New Clevos with Max-Q?

@Ionising_Radiation Expect around 10% lower in Firestrike when comparing "both 1070 versions". I can't release specific results or temperatures (component or chassis surface) though at the moment.

 

Sorry, Clevo doesn't offer a 1080-Q model and their 1070/1060-Q models where originally designed for non-Q firmware and then 'downgraded' for certification purposes, so you will see better results than from an 'average-Q model'.

 

As someone who has been lugging bulky laptops around for 19 years I would love to see a slim notebook with real power but I'm skeptical about this, my "spidy" senses went off during the presentation. Time will tell.

 

You will spend more but get the most performance per watt you can get, obviously with less watts that's less performance if power was not a limit.

 

But that is such an awesome line! Just stonewalling with something other than "I cannot recall" or "I plead the 5th" (a US thing, invoking our right against self-incrimination). Plus, all of that, then they move the CIA director over at the end. But, I'll admit, I like the franchise (although there is a movie or two in there)...

On topic, I don't think I can be convinced to buy gimped GPUs at a premium...

Sent from my SM-G900P using Tapatalk

 

Just saw AMD presentation, not very promising. No fps values and require two card to run prey at 4k.

 

Here you can see the Clevo system before the Max-Q certification:

http://forum.notebookreview.com/thr...hp6-review-by-htwingnut.805339/#post-10535208

After Max-Q certification the same systems GPU base frequency is going to be reduced to 1113Mhz (down from 1405Mhz)...only because Clevo is afraid that the end-user are uneducated enough to insist on a system with Max-Q sticker.

 

Let's hold everyone to the gimped bs of thin and light... Smfh

Sent from my SM-G900P using Tapatalk

 

Guys guys guys, there seems to be more speculation here than in a Flat Earth Society meeting. Instead of smack talking about these laptops/GPUs first see the reviews and what they perform on benchmarks. I read a website which apparently said that the 3D firemark test(not any more specific sadly) was conducted on the new Triton Predator and it scored about 17,000 which is slightly less than the normal GTX 1080(the website also said they got the result directly from Asus so take that number with a grain of salt). It doesnt SEEM to look that bad considering it(apparently) cuts the power usage and the temperature output by a lot so in my eyes it seems alright. Zephyrus has the same GPU with a cheaper pricetag of 2.7k so I am not terribly concerned about the price(of GTX 1080 Max-Q).

 

MAX-Q !!

 

If that firestrike score is true and is for the graphic score it puts the 1080 Max-Q at regular gtx1070 level...

 

Lol.

Before I bought my MSI GT73 I was looking at the EVGA SC17, I wanted to believe it would satisfy my needs but the slim laptops all had heat and noise issues. I'll keep an eye on the MAX-Q and wait for some real users get their paws on it.

 

While consuming a bit less power most likely.

 

It would make sense that would be like 20w less, or something like that. Less MHz tend to less leakage and better efficiency.

Maybe enough to use less rpms on these thin laptops.

Enviado do meu MHA-L29 através de Tapatalk

 

Oddly enough the 1060 & 1060-Q have the same TDP (depending on vendor)...

 

well maybe aside from extra throttling they actually did put in some work and binned some chips for higher efficiency...

 

If they would have done that it wouldn't need the same power to run lower clocks...rather the opposite.

 

Obviously Max-Q means max-quality which means your games will run smoother than the full power variant!

nVidia Magic!

 

right! duh, brain-fart

 

As an analogy, I used to despise the brand dilution of things like the M badge when I'd see it on a 118i/318i/X1/X3 etc. (paid for as a factory "M-kit" option I mean, not kids buying the knockoffs from ebay). I don't hate on it anymore, these people paying through the nose on these pointless "M package" upsells helps the accountants in head office sign off on the R&D on their lower volume, proper performance cars and technologies.

Nvidia is a publicly listed megacorp, they will screw us all for a buck if they can get away with it, which they are, just look at their financials explode over the last couple of years while their various lockdown - clockblock / overvoltblock / anti-free-use behaviours and quality issues like the 970 3.5Gb fiasco - and broadly speaking its consumer base is still voting with their wallets saying "Thank you sir may I have another"

 

Is it me, or is TDP somewhat inaccurate with Pascal though? Particularly when there seems to be a growing gap between thermal output and power consumption.

Pascal seems to just generate far less heat at the same apparent wattage than previous chips. The other consideration is whether Nvidias specified TDP is a "maximum possible" or "typical real-world" number. Or Nvidia are just doing what Intel does and slap the same TDP across all iterations of their core types regardless of actual thermals or power consumption

What I'm most interested in, is if there's any difference in "Max-Q" systems as far as power delivery. The Max-Q premise appears to revolve around sticking "the next model up" GPU into a laptop and down-clocking it to suit the existing cooling/power system.

Using a 1060 vs 1070MaxQ as an example, it might suggest that a larger GP104@80W is faster than a GP106@80W which is rather interesting. I guess we'll have to wait for a review to see how much power one draws compared to heat output.

 

It's an artificial cap to prevent the chip from exceeding hardware component limits.

More isn't always better, it's how it is being utilized. We can pull 300W+ from a mobile 1080, but all this power just goes waste if not done properly and simply ends in mass RMA.

Like all these people who have recently been flashing a higher TGP 1070 vBIOS on a smaller 90W MXM model and burning through an entire production run by destroying their MOSFETs.
Since then the manufacturer has recalled and stopped selling the model for investigation...And guess where this end-user stupidity will lead to? More locks and total encryption of future Volta GPU firmware to prevent any kind of cross flashes and alterations.

 

Are the power components of weaker quality on those Max-Q models? Or are they on par with the proper cards? If so, would the 1080 Max-Q *if MMX*, work properly with a nice created modified vBios?

 

In general everything from MOSFETs, power phases to bundled AC Adapter is designed for less load to keep production cost at a minimum.

 

Just saying... I was under the impression that Pascal had worse thermal density, and therefore harder to cool, than Maxwell... @D2 Ultima makes this point fairly frequently.

 

From what I've witnessed, Pascal scales very oddly with power drawn, and depends heavily on how the program using it is using it. For example, I could easily get Runescape to suck 99% GPU usage but it'll sit at about 160W drawn for me, but if I make Overwatch draw 99% usage, it'll suck more. Then if I run OW at 4K and make it suck 99% usage (or any game at 4K usually) it'll throttle my clocks far harder as it hits a TDP limit much faster.

The higher the clocks go, it seems it needs more and more power and this doesn't scale linearly. 2100MHz in benchmarks takes close to 300W as Prema said, but what's the real difference between that and 220W for 1900MHz or 2000Mhz? More power isn't necessarily better all the time (though I still feel if 1060Ns were ~100W and 1070Ns were 130W-140W the world would be a much better place).

It is. Compare the 1080N in the P775DM3 to the P775DM1 180W 980N. The difference in temps should be rather night and day, despite both being 180W rated cards.

Pascal is hotter, easily, make no mistake. As far as I've come to understand, the hottest architecture ever released. People just honestly stepped up their cooling this time. I think people forget how cold 970M cards ran in machines that were well-designed like a P770DM or P650RP6. 1060N doesn't even come close to that level of ice cold.

So does this mean that 1080N Max-Q won't cost $1200?

No?

Oh well

 

Thanks But You forgot.... Max profit!!

 

I'd say that's probably best explained by voltage increases. As voltage goes up power consumption goes up exponentially. I haven't really seen any solid explanation for the ~2100mhz "wall" that Pascal runs into but it seems to me that it's due to being very far up the required voltage curve than we're normally used to. The LN2 record on GP106 was done at over 1.5V

I have a suspicion that what we might actually see is something like the 1080MaxQ with 1070 power delivery, 1070MaxQ with 1060 power delivery and so on.
Basically the down-clocked core paired with the next lower tier power delivery.

 

I think this is the bad chip ones that nvidia needs to sell it to the consumers rather than trash it out. Good marketing for them, bad news for us. Anways It's a good thing that there are forums like this one to enlighten others.

 

less power phases, possible if unlocked vbios will blow card up...?

 

Wow man, 2 savage 4 me bro.....

 

That's all on you man, for me it was a fun humorous take on the situation... lighten up

This one is supposed to be humorous, and informative too
http://forum.notebookreview.com/thr...ts-video-articles.805342/page-8#post-10537676

 

Also I dont think you guys understood when I said they werent too specific. There is like a tons of '3D firestrike' scores to compare the performance of the new Max-Q with(it looks like it is in the middle of GTX 1080 and GTX 1070) and all the number 17,000 signifies is that is will be relatively powerful(in my newby eyes). Also dont forget that the actual laptop may be the cause of a lower firestrike score and other laptops may achieve a better score. My point still stands that you guys are being a little too cranky for something that hasnt even been released yet and instead of roasting the Max-Q GPU now you can easily do it later when you get more info. Nvidia isnt being 'lazy' when it first came up with the Max-Q design but instead they are trying to 'patch up' some issues the pascal GPUs may have in the past. Sure you can say that the release of Max-Q is a bit premature and they should have done the release with a desktop grade version of it, but that doesnt mean it is a step in the wrong direction.

 

I dont think you realise what the use of numbers instead of words mean

 

Enlighten me, please explain the joke... what does too savage for me mean to you?

 

One doesnt simply use numbers instead of words in a serious manner lol. Ever seen a serious piece of writing that replaces words with numbers?

 

Ok, I get it, you meant it ironically?

How could I have possibly missed that?

I thought you were trying to soften the blow, and really thought it was 2 brutal 4 u

 

I could be wrong, but shouldn't it be pretty easy to guestimate the MQ performances? I mean, we know their boost clocks are lower than standard edition base clocks, and up until now we're assuming same # of cores anyway.

Simply extrapolating should get you a ballpark figure, right?

 

Performance is not always linear with clock speed, but in theory yes.

 

A single fire strike run doesn't tell the whole story. It's an extremely quick benchmark that barely gets a GPU up to running temps. Id be looking to see how the following runs compare.
It's pretty lame having 1080 that can only sustain 1080 speeds for a couple of minutes before it clocks down and starts crawling to save itself.

 

Exactly. The firestrike score was the only hint of a benchmark test so i put it in this thread, but it isnt very accurate and easy to compare. 17,000 gives a very vague idea of the performance of this GPU but it is good enough to allow a person to see its between GTX 1070 and the true GTX 1080.

 

My thoughts exactly, lets see how it does after an extended gaming session.

 

Am I the only one here that thinks these Max Q machines are going to be really pricey? Like MacBook Pro 15 inch will be a lot cheaper levels of pricey, which at that point, you could get a normal 15" and with the left over $1500 that you would need extra to get a Max Q machine, you could afford a personal trainer and supplements, make carrying 2.9kg/6.4lb a breeze.

 

Oh, definitely. Why do you think Nvidia made them ? you will pay the price for a 1080 with 1070 performance. So I expect them to be $3000+ for the 1080 Max-Q and $2000+ for the 1070 Max-Q. The 1060 Max-Q will probably be $1-$200 more than the regular version. On noteb.com I will make sure to differentiate normal cards from Max-Q. Only issue if I should change the 1080 from the Razer Pro to Max-Q or leave it as a normal 1080.

 

I hope they won't bring this Max-Q bullsh#t to mid-range laptop . My GTX 860M won't reach maximum speed (GPU boost speed , not stock speed) unless I force it via Nvidia Profiles and that alone is annoying for normal user already...Now this Max-Q will make thing worse

 

It's doing what the M series always did as far as I can see.

 

Except the 1080 Max-Q is going to ship alongside fully functional 1080 laptops, and priced the same - maybe more - for less performance.

All the Max-Q GPU's are riding on the legend of the established 1080 / 1070 / 1060 best performance, but it's being sold as a miracle version that performs the same in 1/3 the chassis size.

It's such a disservice to the public image of Nvidia - yeah, how could it get worse - it's simply poor stewardship of the brand.

They could have sold the Max-Q performance as an improvement on the 1060 in slim laptops as they indeed are. Nvidia provides comparison graphs for performance improvements from the standard 1060.

Instead Nvidia names the Max-Q GPU's as 1080 / 1070 to fool the public into thinking they are getting more than they are going to receive.

Nvidia could have made a big enough splash at getting 1070 like performance into such a thin laptops, instead of deceiving customers into believing they are going to get full 1080 performance.

We've already seen people get excited at finally getting a 1080 into a super slim laptop. It's named a 1080 it must have the 1080 performance we expect, right?

Maybe Nvidia did capture some Max-Q "mini-wormhole's" allowing laptop makers to build cooling that will magically suck away the thermal problems and indeed deliver full 1080 performance.

 

Well they are the same chip, same ram so of course the cost is also going to be there

In a sense they are not really much cheaper to produce.

 

But the result from their use is less value than full performance, so they have been designed to be overpriced for their value.

Perhaps there are a % of GPU's that otherwise don't make the grade at full performance that are being utilized as gimped Max-Q GPU's?

More likely as you say Nvidia is finding homes for as many 1080 GPU's as they can, and since people want slim laptops, Nvidia is gonna give it to them - at full price + a premium for the wonders of Max-Q.

It's not a 1080 performance that's received, so it shouldn't be called a 1080 anything.

Call it a 1075, 1065, and 1055/1060 (depending on performance delivered).

Or, call it a 1080m, 1070m, 1060m, as the public understands what that means, reduced performance mobile GPU's that parallel their desktop GPU brothers, but aren't tuned as high as desktops due to thermal and power constraints.

Which is what the 1080 Max-Q is, right? A detuned version of the full performance desktop / full power laptop.

It's disingenuous to change the name so the public needs to figure out the "scam" all over again.

Maybe we should all start calling them 1080mq, 1070mq, and 1060mq, perhaps that is close enough to the detuned part naming mobile GPU to clue people in it's less than it's number name suggests?

These new Max-Q owners will eventually discover the same things as all the disappointed Razer 1080 owners, now finding out that not only do their 1080's perform like 1070's, their 7820HK OC'd perform more or less like 7700HQ stock CPU's because there isn't enough power allotted to get more than 45w sustained performance.

Nvidia Max-Q has created a whole generation of gaming laptops with disappointing gimped performance engineered into the laptop that won't meet up to the owners expectations for full 1080 / 1070 performance - following the path of the Razer 1080 Gimp Edition, also engineered and produced to deceive and eventually disappoint their owners.

 

GTX 970 3.5GB VRAM all over again . ̶N̶̶v̶̶i̶̶d̶̶i̶̶a̶ ̶n̶̶e̶̶v̶̶e̶̶r̶ ̶l̶̶e̶̶a̶̶r̶̶n̶ . Oh, they did learn sth this time , at least they told us before shipping lol