Clevo 2020

SLI GTX1080’s outperform a RTX2080Ti in BFV.

it scales so well, it is nuts. After watching a video it made me want a P870. You could run high refresh rate with 4K on (2) 1080’s

 

Is there a way to use both vapor chamber contact plates to a single die? How does this work?

 

Last I checked, Battlefield V doesn’t even support SLI.

For single GPU, only one side of the vapor chamber heatsink is used.

 

You use Nvidia inspector tool to enable SLI







So you could possibly run a shim that contacts both side of the vapor chamber.

 

Interesting. Do you know if there are graphical glitches when forcing SLI in this manner?

 

I have no idea. SLI usually causes high latency. And those test are all in 4K. It was a video on YouTube. Just search “BFV SLI GTX1080”

 

Yeah, I was using SLI up until about 2015 and absolutely hated it. The input lag, the microstutter, the graphical glitches, the hours upon hours of Inspector tweaking to try getting it to work in games without official support, and just the overall hacky nature of the tech. And from what I can tell, SLI is even more useless for gaming these days, as the majority of new games are released without official support, and Inspector tweaking doesn’t work for DX12/Vulkan. SLI is only good for benchmarks now.

Anyway, I found that video on YouTube. The microstutter on the SLI setups is really apparent from the framerate/frametime graph compared to the single GPU. And it looks like either the SLI setups are running lower graphical settings to make their performance look better, or the graphics are glitched out, because they are missing things like leaves on the ground, raindrops in the air, and various particle effects during parts of the video.

 

Hey what is your input lag like in BFV?

It feels smooth in gameplay with my GTX1080. I have vsync off in game, future frame rendering off in game, and my monitor overclocked to 154Hz.

It says I’m around 6-9MS in game averaging 124FPS, Maximum 154FPS, minimum 91-101FPS

But when I look at my latency graph in MSI Afterburner it looks like hell lol. A lot of high spikes, up and down graph.

But I’m just not feeling that. It feels butter smooth to me. No skipping, with just smooth performance.

Gsync is on, and vsync is set to application controlled in the nvidia CP.

I am curious what your msi AB graph looks like after playing BFV.

 

Never had any latency issues with my P377SM-A with dual 880M's back in the day. But I prefer a single gpu to a dual gpu in SLI

 

SLI is great for benching.

That was it about 99% of the time for me.

Even back in the day when everyone was all over SLI, I would get an SLI system and immediately downgrade it to a single card from the 400 series through the 900 series, it was my MO.

I've always preferred a single GPU even in my desktop.

 

I wish someone would make a GTX Titan XP in MXM with a 265TDP. We could get some serious work done!

I had a Titan XP Starwars limited Galactic Empire video card. That thing was a BEAST! it was only about 15-20% slower than a RTX Titan.

 

Nvidia said no to a ti mxm module.

 

Great move by Nvidia - after all who would want a really powerful laptop anyway...

@Meaker,
while we have you here, could you comment on the approximate sales numbers percentage of the 751/775/870 chassis in relation to total sales of Clevo built laptops at Sager?
As these have been in production since 2015 I would expect that lower sales numbers did not make a complete redesign desirable for Clevo hence the many iterations that we have seen of these models.

 

I don't have the equipment to measure input lag, but subjectively I've always felt that BFV has higher input lag than BF1. I assume you mean frametimes though? My frametimes are generally pretty smooth even when CPU-bound, aside from the occasional odd spike. Still overall not as smooth or responsive as BF1 though. I switch between both games regularly and the difference is noticeable.

I'm using DX11 with FFR On. FFR setting doesn't work in DX12. I've heard DX12 has worse frametime consistency than DX11, so you might try DX11 if you haven't already.

 

The vapor chamber is one giant shared unit, no need to have both plates contact the gpu die for the card to get benefits from +400W capable heat dissipation.

As per lag, what you're likely noticing is likely fps variation (client side) + slight delay (server side) that messes with your muscle memory, making the game feel out of pace. If you were to have a constant 144fps that choppy feeling would be greatly reduced. Which is why most "pro" players play with low detail settings, both for framerate consistency and their personal performance. Trouble i when you play on a stage where either publisher or alike force them to play on max setting to show the game off in the best of light.

 

Man that 8600K takes a lot of voltage. I really see what you mean. Temps look great though for 4.7Ghz. BFV really hammers a CPU just as hard as any stress test.

You may as well be 5Ghz with -3 AVX offset.

It feels smooth to me, and it does not feel choppy. I’m kinda just wondering my my frametimes look so crazy on a graph. I’ll post a graph up tonight. So you guys can see.

 

Likely from frequency fluctuation as per GPU Boost 3.0.

 

Your GPU temps look great. But then I saw the GPU utilization.

It looks Like you may be CPU limited. Even with my 8086K clocked extremely high, during very action packed scenes, BFV can knock my GTX1080 off balance and utilization will drop some when all 12 threads hit the high 90’s% percentage in usage.

I would increase the BFV resolution slider until you get full 99% GPU utilization. And until your FPS stays the same as where it is at now. Then you will gain image clarity without losing performance, or wasting GPU power.

 

My VID is on the higher side probably because BFV uses AVX instructions. The VID bounces around a bit while gaming, unlike in benchmarks/stress tests where it stays pretty much locked. I'm usually around 1.25-1.3V VID while playing BFV, what about you?

I'm using auto fans as I prefer the acoustic profile. Max fans is about 10C lower on CPU and GPU.

You’re probably not as sensitive to microstutter if I had to guess.

I'm definitely CPU-limited in BFV. I'm deliberately playing the game on almost the lowest settings for the best visibility and performance/input lag. I never let my GPU get maxed out in multiplayer games because being GPU-bound causes a huge spike in input lag. I prefer having the best responsiveness over eye candy in multiplayer, since I never slow down to "smell the roses" anyway.

 

I love BFV. Nice to see someone else who is enjoying it as much as I am. Yes the CPU limitations are real. Running ultra detail with TAA on Max my average is around 125-130+ and my minimum is like 91 I think. I feel competitive, and the graphics are excellent! I do pretty well playing scout recon mostly but every now and then I stop to smell the roses and enjoy the view. The game looks stunning!

It already feels like a Ferrari when I’m playing. I’m only a like level 13. My origin account was hijacked, and I had to start over from scratch. I’ve only been playing this account last Thursday

 

Oh and I’m not sure what my cpu voltage is in game. I’ll check though, I don’t have that in my OSD options didn’t know I could see my voltage during gaming.

 

Haha, idk about enjoying it as much as you. I’ve always preferred BF1 over BFV, and after they ruined the TTK in BFV earlier this month that pretty much killed the game for me. I’ve been playing BFV for about a year, 200 hours so far, around 3 KDR and 1.55 KPM overall. Recon is my least played class as I don’t like sniping, although since they nerfed all the other weapons, the sniper rifles are more powerful than they used to be.

I’m using HWINFO for the GPU voltage/power and CPU VID in my OSD.

 

Such a skilled player like yourself needs a i7/i9. Seriously though, you run all low settings? Do you run g-Sync? Refresh rate on fast? ULMB?

 

Running BFV with low settings in such computer is a no go. I've played 200 hours of bf1, 100 hours of BFV and 3200 hours of BF4 so far (yes 3200). BFV is dead for me. They keep screwing the game and not giving a damn about the community and not even caring at all
I've only played hardcore mode in 2 games and it's my main mode in BF4.

Anyway I've always played BF4 in Ultra with 1080p and 125 % resolution scale with stable 144 fps with g-sync on

 

Haha, I’m not that skilled. I have textures, texture filtering, and mesh on ultra, everything else on the lowest or off. My panel is 120Hz G-Sync. I use G-Sync and VSync enabled in Nvidia Control Panel, and BFV capped at 117 FPS using its built-in FPS limiter.

 

You tried overclocking your panel?

 

No, I can't share that sort of data, sorry.

But in general yes the desktop models get refreshed less often, especially when the chipset basically does not change lol.

 

Not on this one, although I’ve done it on other laptops in the past. The Lenovo Y500 I had a few years ago I overclocked the panel from 60Hz to 120Hz.

 

I always missed the way you could do some real damage like in BF3/4. An opportunistic flank could let you wipe 2-3 squads as a single man.
However, in BF1/BFV you're lucky to mow down a couple people before you run out of ammo.

 

Really depends on the weapon in BF1. I frequently wipe out squads with the cheese cannon that is the SMG 08/18. Before BFV’s TTK shaft it was also fairly easy to pubstomp with weapons like the Lewis Gun and the large mag SMGs.

 

Most of my BF1 was on the PS4 with friends, as a medic w/ autoloading 8 ext and assault w/ model a10 hunter.
https://battlefieldtracker.com/bf1/profile/psn/amn1xou

I got my p775 for gaming BF games but as luck would have it we only played BO4.

 

That's why BFV sucks for me. Playing only hardcore I love playing in my own corner with no killcams and 3d spotting. BFV before the 5.2 patch was OKish but 5.2 simply made it impossible for me to enjoy the game. I finally uninstalled it for good.

There are active BF4 servers still though

One game I've played too is bad company 2.. And bf3 once in a while and I enjoy them..

Now going in a kill rampage in BFV and having to shoot 30 bullets to do 20 damage (because you miss most shots due to distance) to kill someone (when the charger goes off)? Forget it

 

Dang you sure are getting in those kills per death.

 

Hah yea, that also comes with a whole bunch of hate mail.. I exclusively played Conquest and PTFO'd

So CES is getting ever so closer, and now that we know RTX $uper MXM are coming do you think we'll see them there paired with the x170?

"CLEVO @ CES Las Vegas 20202019/11/11

Dear Customers, A complete range of innovative products from entry value to high performance. Please visit CLEVO to find the best solutions to serve your business needs. Sincerely looking forward to meeting you at Mandalay Bay Hotel from 7th to 8th, January, 2020.
For further inquiries, please contact us via e-mail..."

 

Exciting times with the possibly highest performance mobile graphics solution in the form of the Quadro RTX 6000 coming into a 15" ASUS turdbook:
https://www.asus.com/Laptops/ProArt-StudioBook-One-W590G6T/Tech-Specs/

I hope that will be a sign of better things to come in the MXM form factor, too, like a 2080TI for example. The real challenge would be a chassis that can properly cool it but then Nvidia does not seem to be concerned with that any more seeing where they allow the Quadro RTX 6000 to be put in.

 

The leaked RTX 2080 Super MXM specs

RTX2080Ti: 4352
GTX1080Ti: 3584
RTX2080S: 3072
RTX2080: 2944
GTX1080: 2560

It's still nowhere near the 2080Ti performance.

 

Maybe it is just me but it seems that with everything else being the same or scaled accordingly the RTX 2080 Super is not very Super at all

 

It's very 'meh' indeed. And even more pointless of an upgrade if you already have the RTX 2080.

 

That was the case with the desktop super cards too honestly. The 2060 and 2070 supers were a modest bump up and the 2080 was honestly worthless. So not surprised it only has a fancy name without any performance to back it up.

 

The desktop 2080 super was not that exciting either, the core does not have much more in it.

 

Put it this way. Put 2080 and then 2080 Super in desktops and run them with exactly same TDP. Nvidia bumped up max allowed power consumption for Desktop 2080 Super. With notebooks you’re screwed. Forget a decent increase in performance with Nvidia’s newest Mobile Scam. What’s the point with a small bump in cores and memory if everything else is capped to the bone? Only much more cores with the Ti chips would make it better.

The sales point with mobile Super is to provide slightly better RTX performance. But still worthless for Mobile cards. This even with the better desktop cards.

 

More like a 2080 'Meh' MXM to me

 

Binning and tweaked power tables probably could push clocks up, another 10-20w would help too.

 

From 215w up to 250W. In our testing, however, the RTX 2080 Super is only about 4-5% faster than the RTX 2080 Founders Edition, even in synthetic benchmarks like 3DMark TimeSpy Extreme. And the Mobile Super card will most likely get same crippled TDP as 2080 https://www.techradar.com/news/nvidia-geforce-rtx-2080-super-vs-rtx-2080-what-has-changed

The Clevo card will finally get 200W. Aka won't be the most crippled anymore vs. some other brands models.

The desktop cards will get another summer with new, while the mobile is re-hashed old one.

Nothing to be happy about.

 

It'll be interesting how they manage the thermal headroom in the x170, when Intel's 10-series is going to run hotter than the crispy 9900K. That unified heatsink is for dumping the molten cores of the 10-series

125W TDP for the 10 core 10900K - that's the 3.7Ghz base TDP
https://videocardz.com/newz/leaked-slide-confirms-intel-10th-gen-core-k-series-are-125w-tdp

And they appear to be capable of going north of 5Ghz
https://videocardz.com/newz/intel-c...ications-of-10th-gen-core-comet-lake-s-leaked

 

Has it been confirmed that the Clevo 2020 will support MXM graphics cards?
In any case we should have confirmation about what it can or cannot do very soon.

 

So you think the 2080 Super MXM will be 200TDP? This wouldn’t be such a bad option then.

This is also kinda like 10 series GTX I suppose. There was (3) intel generation CPU’s in laptops on the same GTX10XX series video cards in mobile haha.

 

Saved by the Unified heatsink. But in Cpu only load. The Cpu side of the heatsink is intended to remove 140W heat.

Yeah, Clevo was forced to go back to 200w. Not fun be the ODM who offer the most crippled cards. I mean remember MSI went up to to 180 or 200w after short time. Clevo stuck with 150w @Prema

https://twitter.com/PremaMod/status/1097220895834161152

 

Hotter but no more thermally dense.

 

A unified heatsink is nice when only the CPU or the GPU is under heavy load so it will be an advantage for most DTR users and it could even be an advantage in some games where CPU load is lower.

So buying a P775TM now it should have a 200W TDP for the RTX 2080?
And yeah, not much fun to only have 150W, but also not that much fun with a 9900K and the RTX 2080 drawing well above 300W in a P775TM chassis that is designed for a TDP of 250W!