Clevo 2019

Haha, yeah, but I'm guessing that notebook heatsink screws bottom out and are height controlled, so you'd have to make the gap up with something.

One of those new thermal pads?

 

How close is the RTX 2070 in the new Sager NP9176 to other mobile RTX 2070s? Also, how does it compare to the RTX 2080 Max-Q? I'm probably going to go with a 2080MQ anyway. I realize the combination of CPU and GPU is what actually matters for graphical performance. I already know that the Schenker i7-9700k + 2070 is about the same as or slightly better than the i7-8750h + 2080MQ.

 

Want 2080Max-Q ? http://forum.notebookreview.com/thr...iscussion-thread.827176/page-11#post-10858439

 

Oh yeah I forgot to watch that video. Watching now. I have a much lower than usual price on the 2080MQ however. If not for that, I wouldn't consider it. Will get back to you in about 3 hours.

 

Desktop CPU models have high clocks in region allowed by Nvidia.

 

Max-Q versions

Code:

NVIDIA_DEV.1E90.7509.1558 = "NVIDIA GeForce RTX 2080"
NVIDIA_DEV.1E90.7709.1558 = "NVIDIA GeForce RTX 2080"
NVIDIA_DEV.1E90.95E1.1558 = "NVIDIA GeForce RTX 2080 with Max-Q Design"
NVIDIA_DEV.1E90.97E1.1558 = "NVIDIA GeForce RTX 2080 with Max-Q Design"
NVIDIA_DEV.1ED0.7510.1558 = "NVIDIA GeForce RTX 2080"
NVIDIA_DEV.1ED0.7710.1558 = "NVIDIA GeForce RTX 2080"
NVIDIA_DEV.1F10.65D1.1558 = "NVIDIA GeForce RTX 2070"
NVIDIA_DEV.1F10.7509.1558 = "NVIDIA GeForce RTX 2070"
NVIDIA_DEV.1F10.7709.1558 = "NVIDIA GeForce RTX 2070"
NVIDIA_DEV.1F10.95E1.1558 = "NVIDIA GeForce RTX 2070 with Max-Q Design"
NVIDIA_DEV.1F10.97E1.1558 = "NVIDIA GeForce RTX 2070 with Max-Q Design"
NVIDIA_DEV.1F11.65D1.1558 = "NVIDIA GeForce RTX 2060"
NVIDIA_DEV.1F11.7509.1558 = "NVIDIA GeForce RTX 2060"
NVIDIA_DEV.1F11.7709.1558 = "NVIDIA GeForce RTX 2060"
NVIDIA_DEV.1F11.951D.1558 = "NVIDIA GeForce RTX 2060"
NVIDIA_DEV.1F11.95E1.1558 = "NVIDIA GeForce RTX 2060"
NVIDIA_DEV.1F11.971D.1558 = "NVIDIA GeForce RTX 2060"
NVIDIA_DEV.1F11.97E1.1558 = "NVIDIA GeForce RTX 2060"
NVIDIA_DEV.1F50.7510.1558 = "NVIDIA GeForce RTX 2070"
NVIDIA_DEV.1F50.7710.1558 = "NVIDIA GeForce RTX 2070"
NVIDIA_DEV.1F51.65D4.1558 = "NVIDIA GeForce RTX 2060"
NVIDIA_DEV.1F51.7510.1558 = "NVIDIA GeForce RTX 2060"
NVIDIA_DEV.1F51.7710.1558 = "NVIDIA GeForce RTX 2060"

 

It will be interesting to see what balances designs come up with.

 

How’d you manage that?

 

So I did some digging around and checked all the sellers on both Taobao and in the USA, finding that it is best to purchase a 2070 from Rjtech, even with CA taxes on me.
1075$ vs taobaos lowest at 860~ and eurocoms at 1050$ both without shipping and and no new heatsink. With shipping/heatsink outside of CA, it'll be closer to 970/980$.

 

Heat from GPU exits to CPU plate, is absorbed by the heatpipes next to it. The core of the GPU is under the middle point of the heatpipes so the heat goes out like <[]> instead of []>>. Since the plate on the CPU side is not as big as the GPU side, it transfers heat quicker without affecting temps from the CPU by much (because of the CPUs integrated heat spreader, it already has a thermal sink and doesn't require much of an additional one.) All in all it'll be an increase in cooling factors for the GPU.

 

Except at peak loads, it just helps gpu temps at idle and gpu biased workloads.

 

I was looking in to this taobao thing but without knowing Chinese language it might be hard or even impossible. Can you recommend any website that is translating it for ex English?

 

Walmart.

 

It's kind of like going through a market there are deals to be had but you need to be careful.

 

So how bad will these 16.1" Clevo P96 underperform using the same 180w PSU?
P960ED RTX 2060 (I'm probably fine)
P960EF RTX 2070mq (I wish I was 230w)
P960EN RTX 2080mq (forget afterburner)

 

8/16 H-cpus are on the way) Wish Clevo could install those instead of 6-core 8750H

https://twitter.com/TUM_APISAK/status/1093101201309163520

 

More cores on the same process with the same power limits. So you would rather those over the socket CPUs?

 

Well, for slimmer models there's not much choice - H cpus or bust) at least some overclockability wouldn't hurt as they are unlocked.

Models with socket cpus are waaay too old design wise with not too much of cooling improvement, if any, of 2 currently available models.

Sent from my SM-G930F using Tapatalk

 

With the new Alienware Area 51m release, I'm hopeful that Clevo decides they need a new design model with a modern, high-tech cooling system rather than a refresh of an old design.

 

I actually exclusively use their site for low level items like heatsinks or specific boards impossible to come by in the USA. I use Google translate through my phone with a feature that converts screenshots to text.

 

If it ain't broke.
If you want high tech heat removal it's a far cry from what we have at the moment.

 

In regards to your first sentence, I'm not sure I understand. Are you saying, if it ain't broke, don't fix it? As it is definitely broken with the latest CPU/GPUs. All one has to do is look at the amount of thermal throttling we see in the i9-9900K, 2080RTX at stock levels (definitely in the P750TM, and I think the P775TM also has a bit of an issue), they will have to use the vapor chambers and do some more heatpipe/exhaust redesign to go beyond what's currently released.

 

Usually refreshes are saved for actual chipset changes so that effort is not wasted.

 

That is fine by me. If they solely stick with Intel solutions, they can be sunny, icy, covey, lakey or whatever the case may be in 2019-2020, just as long as Clevo really thinks through and implements a solid design around at least a few non-BGA models.

 

Update: SAGER’s lineup:
NP8961 RTX 2060
NP8962 RTX 2070 Max-Q
NP8963 RTX 2080 Max-Q

Both of these have a low price for 2080 Max-Q compared to other slim mq’s but a 180W power adapter at 19.5V, 9.23A is also the smallest I’ve seen in the MQ 2070/2080 category.

I imagine the previous models NP8954 / P955ER with the 1070 Max-Q and i7-8750H must of had its share of power throttling using only a 150W PSU.

Otherwise I do like the thicker chassis for cooling with better fans and hidden finger print sensor, and slimmer bezels to fit a 16.1” 144hz 1080p display.
edit add:MQ

 

Certainly the 2070 max-q wont turbo as hard as the 2070 but that's kind of the point.

With the 2080 you get more RTX cores for ra tracing which will help that side.

 

For Max-Qs I meant:
compared to other slim mq’s but a 180W... the smallest I’ve seen in the 2070/2080 category

I'm more concerned about the 150/180W vs the majority of 230W Max-Q laptops out there.

 

Well, if the TDP really is 80-90W I think that a 180W PSU is enough. I have a 980M, with (AFAIK) the same TDP and the same PSU, and I have experienced no problems so far.

 

the 980M is actually closer to 125W

 

Still it had no issue with the 180w PSU, at least in Clevos, right? I remember Alienwares having problems and requiring a 230W one though

 

Yeah I suppose if a modern GTX 1060 i7-8750H laptop with a 144hz display that comes with a 180W power adaptor has the same power consumption as RTX 2080 Max laptop there should be a little wiggle room left for overclocking right?

 

Well I don't think you will have much room for overclocking a 2080mq with a 180W PSU. The system will probably be already very close to this limit at stock already - both in terms of power and of thermals.

But at this point it's just speculation. We need to wait proper reviews.

 

Doesn't seem like it's turning out that great now that the gag order is lifted. Not quite a match for 2080 FE, and sadly, it won't work in CrossFried configuration.

Maybe the drivers are not as great as originally thought. Still waiting for more conclusive evidence on whether or not it it will eventually be overclockable enough to be taken seriously as an enthusiast GPU and not just an alternative for a gamer-boy GPU. According to GamersNexus, overclocking is totally broken (as usual for AMD, especially their GPUs). For now, the Green Goblin is still holding the crown and probably won't feel much pressure from the release of this product.

 

A power hog before you start overclocking. No power efficiency from 7nm due they had to up the clock speed Same old but in new clothes.

 

That is the trouble when people get hung up on a magic number, like "7nm" and everyone gets excited about it. Smaller is not automatically better. At the end of the day, you measure results. If the results are better, it's a better product. If not, then the product is either a waste of time and money, or just a total flop if the numbers go the wrong direction. I think in this case it is a better product, but still just not good enough to want it. And, not any better than what we already have. Had they released something like this in 2015-2016 as competition for Maxwell, it would have been really exciting and amazing.

 

According to Notebookcheck, the Razer Blade with 8750H, RTX 2070 Max Q and FHD 144 Hz screen consumes a constant 155W during Witcher 3 and gets up to a maximum of 182W (having a 230W adapter).
It will be interesting to see what the results are with the 180W adapter in the Clevos. Let's hope it's enough to not limit the system

 

I believe you are right. I also want to know if this is 80w or 90w 2080MQ. Heck I wish we knew what MQ we are getting for all models.

 

180W wasn’t a problem on the P650SG because it would disable Turbo Boost on the CPU when the GPU was loaded. After flashing a non public release EC mod on my P650SG that disabled the CPU throttling under GPU load, it could fully utilize the 230W aftermarket adapter I bought when running both CPU and GPU overclocked.

 

If this Acer Triton 500 Review is any indication - 180W PSU will cause battery drain during gaming sessions...
"This laptop charges via a standard barrel-plug and our pre-production sample was delivered with a compact 180W charger.

That’s however under-powered for an RTX 2080 Max-Q configuration, and as a result the battery slowly discharges during demanding gaming sessions, so you’ll have to take breaks after each couple of hours of playing. We’re looking for clarifications whether the final versions will also ship with a similar 180 W charger or a more adequate 230 W brick, which other OEMs are bundling on their RTX 2080 Max-Q variants."

I am wondering if an aftermarket slim 230W power brick would be an option, the price is still quite good for the spec....

 

On the same arch it certainly better, this is just a workstation card that was only feasible because Nvidia set the prices so high.

 

Oh I totally forgot about that review! Yeah I wouldn’t mind picking up a compatible 230w later if that fixes the drain issue. Kinda messed up though.

 

Now I’m really torn now. You could save a lot going with the P960EN vs the GS75 especially if the $2800 GS75 is only single channel. I guess I will wait for reviews. I don’t mind buying an larger power brick if one can be used travel and the other left home for greater power usage.

 

I think this is the part that really gets me. We've gotten to the point where the hardware is powerful enough and price ranges are so wide that market segmentation is an entirely different beast than what it was when I last bought a laptop 5 years ago.

This isn't even a viable product if RTX was better value or 1080ti production continued.

 

Yup, that's what I thought when I first heard of the lower wattage PSU's being shipped, that they will be using battery boost techniques to make up the difference.

So it might actually be worth plugging in a compatible PSU with higher wattage to see if the laptop + battery will draw more power and stop the battery boost from engaging.

I have done that before for a wide range of laptops, but recently - last couple of years - the firmware has been locked down to assume a maximum power draw based on the shipped PSU wattage... it's worth a try though.

 

Yeah a decent review would be really helpful right now lol.

I honestly don’t think a 2080 max-q is worth more than what clevo/sager are asking from the reviews I’ve seen... If this unit ends up being a dud then I will probably either go with a cheaper 1080mq unit or just hold out...

 

I think that the sweet spot for me is the 2070 (either maxq or not). 8 GB di RAM for my dev needs and enough of an update for vidya from my 980m. And, most importantly, is not overpriced as hell

 

Yeah the 5lb 17” Mech-17 with a Full RTX 2070 seems respectable at $1700-$1800. I get pulled in so many directions.

 

Vega is that expensive to produce though, half the cost is the ram. There is a niche of professionals who will find this a bargain.

Totally unusable in a notebook of course.

 

Too bad that AMD is years behind in the field of machine learning. I would kill for a 16 GB CUDA capable video card under 1000 $/€.

Tell that to Apple...

 

I was just reading the Radeon VII review on PC Gamer. Having many Tensor cores is a must! To quote: “As for deep learning, AMD has decent FP16 performance—about 28 TFLOPS, or twice the performance of FP32. It's not at the 80 TFLOPS level of the RTX 2080's Tensor cores, never mind the RTX 2080 Ti's 108 TFLOPS, but FP16 isn't used much in games anyway. For most gamers, it won't matter.