New Clevos with Max-Q?

That is my concern...obviously I am not interested in the whole Max Q thing, overpaying for an underperforming card...so I am interested in uprading in the next round...volta/coffelake or something...but if they implement all these max q stuff from the beginning it will take a while to sort out all the options for the best upgrade path for me. Hopefully my current laptop holds up till then...

Sent from my SM-G930F using Tapatalk

 

I must add, by now it's likely cheaper for Nvidia to produce there Max-Q chips using 1080 dies anyway. It's likely the left over junk on what should be pretty good yields by now. No one is desktop land is buying 1080's over the TI, they otherwise get the 1070 if they're value orientated. I don't see them suddenly changing ethos and going above the call of duty on their dime.

 

Maybe the Redmond Mafia will sell a Surface with Max-Q. Seems to be a good candidate for fancy overpriced disposable crippled filth.

Literally disposable digital dung. See the video. (thanks @Mobius 1 for the link)

 

What you seen on this video is the future for all laptops, and from every brand Nice
And people complaining on Apple

 

Microsoft is like....right to repair? Sure go ahead and try...as long you get duck tape....it might stick together.

 

Technically, if Max-Q is designed to run most efficiently then they would actually be higher binned parts. Better binning means you can tighten the voltage variance, especially at lower clocks. Lower binned parts are easily overcome by throwing more voltage at them which is the opposite of efficient. That being said, binning variance has shrunk dramatically recently anyway.

So from a value standpoint, they would been worth more (their production cost is identical though).

 

^This . People who is going to buy GTX 1080MQ prepare yourself , your about to get trolled by Nvidia (Nvidia released GTX 1080ti/GTX 980ti a few months after Titan X Pascal/Maxwell)
Volta is coming this year , or at least first quarter 2018 if you don't trust Wccftech
http://wccftech.com/nvidia-volta-v100-gpu-tesla-v100-pci-express-announced/
Remember GTX 1060 ? it's on par with GTX 980
GTX 1160 vs GTX 1080MQ . I bet $20 Volta gonna kill it badly , in both power consumption and performance

 

Could be. Just as much as it could be a way for nvidia to sell GP104/106's that didn't make the power/volt/clock targets to be called a standard 1060/1070/1080.

I guess we'll just have to wait to for some carefully constructed tests to see whether these are any different to a standard 1060/1070/1080 running maxq clock/volt/power profile.

 

I wouldn't put too much stock in that (especially given WCCFs reputation of speculating out of thin air). Volta was announced at ISC which is a supercomputing/computation conference and I expect we won't see a consumer Volta card for a year. They're not going to cut their own life-cycle short now, especially given AMD still hasn't released Vega yet and the RX400/500 series is basically sold-out worldwide due to the mining boom.

As far as performance is concerned, Pascal->Volta is only shrinking from 16nm->12nm so the performance increase won't be that big. You can see the FP32/FP64 perf numbers here for both P100 and V100:
https://en.wikipedia.org/wiki/List_of_Nvidia_graphics_processing_units#Tesla
Per-SM the raw performance only jumps ~5-10% which is about what you'd expect from the relatively small process shrink (nowhere near as drastic as the 28nm->16nm shrink we had with Pascal).
That being said, there's about 40% more cores on the V100 for the same power budget so overall per/watt appears to jump by the same amount. I highly suspect the large jump is due to the use of the "Tensor Cores" which are "mixed precision" FP32/FP16 cores, so we don't yet know if they will be included in GeForce cores, or even useful in gaming scenarios which benefit mostly from pure FP32.

You're not getting me. The point of Max-Q is voltage efficiency. You cannot get that by using the lower binned parts, you would have to use the higher binned ones.
ie. Theoretically, a 1080MQ would reach the same clocks at a lower voltage than a regular 1080 under full cooling and power delivery (assuming you had the same vBIOS without limitations etc etc).

Personally, I don't think there's much binning at all. Almost all Pascal chips perform the same which makes binning rather pointless.

 

I'm also highly skeptical that they'd be offering highly binned chips. If ever there's a legit reason for a 1080 to underperform it'll be from wedging one in a thin'n'light, they can at least run with that being the reason for performance deficits .Honestly, although I don't have to like it, this is a pretty good and timely way for them to clear out their leftovers.

 

Binning will likely be irrelevant once the brain cancer has metastasized. I suspect most self-anointed 'hardcore gamers' don't even know what that means anyway.

 

Well said.
By the way: nVidia must be completely not afraid of incoming AMD mobile VEGA if they launch "retarded" 10X0 series.

 

If it was **ONLY** lower voltage - and all else was equal - I'd agree that means a better binned chip
BUT
MaxQ spec compared to vanilla is lower voltage AND lower base and boost clocks AND lower power limits AND lower performance. You cannot impute chip quality from this scenario because every piece of silicon ever made behaves in this way. Undervolt your 6700HQ, turn off turbo, restrict TDP lower than stock, and run tests on power draw and temps, did you just magically make yourself a super efficient Intel 6700HMaxQ?

And would you pay more for that nerfed CPU than a regular 6700HQ.

 

To judge chip quality you would need a characterization of the process, how the average chip reacts to voltage vs frequency and see what % of chips can manage to hit that set of values.

Without that information you are just guessing.

 

That's right. Not even ASIC quality can be identified any more. I doubt that is by accident and suspect it was a self-serving strategic self-preservation maneuver for the NVIDIOTs.

The real test will be how well they perform and what their capacity is for being overclocked to perform like a normal GPU bearing the same title. Capacity for overclocking is where the rubber meets the road for CPUs and GPUs. The good ones will take all of the crap you want to dish out and the bad ones will not.

 

The goal would be to make the upper TDP close to the thermal limit of the machine and the GPU would clock itself up as far as cooling really allows.

 

People make strange decisions. Expect lots of people will

Aka more crippled you make the machine the more crippled will the performance be. You can't put in the bigger cooling from the P870 series into laptops in Razerbook's size. And all of you know how this Apple similar laptop perform.

 

Apple also has a closed unibody design and are as thinner.

 

But they don't put in 1080 performance. Maybe a reason they don't screw up even more?

 

Again..missing my point entirely.

All I'm saying, is that a "high binned" (ie high ASIC quality) chip will have better voltage characteristics across the ENTIRE clock range, both high AND low. They're literally more efficient at any point regardless of clockspeed. So lets say you have a fixed power budget of 100W, then if you want the absolute BEST performance WITHIN that power, then you use the highest binned chip at your disposal and clocked to suit.

To summarise:
Is Max-Q using high binned chips? I don't know.

Would using highest binned chips achieve the best possible perf/watt? YES

Is there any point to using high binned Pascal chips where per/watt doesn't matter (ie desktop cards)? NO (due to the power wall at >2000mhz)

Should you pay more for it? Depends.

 

Since when are apple the gold standard for people who actually know what they are talking about?

 

Harsh, but zing!

 

ASUS GX501V Zephyrus Unboxing!
Still unreleased...so no power on, just unboxing.

Release countdown page...
https://www.asus.com/Laptops/ROG-ZEPHYRUS-GX501VI/

 

I really don't like the keyboard/mouse layout.

 

That's how they made it thin with enough cooling to get the 1070 level performance out of the 1080.

Trying to put the keyboard on top of everything would make it too thick, and the whole design is to be thin.

MSI did the same with the GT80 series, but in this case it was "relative" thinness, and more to fit the full sized mechanical keyboard.

 

you're going to love how the acer max-q machine look like

 

It's got that crazy touchpad close to the screen between the screen and the keyboard, wild...another $3000 1080 laptop that performs like a 1070...
Acer Predator Triton 700 hands-on: Thin gaming laptop

https://www.windowscentral.com/acers-predator-triton-700-will-use-nvidia-max-q-and-gtx-1080
Acer Pancakes a GTX 1080 Laptop: Acer Predator Triton

 

What good is an ultra-thin, ultra-portable, crippled and overly expensive laptop if you'll have to carry accessories like a palmrest and LAN dongle around anyhow?
One could just as well get a proper, uncrippled laptop instead if you're going to need a spacious bag anyway.

That said, I find the "lifting for more airflow" to be an interesting concept with some potential, although it does look rather fragile on this unit at least.

 

Damn nice touchpad design if you don't use a external mouse. Trash being even worse. Maybe as well use a touchscreen. Everything comes to a cost. Smart move from Nvidia.. Fully locked down for vBios mods for the normal graphics so Max-Qrippled don't look too bad in comparison. And yeah, all of you looking forward to Volta. Max-Qrippled with Pascal was only the start shoot for more ugliness from Nvidia.

 

That lifted gap open to the interior of the laptop, flush with the table, looks like a magnet for drink spills... that would be a spectacular flame out.

A drink spill would flush into the fans - blowing all over the inside of the laptop - coating everything with that beverage.

It might even provide a flying sparks show as it shorted out...yipe!!

 

If you listened to what he said during the unboxing. 4-5 times Gtx 1080. Even fully Gtx1080. He does not mention that this garbage is the power gimped Max-Q variant. How many buyers with low or lack of knowledge will be dazzled with fake advertising? The Sellers and websites will go over dead bodies to try to sell this ****y. Mark my words.

 

Yup, it will be interesting to see how many market these POS unscrupulously to their clients.

A good way to pick a seller, not one that sells Max-Q laptops as "1080's".

 

Well, we saw that coming. At least in these threads. He should be referring to it as 'the MaxQ' but yeah...

 

Right, to cover their butt's sellers should always add that Max-Q before the 1080, otherwise when their customer calls / emails complaining that their "1080" isn't performing anywhere near what a 1080 should be performing, they can say, hey - we said it was a "Max-Q 1080", not a "real" 1080

 

Yep, preempt the inevitable disappointment.

 

The inside of the triton:

 

That kind of layout has taken way longer to come into force than I expected. I suppose M.2 drives are a fair amount cheaper now.

 

The cooling looks solid on the Predator.

 

What the 21 inch one? It won't matter to many people.

 

I have only had time to scan through the videos, I assume there are no benchmarks yet?

I would also say that we here can see the scam but many won't and I don't blame the companies for trying and I wouldn't blame a re seller for selling a top item as long as they don't misrepresent it, they all have to make money and it's always up to the consumer to protect himself/herself. I don't trust any big company, it doesn't mean we can't do business however, I'll go in with as much knowledge as possible and call BS when i see it.

 

I doubt the manufacturers will ever mention the Max Q in their marketing....hopefully some resellers will point that out.
The Asus marketing is WORLD'S SLIMMEST GAMING LAPTOP WITH NVIDIA GEFORCE GTX 1080.
Hopefully in the next few days benchmarks will arrive for the zephyrus, then later from the other machines in line for release.

 

The processor thermal throttle in Acer 21X. Why should the new Acer Max-Q models be any better? New ain't the same as better.

 

You never can tell between models, you always need to see reviews first to be sure.

 

There are 2 1080 Max-Q benchmarks that I have seen, they show +1% faster than a 1070 laptop average result, and +3% faster than a 1070 laptop average result, both about 15% slower than a 1070 desktop average result.

Here is one, I'll see if I can dig up the other one... @Papusan ?

NVIDIA GeForce GTX 1080 Max-Q
https://www.notebookcheck.net/NVIDIA-GeForce-GTX-1080-Max-Q-GPU-Benchmarks-and-Specs.224730.0.html

Benchmark results from their test:


Notebookcheck has pages for 1060 Max-Q and 1070 Max-Q results, but no results posted yet...

NVIDIA GeForce GTX 1060 Max-Q
https://www.notebookcheck.net/NVIDIA-GeForce-GTX-1060-Max-Q-GPU-Benchmarks-and-Specs.224734.0.html

NVIDIA GeForce GTX 1070 Max-Q
https://www.notebookcheck.net/NVIDIA-GeForce-GTX-1070-Max-Q-GPU-Benchmarks-and-Specs.224732.0.html

 

That test on notebookcheck was taken from the Acer presentation of the Triton in April. It is the only benchmark I have seen so far for a Max-Q.....lets see how the Zephyrus does soon.

 

You won't see any performance figures until the NDA lifts in a bit over a week. Said NDA may also have some bearing on how machines can be marketed (ie saying if they use Max-Q or not).

 

Right, 1 benchmark does not a trend make

There could be lots of reasons why performance was so low; like under powering the 1080 by 50% (90w vs 180w) so it can't reach full performance.

Perhaps there will be some magic that allows the Max-Q laptops to defy power and thermal limitations and outperform our expectations...

 

Everything that is mobile platform e.g laptops, tablets, smartphones and other portable devices are disposable...nothing in that platform can pass disposal, not even Clevo.

 

Still, Magic can't change the Max-Q to perform as the real cards. Even God can't do that. First time crippled... Always crippled!!

 

Well, when you get down to it even large tower desktops are disposable, too. So are cars, truck, trains and planes.

Looks at that video and you will see what I am referring to. It's inexcusable.