Clevo Overclocker's Lounge

Then try the insider version of RS4 which is pretty much better than RTM aside from UI issues such as Checkboxes are mysteriously disappeared in most cases. New UWP Aero or Fluent Design UI is complete and consistent.

 

@j95 has suggested I try RS4 as well. I might, but I have a personal hangup with being one of their guinea pig contributors. The thought of doing something that might help W10 become more successful in its current state does not set well with me. If they did away with idiotic things like the forced updates, UWP filth, diminished end user control of the OS, and making silly changes for no legitimate reason other than to say it is different than it used to be, and acts/feels like a smartphone, (as if that is somehow a redeeming quality,) I might not have as much contempt for them as I do. As long as it looks, feels, acts and smells like it does today, I do not want W10 to be a successful product. It deserves to fail based on those things.

 

It looks more like my phone in RS4. I tried it in a VM and used NTLite to remove some unneeded UWP apps like 3d viewer,Onenote,Office Hub, Xbox Game Bar,Xbox and kept only Zune Music and Video. Removing some UWP apps will break some restriction in uninstallation of UWP apps so I can freely remove some apps using CCleaner/Dism++/BCUninstaller.
Windows Spatial Audio has seen some hefty improvements. By default turning ON Windows Spatial audio defaults to 16 bit 192kHz instead of 16 bit 44kHz. Audio quality is great on Groove Music or any UWP apps. Felt almost I had Linux. On Linux hearing 32bit audio is amazing and very immersive. I bet SB recon or Nahimic can take full advantage of immersive effects in RS4.
New entry is the ability to check Startup apps in Settings, Downscale video quality for better battery. Only downside is, it needs huge IO speed/throughput for better response time; so SSD is recommended. That means my old laptop and my sis's must be upgraded to SSDs.
For now I tried only Home edition. Maybe in a few days I'll try Pro and Enterprise editions. The insider retail ISOs contain Home,Pro and Enterprise editions.
EDIT: I'm staying on RS2 until EOL.

 

How do you guys like the logo I made for the Prema Mod 'Team' aspect of the project (left side of my signature)?

 

Make mine blue and i'm in.

 

I like it, purple's pretty great for me

 

It's very you

 

Looks very nice. Great job.

 

Matched it with no trouble.

 

nice! Look like an eye or a magic mushroom

 

Hey @Mr. Fox I think you'll enjoy this in particular

https://clips.twitch.tv/PlainPolishedLasagnaFreakinStinkin

 

LOL, yeah that was excellent. Thanks for sharing it.

I love Doom.

 

Here's this... http://hwbot.org/submission/3756295_

(Spent a little extra time with trial and error finding out how many of the additional worthless services could be killed off in the Windoze X ass-ware OS and snagged a few extra points.)

 

https://clips.twitch.tv/AntediluvianAbstrusePuddingSSSsss here, enjoy this one too. I think my response is completely fitting for what I found, don't you?

 

Absolutely appropriate with that chain gun.

 

And, Windows 7... http://hwbot.org/submission/3756315_

 

Nah. Oh hell to the Nah!!!!!!!!!!!!!!!!! Just say no!!!!!!!!!!!!!!!

How you gonna get the bomb weapon and just pan out after that??????????? Just say NO! You are just wrong! I was looking to see it's damage and you just killed that off.

 

@Johnksss I was able to overclock the ram to 2933, by doing 133/11 instead of 100/15 and now it seems stable. It would be nice though to reach 3k...

 

Haha the clips can only be 60 seconds unfortunately

Sent from my OnePlus 1 using a coconut

 

@Johnksss @Mr. Fox try a fresh server 2008r2, 2012 or 2016 install. they come with the least amount of garbage. of course just like win7 > 8 > 10 in benchmark score, 2008r2 is the best among them.

and apply registry tweak to turn off window core park to reach true cstate off.

 

How about you go ahead and run the benchmarks so we can see them. You already have the os installed.

Sent from my SM-G955U using Tapatalk

 

my OS is full of garbage of close to 50+ software and a bunch of services. I can do a fresh CB R15 with nothing else except no throttling when i get off but its on a fresh 2012 with no hwinfo and only TS for settings, can prob do 5ghz anything higher i dont have ways to cool it, maybe you can do a 5ghz to compare after. dont have other OS installed.

edit: on 2nd thought, cb prob dont differ much. where can i get 3dmark or other software that shows win7 > other OS so i can use them to bench it.

 

regarding meltdown and spectre

 

Not sure what that had to do with you running the benchmark? "Scratching head" It's not for you to compete it's just to see what it can do. I mean, have you actually beat any of us before with a tuned system? I mean I did go check and all, but couldn't really find a case for that.

 

8700K physics score over 20K...

3DMark 11 - https://www.3dmark.com/3dm11/12595237 | http://hwbot.org/submission/3757209_

 

honestly if we were to bench to see if fresh server OS would be better or worse, it would not be about high end overclocking score per se. it'll be about matching similar conditions in our system. 1 bench software running, same services running, but different OS like 2008r2 vs win7 (enterprise vs consumer). also would not be using AC since thats temp bound, we'd be doing more of a 4-4.5ghz overclock with zero temp issues, and finally compare the score as they SHOULD be similar but thats what we're trying to find out.

is 3dmark only cpu involved? i only got a 1070 until volta mobile comes along, assuming it works plugnplay

 

No, but 3DMark 11 and Vantage are the most demanding in the CPU/Physics tests of any Futuremark benchmark. They carry more weight than they do with the overall score than the fluffy "feel good" GPU benchmarks like Fire Strike do. 3DMark Sky Diver also gives extra credit for an awesome Physics score. And, HWBOT has a special CPU/Physics-only category for 3DMark 11.

3DMark 11 Physics Test Ranking - http://hwbot.org/submission/3757243_

 

no sorry, what i meant to say is, do you notice a difference between 7 and 10 running those software? i dont have 10 so i simply assumed that 7 and 10 would have similar CB scores at same frequency with similar service, but that could be wrong. if CB ends similar theres no point using CB as test to try and verify OS differences, however for other software i do not know never used them. thats why im asking you and john

cb is a pretty well optimized software, for legacy extensions

 

@ole!!!
Hummm, ok. nevermind. I'll do it my self when i decide to get around to it.

 

lol. if say i do a 4ghz and you do a 4ghz for server os vs consumer os, one has higher score, then answers out. 5 or 5.1/5.2 ghz would mean the same would it not

 

I don't want you to take this the wrong way, but um..... All I asked was for you to run the bench. You could have just said no and I can respect that. Nothing wrong with that at all. Was not really looking for 50 reasons on why not to run it.

Thanks though.

 

kk i'll run it brb

 

Yes, absolutely. Windows 7 scores higher in 3DMark 11, Vantage and Sky Diver, sometimes Cinebench and wPrime because Windows 10 (and Windows 8.X) hinder CPU physics performance. Windows 8.X and 10 have a slight edge in graphics peformance on some benchmarks (like Fire Strike), but definitely nowhere close to all of them. Overall, Windows 7 consistently outperforms Windows 8.X and 10. But, that's why I dual-boot both OSes. I use the OS that returns the highest scores, just like I use a hammer to drive nails and a screwdriver to drive screws. All about using the right tool for the job.

Windows 10 is also severely bloated with a crap ton of garbage services that are worthless and unnecessary unless you enjoy the trash payload that goes along with all of that filth.

 

People should take a look into Black Viper’s Windows 10 Service Configurations (Updated for Fall Abortion).

 

Definitely agree. Killing unnecessary services make a measurable difference in benchmarks. I also disable WiFi and audio service when benching. They are not unnecessary for normal use, but they do not do you any favors for benching the CPU. They consume clock cycles and slow down the CPU a little bit.

http://hwbot.org/submission/3757262_

 

Not only for benching. Not all the services need to be running 24/7 to have a functional OS.

 

C'mon bros... @Prema HWBOT Team only needs a few more points to overtake the EVGA Enthusiasts Team... post those benches. Every point counts!


@Trafficante @Mobius 1 - add your newest and best scores and it bet it will do the trick.

 

Just did an install and ran it. No tweaking.

 

@Johnksss @Mr. Fox so a short server 2012 bench turned into overlocking game and trying to find the best settings.. most services are disabled except a few.

at 1646 im 5ghz a tiny bit of throttling, for 1648 and 1650 is 5.1ghz so its hardcore throttling at that point. can't go over 1.3v so had to lower temp as much as possible so chip wouldnt use too much voltage and wouldnt crash with too little voltage. ultimately temperature is my limit unless i use AC, i even took off the bottom cover and had coolingpad blowing air directly too.

only 2400mhz ram tho, if i do get 3200 ram working in this that'll be a decent jump as well, just ram bandwidth alone is 33% increase going from 2400 to 3200.

but honestly i dont see the point of this because we were talking about server OS vs consumer OS.. john doesn't have the same setup as i do so how would we compare.

 

Is this a P870TM1? With a delidded CPU? Your signature only lists the M18x R2.

Also, I cannot be completely sure but I think higher speed RAM, in reducing CPU load, causes the CPU to work harder (hotter) for the same utilization %. So... if 2400MHz cl17 at 5GHz 100% gives you a score of 200 in benchmark X, and 3200MHz cl16 at 5GHz 100% gives you a score of 210 in benchmark X, your CPU will also run hotter/consume more power during that test. This is how I understand it but not currently testable by me. I remember a CPU temperature jump when I upgraded RAM in my last notebook from 1600 11-11-11-27 to 2133 11-11-11-31, though.

 

Yes, 100% load in sensors really does not mean much, hence why a power virus type application can consume way more than something else at "100%".

 

@Prema intervened So now 3000Mhz ram is a walk in the park! Was even able to tighten the timings as well!
Prema is blessed with a powerful talent to fix our notebooks, no doubt about that!

 

That's right, 100% includes "busy/waiting" time. It's exactly why Hyperthreading and Simultaneous Multi-Threading work in the first place. It simply feeds more data to one core, making it work harder. We almost never see true "100%" loads on anything.

 

One did with 2400 mhz ram and the other at 3200 mhz. No tweaking. No AC. Just sitting on the table with max fans. Just jumped up and ran it before i'm leaving for CES.



The purpose was for me to see how it stacked against windows 10/7. Nothing more.

 

So is 5.2Ghz the max you guys can reach (It's incredible, I'm just curious)? What's the limiting factor here, power or thermals?

Side note, does the IGP even get used from the CPU? Or could you go about setting the clocks to 200Mhz and -180mv if it's solely using the dGPU. I attempted this on my 9560 and was able to mitigate 2-3w on the chip TDP. Different systems I know, but I'm just curious as I'm thinking about pulling the trigger on one of these. Spending $~7,000 on a laptop is not something I do on a whim haha.

 

Maybe on a good day you can run a few things in the 5.2 ghz range, but for almost everything else. Nope. Not yet at least. 5.0 or so depending on chip and setup.

 

i can confirm this, with avx load faster ram gets you more heat quicker to notice, also harder on cpu/ram. I will need help in cooling either way to go any higher, with either faster ram or 5.1/5.2 at lower voltage. it is 870tm and too lazy to change up sig.

like john said you'll need a good chip from silicon lottery or from whoever you're purchasing from. i got the ram/cpu/SSDs and laptop barebone separately for around 3700 USD, you just have to pick up and find the parts and trust that it'll work, only a 1070 though.

 

CES, picture plix!

how you get yours to 165w without throttling? i can't seem to go over that 1.3v limit and it'll always be around 145 to 150w for me, and you did them without throttling.

also, how did it compare versus consumer windows

 

Here's this with the latest @j95 mod... last GeFarts driver before the cancer fix for the security issues.

Time Spy: https://www.3dmark.com/3dm/24520450# | http://hwbot.org/submission/3757694_