion and it did not happened on bios B14 i think it's something with .29 and the 4850
cause i'm not runing oc and i can feel it everyday when i get some heavy task on the machine it start going slowmotion until it completely freeze but when i'm playing some music it just play it but slowed it's not stutering but just slowing down -
cookinwitdiesel Retired Bencher
Try this, run the quad with only 2 cores being used, you can set windows to only see/use 2 cores. Maybe it is something to do with the way the driver is being used with a quad vs a dual for the 4850s that is slowing the system down (trying to think like John lol)
-
mandrake...seems you have two things you can try right now
switch the bios
pull one vga card -
cookinwitdiesel Retired Bencher
180 seems to stick in the memory, can't believe it was that big of a delta between that and the M17x though!
-
-
nah..not just a bios option...you can do it per application as well, simply by using cpu affinity.
scook.....you know me...always trying to think outside the box...lol -
-
cookinwitdiesel Retired Bencher
no, it is part of windows startup
go to msconfig, boot, advanced options, check the box for processors, select 2
-
cookinwitdiesel Retired Bencher
You can set the OS to only use x cores regardless of individual programs John, so it will show up as a dual core
-
make sure you pick the right two cores!
on the d901c...the fastest two cores we're 0/1 -
B14 bios is set to 1.163v just like the M17x but then you have only the FSB to crank up the speed and no multi to mess with...
-
there should be 0 reason for the .29 to have any issues with the 4850 unless they are pulling more than the hardware will allow as ive stated many times. even though i pulled the voltage limits out of the bios remember the mobo has its own limits in the psu section as i said before its semi weak imo from what i remember off hand actually it was only a mere 3 phase design and not a great one at that. so with the overclocking it may just be saturating the psu section of he board. the b14 bios will limit you to so much voltage BEFORE the mobo hardware steps in. this will not happen as ive said with .29
edit if you get a good psu on this like a real ac-dc unit and it still happens imo its hardware releated not the bios or software. these are some of the reasons way back when that ocz skipped them in their testing and also why flex kept holding off on them i remember even telling don this when i had them in hand a long while before aw ever did. i had hoped this was all resolved by now though
have you tried undervolting everything? as much as possible? also remove 1 ram stick run with one hdd, no usb devices etc try to get as much load off and see if that helps
edit just looked back at my pics ITS ONLY A 2 PHASE MOBO LOL.. even the cheapest asrock desktop g31 mobo's are at least a 3 phase design -
also i have no multi to mess with in the first place XD
i switched to 2 core and put the voltage to 1.1625 and rebooted the thing feel like it's on steoridswell see if i feel the slow down if i don't i'll switch back to 4 cores
zfactor i have not even tryed to OC yet with your bios cause has of now i'm only searching if i can tweak it to run stable and even then i feel the same things that mandrake told while they where not there on b14 -
mandrake said he had no issues with the 3870's at all so imo its the extra current draw of the 4850's just putting it over the top
-
and how come it happens to me even if i had 4850 and a q9000 on b14 and it did not happened and it happen exactly the same now that i'm on .29 it's the part i don't get ....
-
?? not sure i get what you mean? you mean it happens to you on b14 and him on ,29 or you on both or???
right now you say you are only using 2 cores and have the 4850 at 1.1625 correct and you dont have the issue?
there is nothing bios wise that would limit the 4850 in anyway shape or form thats what im saying -
cookinwitdiesel Retired Bencher
so in the case of the 4850s, a beefier power supply should go a long way in helping the system perform to its true abilities? Since they are overdrawing the stock as adapter?
-
No i think what Z is trying to say is THE SYSTEM is having issues doing it and not much can be done for it
-
it happen only on .29
only using 2 core at 1.1625 atm
and i had 4850 all along stock
for now i don't see the issue
my hardware was the same all along since i don't have any other cpu or gpu to swap
Sry for my english it's not my native language -
so its possible b14 is LIMITING the system to keep voltage in check is the only thing i can think of that could be different but that will suck for overclocking
and yes it imo is more hardware related -
in the laptop world..it's over looked.. with a beefier power supply...you might be able to do more with less voltage as well..(speculation of course). could be the oppisite.
gangsterone swore by his. its a real power supply with your laptop tip on it. and regulated for your amprage, but puts out more watts..(more speculations) and it's of course...custom built...let me see if i can find the place again. i would have one if i didn't sell my laptop. because it can pretty much be used with any laptop...as long as you ahve the right tipped wire on the end. -
wow yeah i went back through pics and remember how crappy this mobo really is (no offense to anyone who has them heck id love one again). doesnt even use a separate phase for the ram simply a fet wired into the 2 phase psu section from the main input etc.. really crappy as most laptop boards are but even clevo has a MUCH beefier power section on the 901. i just think you guys are at the limits of this board im sure john knows what i mean especially when it comes to desktop boards
-
it's a little wider than the normal power supply. -
lol wow what a wimp... you should see mine i run on these....ill take a pic
-
it don't surprise me most lappy are crappy
but i'm wondering how come all the laptop i get have a bad crappy power section ....
Edit back on 4 cores for now i'm trouble free i'll see if it does it again -
are you trouble free with the gpu's at the lower voltage though? or back to default
-
cookinwitdiesel Retired Bencher
With power supplies, (and you should know this John!
) the voltages are always the same, more watts just means higher current on the already established voltage rails. I have already explained how higher current correlates to a higher performance at a given voltage. Higher voltage specifically to the CPU is different from more watts coming off the power supply. Although in the end, more electrons either in the form of current or potential, will help performance along
If the actual hardware is what is limiting the system here (crappy voltage regulation on the motherboard), then I can happily say thank god I have MXM 3.0b now with its higher wattage standards
How many phases is the voltage regulation to the cpu on the M17? The M17x has 3
edit: just saw your post z, is what I thought. can someone put up a good picture of the motherboard with the cpu cooler removed please? -
I have no problems with slow downs with my M17 on Windows 7 with the B10 bios and dual 4850's.
Though i don't use a quad.
I did have stutter with 3870's and vista...not sure if it can be OS related lol.
-
ill up a better pic when i get my camera back from the office -
cookinwitdiesel Retired Bencher
-
cookinwitdiesel Retired Bencher
It is like we are trying to super turbo charge a non-reinforced engine -
the specs for your laptop are a guideline and not a max..once you get past this...you will go further Danial-son...lol -
cookinwitdiesel Retired Bencher
But the higher standard will mean that beefier power circuitry is present
Which I can then torture to the point of my motherboard failing....wait, did that already! haha (and the new one works with my battery!!!)
All this is making think I will most likely just buy the 4870 upgrade (if that is possible) instead of getting an M17 all over again. Although they are undeniably fun notebooks!
When we were doing the motherboard swap I just had a warm happy feeling seeing that trio of voltage regulators lined up by the CPU socket, they even get cooled! (unlike the M17) -
and where are the new benchmarks at????????? -
lol this is one of my el cheapo ones i use for laptop testing i have over a dozen of these in my shop and one that cost 5-6 times what these do this is what im talking about when i say im using a power supply for testing lol this one is a bit dusty from being in storage i just pulled it out for some stuff at home im going to work on:
Attached Files:
-
-
i have one of those...but not with all the fancy stuff on the front.
-
yeah but most people here dont. is yours the same model? minus the gauges. this was one of aries best units..i also have a few of these at my shop among others:
http://cgi.ebay.com/ws/eBayISAPI.dl...d=280414615663&ff4=263602_263622#ht_958wt_986
http://www.testequipmentconnection.com/index.php?main_page=product_info&products_id=2600 -
nah...i was using mine to power my car stereo system during extreme testing. never used it for what your doing. you just be giving me more and more ideas over the years...lol
-
with the variable voltage dial you can tune it in to what the system needs use a good meter to be sure i use mostly fluke ones or my scope at the shop. and simply wire up the end that plugs into the system to one of the outputs depending on which amperage you need and you have one massive psu
-
so...what kind of power are these laptops drawing??
-
so 2 hours later i played some game seem the new voltage did the trick for me but the issue apeared only yesterday so might just come back in 24 hours i'll keep you updated
i'm back at quad with 1.1625v and for now everything is fine -
nah, it's not like yours. it a beefed up replacement power inverter. one you throw in your laptop bag and go...lol
it replaces your stock power inverter -
cookinwitdiesel Retired Bencher
So that is just a high end DC power supply, we use those all the time in the engineering labs at my school haha
Are you using that to power the entire notebook, or hooking up to individual devices?
I replaced the board, the tech got a laugh out of it all. He got there and I was all ready to go. He said with the old alienware units this was usually how it would go lol. Started trying some runs again, but the stutter was so depressing for me I called it a day. This needs to be reconciled NOW -
Going back to the .28 bios didn't help. Removing unneeded hardware didn't help. Disabling hardware that is non-removable doesn't help.
-
as i said imo this is a mobo power issue again this board has a VERY weak power section
-
Mandrake how frequent do you get the stutter? Is the problem only occur with 4850 and Quad?
-
TurbodTalon Notebook Virtuoso
Are the people who got the M17 with 4850s right before they discontinued it experiencing these problems?
-
hello z ..... good job you doing by the way ( like trying to fix it ) as you are the only person in the world doing this at the mo i think.
so i want a qx9300......
doeas this laptop (m17) not have enough power available to it... or can it be never fixed.
how many amps would you say the laptop pulls?
to you 2 up their the stuttering has been a problem from the begining quadcore with 3870s ect
M17 OC Benchmarking - The Next Generation!
Discussion in 'Alienware Area-51/Aurora and Legacy Systems' started by Mandrake, Dec 11, 2009.