*** Official Clevo P570WM | P570WM3 / Sager NP9570 Owners Lounge ***

+1000 on this.. I always do fresh install, clonning causes certain irritating issues...Clean install will prevent quite a bit of hassle even though it might take time and effort..

 

I will have to do that, if I fail to initialise the RAID or the initialised RAID is instable.

 

You may want to format a little 100mb partition on the raid set to try and keep it aligned if you are doing it manually.

 

I always get a chuckle out of "Better than 99%..."

 

Very nice

 

i thought u sold ur 570 meaker lol

 

In the process of selling to a user who will give it a good home.

 

Have you decided what will be replacing it yet?

 

I already did and showed it off at a gaming convention. The question is now what I do after that

 

how much are you selling ur 570wm? meaker bro?

 

I have a buyer and would not be able to discuss prices on here anyway

A damn good price though.

 

that skylake E seems so far away..

 

in Q1 2016 Broadwell E will come. In Q4 2016 or Q1 2017 Skylake E will come

 

Now did you manage that score with those clocks? O_O. Last I remember, Mr. Fox got a lot lower score with better speeds (except CPU).

 

Remember my vBIOS keeps the Turbo intact, so the Mhz shown in 3DMark won't show the actual speed of the run (THX@3dMark for not boosting the GPU long enough to get accurate readings)

 

http://www.kitguru.net/laptops/mobi...4gb-3ghz-ddr4-so-dimm-memory-kit-for-laptops/

all we need now is skylake-E quad channel memory in this updated p770wm or something along that line. quad channel memory at 3000mhz per dim.. 64gb total holy smoke.

 

I was under the assumption that Mr. Fox was using your vBIOS, but if he was using svl7's then that makes more sense; I'd expect Kpaxx had a higher boost. But still, what a huge difference too.

 

I think he was using his own mod at that time...

 

I know!
I was talking about @Mr. Fox old benches...

 

I will never have scores like those Brothers @Kpaxx and @johnksss do until I can afford a 4960X. I've got an excellent 4930K, but it cannot reach those 4960X overclock levels. That extra 200MHz makes a big difference on benchmarks where CPU clock speeds matter. I haven't been benching for a long time because I needed a new CPU heat sink. I have one now and it's much better, so I may find some time to get back to benching again. I also did not have my portable AC cooling all set up for benching for several months. (I moved and it took a while to get things in order.) I can use Liquid Ultra now and my CPU temps are much improved because the heat sink actually makes good contact with the IHS. The old heat sink barely touched it in one spot and the temps were not very good. I had to use a huge amount of IC Diamond to up the air space, LOL.

I am primarily using @Prema's vBIOS now. It is an excellent vBIOS. I bounce back and forth between it and another one, depending on which one returns the highest benchmark scores in a given benchmark. It can vary a little by benchmark which one produces the highest number. I really like the boost working with Prema's mod, but also like having the Power Target (TDP) slider because that shows me data on % of TDP in my OSD. That data does not show without the TDP slider unlocked for some reason.

 

Was that the original heatsink you had received with the laptop? I don't remember seeing noticeable differences in the cores on my Xeon but then again it wasn't an X or K processor so it always stayed cool.

Did Eurocom provide a replacement?

 

Yes it was the original. Temps were OK running stock clocks. At 4.3GHz and higher it was hitting thermal junction and shutting down except with AC cooling and paste jobs were not enough to fix it. I posted photos earlier in the thread here showing my testing with pressure test film. There was essentially no contact to speak of. Testing with the pressure test film still shows the contact pressure is still very weak, but at least the coverage is much more even and temps are much better now... to the tune of about 10 degrees lower at 4.3GHz and no thermal shutdowns.

I paid for the replacement part. I communicated with them and all I got was suggestions on things to try. That lead nowhere so I just asked for a new heat sink since it was clear they did not understand the situation. And, there was no way in hell I was going to send them my machine for warranty service. They gave me a good deal on the heat sink though. Since I am no longer an Alienware guy, my days of buying extended warranties are over. I don't send my machines to other people for repair.

 

Yes, but his GPU scores were higher even though his core clock showed up as being lower (is what I meant). I know his CPU was 100MHz faster than yours at least.

Does the Prema vBIOS not have the TDP slider unlocked in nVidia Inspector etc? I thought it did. Or am I misunderstanding what you mean?

Have you tried lapping the heatsinks at all by chance? It may prove worthwhile to improve contact. Even your GPU heatsinks may benefit. When I eventually swap out my CPU/Slave GPU heatsinks, I'm going to lap them before I even consider putting them on; I want to make sure I get the most cooling out of it XD. In fact I think I'm going to practice on my old GPU heatsink and see if I get it done properly so I don't mess up with the new heatsinks.

I know that feeling, and it's not the first time I've seen something like this happen. I'm glad you got a new heatsink easily enough, though I feel bad that you did have to buy it.

 

TDP report is disabled as the values do not represent any actual TDP.
Part of it being the highly raised TDP base that is braking the algorithm (done in the vBIOS Mods that have this feature enabled).
e.g. when GPU-Z displays that the card is using 100%, it is in reality already running on something like 165% (that being without the TDP slider even touched).
Thus giving the user a wrong impression...

 

@D2 Ultima - no, lapping would not be useful with the old heat sink because it is apparently deformed or assembled from the factor incorrectly. It wasn't making contact effectively enough to consider lapping, in addition to low pressure. Even with the new heat sink, contact pressure is low, but it has even contact (versus almost no contact) across the IHS. Lapping may help the new one a little bit, and I did polish it up with emery cloth. I also bought some large copper shims that will cover most of the IHS that I plan to try on the Panther to see if that will prove thermals by increasing contact pressure, but I haven't gotten around to testing that yet.

Another issue is the IHS should not be polished and if the IHS is not perfectly flat there is no way of correcting that. (Polishing the IHS would also remove the serial and ID numbers, which would be bad for anyone concerned about warranty... which I am not.) Ideally, I would de-lid the CPU, but everything I have read indicates doing that will kill an Ivy-E CPU because the IHS is soldered to the die on Ivy-E. I haven't seen an example of a successful de-lid on an Ivy-E CPU. Maybe @johnksss can chime in on that since he is more in tune with desktop CPUs than I am.

Low contact pressure seems to be a common problem on laptop heat sinks. This is one area where Alienware really shines. On all of my Alienwares the contact pressure is very high and turns the pressure sensitive film bright red. The contact pressure is also very low on the GPU heat sinks on my Panther, but the GPUs are still running cool enough with Liquid Ultra thermal paste applied.

 

So basically... don't worry about it? XD. Silly maxwell cards >_<

 

The only way to de-lid an IVY-E is to de-solder the lid using heat and that is very, VERY risky.

 

Any who have tried this with success?

 

Yes it has been done, you can look up desolder IHS to see how it has been done in the past.

 

Probably not worth the effort IMO... Anyways has anyone seen the 6700k being delided?

 

So, what the point of doing so?
Tin is a very good conductor.

 

But still a thermal interface and not perfect, better to have as few a layers as possible.

Is the difference big? No, but for extreme clockers after every Mhz then they may want to go to the effort.

 

Don't worth the effort.
Too risky.
For my opinion.

 

That is correct for the vast majority of users yes.

 

Doesn't really work as well as people seem to think it does.

 

Not without extra work no.

 

Anything above 0 degree water on a delidded Ivy cpu is pointless.
Delidding any Sandy bridge cpu is pointless at any stage.

 

IVY-E I agree, it is more for the sub zero crowd which is why I said the vast majority would find it pointless. Any of the paste chips it is a different matter.

 

Sorry. My team puts this to the test all the time with each chip that comes out. This is usually how the rest of the internet finds out what a chip is made of...

Edit: With the exception of mobile, which is already delidded.

 

aren't all HEDT CPU soldered IHS? only starting from ivy and above mainstream desktop uses crap paste.

johnkjsssssssssssssssssss r u getting skylake soon? dat asus gx700 water cooled yeah?

 

I was only replying to cpu delidding or whether it's really worth doing.

Nope, just going to sit back and wait to see if they come with the 990M for us. Although the screen is nice and water too, but if it's only one gpu & a quad, so it still may not be the fastest laptop, but it would sure have all the nice upgrades though.

 

This is because you would need to alter some more code to get it to read below 100% (And this is with using stock clocks and boost working with both sliders unlocked.) while still keeping the slider working. If you make changes as is, it will not work correctly unfortunately.And the more you get away from the 100 percent the worse things get as you go higher....observation only.

 

with pascal it'll be a sweet upgrade but yeah quad with a junky cpu high voltage low clock is a shame. it would have 64gb ram though at 3000mhz, and hopefully some pcie raid 0 via chipset.

 

It's only a numbers game for me at the end of the day. If it can't beat what I already have then it wouldn't be to beneficial for me as a whole unfortunately. Already did the super high speed drives so that would need to be faster than 2300MB's

 

You are right...in my testing back then (with adjusted base line) it jumped from 100% to only 120%-ish with sporadic fluctuations all over the place while the card was pushed (1450Mhz @ 1.2v). (Looks like the underlying base algorithm becomes useless once the upper values can be flexible.)
So unless it can give the user an accurate reading all the way it won't be as useful as we hoped it to be.

Will do my best, hardware wise (as long as they also publish the MXM B version) I don't see a problem for the P570WM.
They in the end also didn't go with the "GTX990M" branding for the "E-GXX" after all:

http://forum.techinferno.com/clevo/10770-mobile-geforce-gtx980-incoming-3.html#post144538

 

Wait, so instead of GTX 990M, they name it GTX980 (no space in between)...? Wouldn't that cause too much confusion? WTF Nvidia...

edit: GTX 985M, GTX 990M or even GTX 999M would've been way better than this...

 

Correct, except there is a space (I just never use one): NVIDIA GeForce GTX 980

The whole marketing will probably be targeted towards not being a mobile but real desktop chip, even though it's mobile (MXM-based) card with a lot lower vRAM speed than its desktop counterpart.

 

So we have notebook GTX 980M, notebook GTX 980 and desktop GTX 980...

Great job Nvidia, that won't make confusion, not even the slightest bit...

 

Meaning.... Your baseline needs to start pretty low. Like down around 50% and then go up from there. So when you hit 1500 mhz@ 1.2V it still reads below or around 100%. This is feasible because then it gives you all the useful sensor data back without compromise. And the system no longers hinders the GPU.(Speculation - As long as all other settings line up)
My clocks are 1348/1463/1407
http://www.3dmark.com/compare/sd/3361184/sd/3371052

I only mention this because people like myself like to run osd software which looks nice in video and picture illustrations. And to gage just what ones system is doing in real time.

 

In my eyes the "ideal" Mod should show the actual TDP in use while using the stock vB TDP as its baseline.
That way people can see how far they are already hovering above the designed limits even under stock clocks and voltage.
The user should not think that his engine is running with 3000rpm while it's already at 4500rpm.
But once adjusted that way it somehow broke the algorithm and almost didn't react at all anymore...

Sweet benches there guys, what do you think causes the differences between them?
(Do you know if you are also both using same RAM at same speed and timings?)