***Gigabyte P35X owner's lounge***

Also what is a good basic memory overclock? Like 300?

I am currently using MSI afterburner for both.

 

Fan profiles must be different on the p35w v3 because my fan doesnt get very loud on auto-high, only time it gets really loud is on max fan.

 

I am actually using Nvidia Inspector.

+135 GPU
+300 Memory

Nice and smooth

Just wondering about overvoltage GPU. How much you think with XTU?

 

I wouldn't OC on this machine.. Power Supply is insufficient I believe.

 

I do want to try to over volt my GPU. Since less power is going to wards CPU.

What do you think would be good for:

Processor Graphics Voltage Offset

+25 or +70

 

Guys btw I am experiencing the ridiculous Dragon Age Inquisition "cannot play fullscreen" bug, anybody tried it on our machines?

https://forums.geforce.com/default/...agon-age-inquisition-can-t-play-fullscreen/2/

Cannot run in full screen mode! - Page 3 - Answer HQ

 

Ok here is last update for a little have to get back to normal life.

I am using Nvidia Inspector:

+135 Clock
+300 Memory

XTU

Processor Graphics Voltage Offset = +70

Noticed +25 to +35 FPS increase on Ultra maxed out. Plus the XTU settings I had before.

CPU throttling now again so I might back it down to +50 or +25. Still have to mess with it for a while on GPU.

But I'll let you know.

Update:

Wait maybe that was a fluke.

No thermal throttling at +70

Will test for a while later and let you know.

Update:

Came back just a little bit.

Going to try +25 and probably leave it at that.

...

I am just trying to find the perfect balance with no throttling of any kind and good temps.

I think I found it



Update:

OK I think I have come to my conclusion:

If you overclock the GPU and the memory

It throttles cpu

If you overclock the GPU and over vault GPU

It throttles cpu

I will have to admit for some reason it is really smooth game play though and really high FPS. I think smoother than just running everything at default and letting cpu throttle like normal.

...

In any case I will have to be honest and say a part of me wants to return it because in a sense this is not as future proof as I thought.

I mean you are getting exactly what you paid for at this point in time.

That cpu you is not going to ever perform better for this gpu no matter what you do.

So the million dollar question is ...

Do you just be happy with a 980M that will play all the games at this level or do you go back on the hunt.

Personally I have no idea what I would look for because everything "is what it is"

The only true thing you can always get is a desktop. That will always give you most bang for buck and allow upgrades.

There in my book not really any other options except Aorus X7 Pro. But that CPU Throttles at 70 from what I hear. And the battery life is horrible.

I love the GS60 and think that is perfect because you get great performance and you will never think about maxing everything out. You are just like this is a great overall laptop. 8th just the horrid battery life as well.

...

If i had to be honest I would say that if the GS60 had the same battery life as the P35x I would probably buy that instead.

BUT

It does not. With undervolting I will probably get 7 hours battery life with word and 0 brightness with this laptop and no bluetooth or wifi in class.

It is way series than I thought it was Going to be and the keyboard is perfectly fine.

980m smokes BF4 on Ultra maxed out with everything and if you use nvidia inspector and overclock and undervolt cpu You will get 95-105 FPS with minor cpu throttling and 65-75 with zero cpu or power throttling.

And that the 970M will never be able to do unless extremely stressed.

So all in all I think I will probably keep this laptop for the very reason I found it in the first place and that is because it's great at everything and not just one thing or two.

Great looks
Great keyboard
Great temps
Great power
Great battery
Great form factor

So this should be a keeper unless something crazy happens.

One thing to keep in mind I am much more knowledgeable about temps and throttling than before for some reason, but rewind a few years when I never really thought about it. This laptop would do anything just extremely fine.

Oh and zero driver or BSOD.

Let the games begin.

 

Why do you want to overvolt your 980M GPU if you've already tested and run stable at the +135Mhz on the cores - that's the maximum that you can overclock with a stock vBIOS, so you don't need any more voltage. Increasing the voltage only increases the voltage of the core too, not the VRAM. Or are you talking about overvolting the Intel GPU? Why would you want to overvolt the Intel GPU? (To me it sounds like you're confused about how all this stuff fits together.) (Intel GPU is not really used for gaming, just used & active for desktop stuff, then when the game starts up it switches over to the NVidia 980M GPU - a system called Optimus, so not sure what you want to gain by overvolting the Intel GPU).

 

That clears things up.

I just used inspector at the end and do:

+135
+300

Get very minor thermal throttling, probably less than normal defaults but it is there.

However there is a combination to where you get 0 thermal and power throttling. And runs everything cool.

Will net you 65+ FPS on BF4 Ultra with everything maxed.

...

Now i just have to decide what do I want ...

 

2 hours of maxed Crysis 3 (including 4xMSAA, rest maxed): 83C on GPU, 90C on CPU. I think it's not bad, but it might need a repasting. BTW on the -70mV undervolt, I got a computer shutdown. It might not be stable.

 

Did the CPU throttling affect the in-game performance?

It's been seen that cooling pads (even mediocre ones) have a sizable effect on CPU temps and performance - and that liquid ultra repasting is pretty effective. I belive combining the two, along with an undervolt, will make things near perfect

 

Try undervolting the cache as well -70 I have no shut downs with both.

Something I might try for fun. Leave all defaults like out of box and use nvidia inspector and go +135 and +300 and just see how performance is. It will throttle but I bet you. Get 100 FPS easy out of it.

 

I've been watching this thread for a month now, and I decided that the good outweighed the bad (thanks to all the users who described their experiences), and so I ordered the CF4 from XoticPC.

If anyone out there is looking to buy the p35X, do NOT buy it from XoticPC. I placed my order on December 2. Eight days later, they finishing processing my payment. Nine days have passed since then, and my laptop is still in Phase 1 (Payment Processed & Inventory Gathering), and they have not responded to any of my emails asking how much longer it will remain in Phase 1.

Now, I know a lot of people have had fine experiences with XoticPC, so maybe it's just getting parts and building the p35x specifically that's a problem, or maybe they just hate me. At any rate, I'll be lucky if I receive the laptop before the month is out.

 

Just talk with either Josh or Hutsady, they are excellent customer reps that really pay attention to your problems and answer all your questions. Probably you are writing to your sales rep who is in contact with a lot of orders and busy dealing with them. Indeed Xotic had their batch delayed for whatever reasons. But get in contact with either Josh or Hutsady.

 

Full Tilt - All In

All stock except for Inspector:

+135
+300




30 Min of BF4 Ultra Maxed Out.

Seems to me full option and overclock ed is really no different. Even though the graph goes all over the place on throttling the average is only like .07

Get like 100 FPS this way.

Might just leave it.


I guess you have two options.

Undervolt and have have consistent 65-75 FPS.

Or this option and have 90 - 135 FPS. With throttling.

I really do not notice a difference between the two. When it's throttling during gameplay I can't see a difference. The only time I even know is when I flip back to xtu and it tells me or if I have FPS on and see it drop from like 100 to 65/75 then right back to 100.

I might just leave it undervolted to keep it a little cooler and have longer battery life. Because really I can't tell the difference.

 

I dont think undervolting will give you longer battery life simply because running ETU uses more battery than you would save with the undervolt. It definitely would if you could do it without having a program running in the background. Its still a good option to lower your temps tho.

I actually tested my battery life with a 70 mv undervolt and got an average of 4 minutes per 1% battery, without ETU running it got a minimum of 4:30 min per 1 % battery life. So with undervolt I was getting around 1.5 hours less battery life.

 

You dont need to leave XTU running for the undervolt settings to stay. I'm sure that drop is battery is because of XTU using extra cpu time to monitor the cpus

 

I think you mean "underclock" not "undervolt", undervolting doesn't reduce performance unless you lower the clocks, lowering the clocks is underclocking (the opposite of overclocking!).

 

how do you keep the undervolt without having ETU running? I though when you closed the program everything went back to original settings

 

Something screwy with your results there, having a program running in the background isn't going to reduce battery life by 1.5 hours unless it's using a significant amount of CPU power, I don't see how XTU would use a constant 'large' portion of CPU power - there was probably something else that influenced your results at that time. You could always use Resource Monitor to find out what the Average CPU consumption is of the XTU process.

 

Anyone planning on doing a liquid ultra repaste? I'd really like to see that. I've seen some amazing results. Mr fox for example went from 98C max to ~82max on his alienware with overclocked intel extreme cpu.

 

All i can do is post my results, you can take them however you please. It appears tho that they dont really matter anyway if you can maintain the undervolt without haveing etu running. And I think jiexi is right about that, i check my voltage on hw monitor and it remains at the undervolt voltage even when i close etu. Thx for bringing that to my attention ive never undervolted before, and always though you had to have the program running for it to take effect.

As for the accuracy of my results, I opened etu, monitored how long it took for 1% point of battery to decrease, repeated this 3 times and got an average of 4 minutes. Then the only thing i did was close out of it and repeated the same thing and got an average of 4.5 minutes. Cant say why it made a difference or if it was just coincidence but it did drain the battery faster with etu on.

 

98 degC to 82 degC! That was probably comparing stock factory paste installed at the factory to Liquid Ultra applied by Mr Fox? Liquid Ultra alone doesn't make 16 degC difference when compared against other good thermal pastes if both are applied properly. More like 4 degC difference as seen here:
Results: CPU Air Cooler; Low Mounting Pressure - Thermal Paste Comparison, Part Two: 39 Products Get Tested

 

Yea the extra battery drain was definitly from etu running. Thx for bringing it to my attention that you dont need to keep program running that makes undervolting even more rewarding.

70 mV seems to be the max for my computer too, I crashed at 80mV while running 3d mark. Been gaming on 70mV for a few hours and no problems. Still a small amount of throttle when fans set on auto-high but definitely lower than it was.

 

Definitely debateable like I described - no need to spread potentially false info! ;-)

 

I was wrong. It's actually 98 to 80 . Check the last two pictures in Mr Fox's album: http://m.imgur.com/a/JUQvK

No no definitely not stock pastejob he's comparing it to. If you'd know him, you'd know he's one of the most hardcore overclockers here on notebookreview. He compares it to his IC Diamond temps. After al the repasting he's done I'm pretty sure he knows how to do it, so the IC Diamond wasn't a bad pastejob.
Edit: just read your sig and massive rep, you must know him. .

Here's the thread/post on it. Mr Fox's impressions are on post 47: http://forum.notebookreview.com/ali...y-liquid-ultra-any-tips-before-i-start-5.html

If this works as good for the p35x, think of the possibilities! There will be thermal headroom to overclock the GPU while the CPU is at full turbo. (Assuming we don't run into power problems).

I should note that the liquid ultra will erode aluminium, so it can only be used on copper parts of the heatsink.

 

Thanks for those links. He definitely seems to have tested it thoroughly from me having a quick skim of it now, it was hard for me to see the detail of his results though because the pics were extremely small text & couldn't read them easily, but I'll take your word for it. I do have some questions though, is the 98 to 80 temperature difference based on a like for like situation, because I know I can see he did measurements after degredation as well as before degredation. Was the 98 perhaps the degraded Noctua and the 80 the freshly applied Liquid Ultra? (I'm only asking because it's hard to digest his posts - can't see the pics clearly.)

 

You should be able to zoom in on the imgur album to be able to read the text. Are you on tapatalk or on a browser?

The IC diamond is 10 days old on the imgur album, while the Liquid Ultra is just new. Somewhere else (I haven't had time to read his post very thoroughly) mr Fox says the temps have stayed roughly the same after 5 months of heavy overclocking. Edit: 5 months with liquid ultra that is

 

Ah, managed to squint and see his data now. Effectively the temperatures of freshly applied Noctua and Liquid Ultra are the same - about 77-80degC. The Noctua degraded after 4 days and delivered the 98 degC you quoted. Now, for some people pastes degrade really quickly & for others it does not degrade as quickly - often the degredation happens if it pumps out the thermal paste, but that's normally indicative of some kind of misalignment of the heatsink (some kind of manufacturing issue) I think. I'm on Arctic Silver 5 and it doesn't degrade at all hardly. If you don't get much degredation of performance then it looks like there's not much difference between Liquid Ultra & other good pastes, which is backed up by the Toms Hardware review I linked, as well as Mr Fox's results of fresh Noctua vs Liquid Ultra.

 

Its not false info, its the results that showed up from the process i described, Its obvious that running a program would use more battery than not running a program lol

 

You're right, I'm sorry I missed that. However the 98C is I quoted is the one of 10 days old IC Diamond, Mr Fox's favourite thermal paste (until he tried liquid ultra), which I presume not te be worn out, I should ask him. It seems LU is better at keeping the amazing first day temps.

Also, in the tom's hardwarereview, the temps are in the thirties. Do you know whether a 3.6C delta at 30C will be a 3.6C Delta at 90C or will it scale linearly, to something like a 10.8C delta? Even ~4C could help out a lot.

 

A simple program that is doing nothing more than monitoring is not likely gonna reduce battery life by 1.5 hours. That's why I said you could verify if Intel XTU uses a significant amount of CPU time by using Resource Monitor to see Average CPU usage of the XTU process - if CPU usage was high for that process then I'd agree with you, but until that point I think there was another factor that skewed your results. That's why I'm challenging it.

 

Like you I thought the same when I saw the Toms Hardware results, but then I looked at the other tests in that review and the temperature delta between Liquid Ultra and the other pastes remained at that small about 4 degC. (Tests like GPU temps which were in the 60's). EDIT: although on looking closer, the GPU tests were with a Liquid Metal Pad, not Liquid Ultra; so possible that it might not scale linearly with higher temperatures, but Fox's results show comparable temperatures between Noctua and Liquid Ultra on Day 1 anyway, so I think main benefit of Ultra is gonna be if you happen to suffer from degredation (not everyone does).

Like you say, it does look like Liquid Ultra is very resistant to degredation over time, so no doubt it's good to use from that point of view, and 4 degC reduction is still worthwhile I think (results from Toms Hardware article); however, it is a 'dangerous' paste to use if you lose track of the little mercury like balls that could short contacts on your motherboard or GPU for instance.

 

All this talk of repasting has me thinking about doing my p35w. If anyone is planning on doing theirs and could do some simple guidance on how to disassemble and then repaste that would be very useful, maybe start a new thread for maintenance?

I'll prob go safe on the paste and use mx4 though

 

Gelid GC Extreme got the best results out of all the 'safe' pastes, (in the Toms Hardware article I linked a few posts back).

 

Just done mine, not that difficult to be honest.

Take off the base panel, disconnect the bass speaker and the battery indicator button, disconnect the battery if you want to be more safe, unscrew the radiator (in any order), note that the fans have warranty stickers on, so better not remove fans unless it's seriously clogged by dust and whatnot.

Remove the old paste and clean the contact surfaces with the pasting kit, use a can of compressed air if got one, put on sufficient amount of paste on the centre of the dies, then carefully position the radiator, screwing the screws IN THE ORDER noted on the radiator (not all the way down at once, do 2 or 3 turns at a time until all of them are secured).

Reconnect everything you've disconnected and put the base panel back on, then depends on the type of paste used, enjoy the drop in temps after burn-in period.

Regards

 

It's a shame they didn't test IC diamond. I'd really like to know if it's as good as so many people say it is.

 

Will you post results?

 

Yep, but if you think about it, it's not gonna be better than Liquid Ultra, and that was only 4 degC better than the regular good pastes, so it's either gonna be the same as the regulars or maybe 1 or 2 degC better at most. Plus, I've read info about it causing scratches on CPU dies & heatsinks, which will degrade thermal transfer after multiple successive re-applications of IC Diamond - I wouldn't use it for those reasons.

 

I agree the difference will not be more than a degree or two. I read that you only scratch the die/heatsink when you spread the ICD yourself like some people do instead of using the grain-of-rice method, but I agree. Better 1 °C hotter than scratching your heatsink.

I think that with most regular pastes, the pastejob is more important than the paste itself.

 

Im going by the numbers on the battery icon if they are wrong thats on gigabyte not me. If it bothers you that much you could run your own test and compare the data.

 

I'm not a benchmarking kinda guy, and I don't have those latest most-taxing-games except Crysis 3 and TombrRaider, which I wouldn't start play until I finished the previous entries of the series. Currently replay Darksiders 1, CPU maxed at 84 degree C, GPU at 76 degree C. I'm using Arctic-something 5 (leftover from previous pasting on my last laptop).

To be honest, don't expect dramatic drop in temperatures, usually get 3-8 degrees at best.

Regards

 

You're not listening to the points I'm making. I'm not bothered by it, I just brought it up so that you weren't spreading inaccurate information.

 

Actually just realised my math was wrong, In my post i said that with etu running i got 4 minutes for every 1% and without i got 4 1/2 minutes per 1% this equals out to 50 minutes not 1.5 hours so maybe that sounds more accurate?
What is the point your making anyway? The bottom line is that every time i turn on etu my battery life drops by a significant amount and thats all i was trying to say. And that point is still true although its only 50 minutes not an hour and a half.

 

Dear Artedes"
What is the Sceen Size of Gigabyte P35Xv3 and what is their website.

 

Yeah, it's just a small point I was making, so it's not the end of the world if we disagree on it, (that XTU does or doesn't use a lot of battery). You have your solution anyway, and no point in running XTU in the background if you don't need to run it anyway.

 

Its a 15.6" screen

Sent from my GT-I9300

 

See post 1500 on this thread. I repasted with liquid ultra, results are summarised there. My main tests were using firestrike and heaven benchmark tools. It clearly helped, not sure what metric best shows by how much.

 

Anyone know where smartupdate puts the driver files so I can copy them.

I can't find where it puts the wireless driver.

I found something in windows temp but I don't think that is everything.

 

I thought I read the entire thread a week back, I seem to have missed that. Great post. I find it difficult however to see whether the heaven benchmarks are run consecutively or not. In another post you wrote that after a second consecutive run there was some minor throttling (still above 3GHz) with -70 and stock Gpu I think. Do you have any info whether this will become worse after a third or fourth runthrough with your system?