*Official* NBR Desktop Overclocker's Lounge [laptop owners welcome, too]

Ha, bloody hell John, shouldn't you be at 5.2Ghz minimum for your 9900K given your normal pushing everything to the limit?!

 

No. Why? I only need to run just enough to beat whoever thought they were on top.

 

It was a joke, but also a question! I don't know who's on top or what you mean by that. I'm sure I've seen many 9900K's at least at 5Ghz, so yours at 4.9Ghz doesn't seem up to your normal character. It was a joke, but also a question.

 

difference being: john's score at 4.9 ghz still beats the 5.0 and most 5.1 scores from regular users

Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk

 

Ah, I wonder how? Unless people's CPU's are power or thermal throttling, or maybe an ultra clean install with no bloat running and reduced processes, setting CB20 to run at high process priority maybe. He can be pretty cryptic/cagey sometimes, and I don't think he got my joke about his usual character and only running it at 4.9Ghz!

 

yep thats john alright @Cryptic/cagey, its what makes him so loveable and fun to be around here at NBR haha no idea, some users like john just bent any random hardware off the shelf to their will while us regular folks have to pay premium dollars to get our hands on a highly binned part to get halfway decent scores haha

Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk

 

Cinebench R20 fun run with my 8086K @ 5.4GHz ...



CPU (Single Core) run...

 

Ha, ok!

 

Nice, but you should really be running that at 4.9Ghz! ;-)

 

50x is my daily for almost everything, but 53x has been rock solid for gaming and encoding, while 54x has been very stable so far. However, 54x isn't necessary nor practical for daily use so I just keep it at 50x for the most part.

 

No LOL or Green smiley face, so it looks more like something else. Who's to say if it was ran at 5.2 Ghz that it becomes, "Why you didn't run at 5.3 Ghz?. I seen you do benchmarks at that before." This is actually usually the route this takes....

Also, this wasn't a post to show me beating anyone, it was just to represent the laptop community. And I generally only do enough to get past the top person then go back to doing other things.

This is totally true, but no one really wins going up against regular users. Those users do not really care. And people using them as examples to show one's superiority is pretty cheesy as well. It only counts when you go up against the people who matter, because those people care. And will take the time to get the machine to do what it was meant to do.

This is a big part of the problem, but when you try to "mention" it, you get the feedback of, "No, not my machine. It's running "smooth as butter" so one sits back and waits till a level headed person gets access to the same machine and presto. This user gets a much much higher score. Why? Because they dropped the cocky attitude and just got down to business. Now the first user wants to start complaining about his machine is sub par and blah blah blah, because his skill level is not on par with the others. This happens way to many times than not. That is actually the part that drives me up a wall the most. Bottom line, you can not win them all and complainers usually never win because their too busy looking for excuses... Plenty of times I didn't have the best hardware and I knew it. I had to suck it up and accept that or find better hardware, but sitting in the forums complaining about it does not make the hardware get any better. (I know your going to assume this is random, but it's not)

And as to running benchmarks, you pretty much summed it up as you have time and time again, so why would running CR20 be different? It's the same exact thing for most cpu intensive benchmarks, but this is only for the "Aftermarket Factory Runs" and not the "Out of the Box Factory" runs. No one ever wins running what the "Out of Box Factory" runs generate. They are generally 99% wrong.

And the answer to this question is......People want everything handed to them nowadays. And when it is and they find out what the people handing it to them already knew.....It doesn't always work. So now it becomes there system this, hardware that, software this and that, but it never ever seems to be the user themselves that is the problem. Or the flip side.(Might be a bit hard to understand, but it has meaning) All of a sudden they are now masters because they took the help of another user and added some fake twist to seem like they made the advice better than what it already was. That one happens allot with new overclockers that beat someone not trying to beat anyone, but only showing some possibilities and a general area what the clock speed should produce under normal circumstances.

And you wonder how I have come to this? Because I use to do the exact same thing (2009) in the very beginning. And someone told me, if you want to advance, you need to understand the whole setup and it's pre limits first... then move forward.

As to the regular user paying premium dollar. That is not always true unfortunately... I'm positive you can recall a few times you spent top dollar and it didn't pan out so you bought the cheaper version and it worked flawless. Sometimes it's just the luck of the draw.

 

Very wise man.

 

Yes he is. @Johnksss , @Mr. Fox and you @Rage Set taught me almost everything I know about overclocking. 75 years old I am and never too old to learn.

 

Yeah, I think I get where you're coming from - people shouldn't stand on other people's shoulders, they should work it out for themselves when it comes to competetive overclocking, so that's why you like to be a bit enigmatic about what you've done. I'm definitely not into competetive overclocking, so you've got a different mindset to it.

 

The best overclockers are. They may share some things but they keep their best to themselves, as they should. At the end of the day it is a competition, friendly, but competitive nonetheless. One thing I've learned in this sport, less is more.

 

word and amen to that @Johnksss

especially the users who want handouts for everything nowadays. one button push solutions for everything, especially when it comes to overclocking they miss the point that the actual fun (at least for me) comes from the tuning itself. the rocky road to stability makes a completed benchmark / stress test all the more rewarding and makes you get to know your system much better.

Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk

 

Not really. I just get tired of explaining the exact same thing over and over for the same exact benchmarks. And half of these things really are not rocket science projects where they need to be over analyzed, dissected and then re analyzed to come to an overly technical explanation for something simple as . "I just set the multiplier and ran the test" This seems to be unfathomable and needs further investigations.... and so on a so forth.

Sometimes, it really is that simple...

 

Well said. I was about to reply to all of this, but you've covered it.

About the hand outs, I completely agree. Too many people (usually kids on the internet) just want everything on a plate without any desire to actually learn. Nope. Not happening and I completely understand where @Johnksss is coming from. This whole "insta(nt)," mentality these days SMH...

The journey itself is indeed a big part of appreciating and learning the system.

(To add) As a reminder, let's also not forget that there are different classes when it comes to benching. Just because one has a high score with a 9900K, doesn't mean that it's all if a sudden ammo to use against someone in a completely different class with let's say a 7700K. That's childish anyway.

There's an art to benching and it is a legitimate sport and hobby, much like and no different to someone playing games all day.

The difference here is that the OC community has a much more comprehensive and respectable culture due to the nature of its accomplishments.



Sent from my SM-N950U using Tapatalk

 

I have done absolutely zero tweaking on my system yet> Mild overclock using afterburner. Smokes my old 2070

 

Sometimes! :-D

 

https://hwbot.org/submission/409555...5_core_i7_8700k_19.01_points?recalculate=true


https://hwbot.org/submission/4093513_papusan_cinebench___r15_core_i7_8700k_1735_cb?recalculate=true

 

This may be the last cold days of the winter season. I have to make use of it before that warm weather begins...

 

Yeah, if it wasn't so heavy, I would have moved my rig to the garage when outside temps were around 8-16F all day long for two days. That would have made for some nice benching... (Too lazy)

Sent from my SM-G900P using Tapatalk

 

Here are some preliminary runs...

Note:
These are actual "stable" settings that I game with and are not just a one time bench setting. I'll get to benching soon, but this is just my 8086K/2080Ti gaming desktop.

Fire Strike:
> Graphics: 40342
As for Physics, I have runs at 23400 .

https://www.3dmark.com/fs/18643361

Time Spy:
> Graphics: 16543
Broke 15K! which was my goal for giggles at its normal desktop daily driver form.
Once I get my 9900K on the test bench with this 2080Ti, it should produce some nice numbers.

https://www.3dmark.com/spy/6519509

Cinebench R15: (1787cb)

 

Nice, your Firestrike GPU score is nearly twice my GTX 1070 score, and your Timespy is over twice as fast. You got 50-60% more on your CPU Physics scores than my 6700K too. What kind of gaming do you do, and what monitor (res/Hz) have you got?

 

Thanks. I use an AW 34" wide screen curved monitor @ 120Hz.

As for your scores, remember that the scores are all in respect to its class. The GTX 1070 is a fine card... I like it a lot and it OC's pretty well from the ones I've had.

 

Yeah, I'm fine with my system, I'll update it with a new GPU, 7nm NVidia GPU in about yr 2020 I expect - I may update rest of system, but don't really want to unless I know it's limiting framerate.

Is that a 3440 x 1440 pixels? Ha, yeah, you need that 2080ti for that resolution & 120fps! Nice choice of monitor, immersive due to size, and 120Hz for the smoothness. If I had 2080ti would be a good chance I'd have a 240Hz 1080p monitor, but I'd probably go to a PC shop & try out one like yours to check the immersive qualities and higher resolution first though to see what I'm missing. As it stands though, they'll be no monitor buying from me until after my next GPU I'm yet to own has expired.

 

Ryzen 2700x CB 20

 

Alright, back to a custom loop. Watched my CPU hit 100c @4400mhz 1.45v before it locked up running CB20. Not sure how accurate aida 64 is with Ryzen temps, but it was enough for me to order another one up. This time not getting the big bottle.

http://www.performance-pcs.com/xspc-raystorm-pro-x4-photon-ax360-watercooling-kit-amd-am4.html

 

New CB top for me

 

hey guys, updated my Cinebench command line syntax thread to include R15 Extreme and R20

Link: http://forum.notebookreview.com/thr...-line-syntax-bat-loop-detailed-output.815101/

enjoy

 

yesssss nice! thx for the update!

may i ask for some pointers? VDIMM, VCCSA, VCCIO? also, any readjustment of cache necessary or did it play a role in there at all?

and did you have to lower the cpu multi to reach 3700?

Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk

 

Voltage was like 1.3V, but it took a while to boot.
Didn't mess with anything else. I just let it try to work itself out first.
Not sure what it would do with the 9900K, I haven't tried yet.

 

cool sounds good yeah still trying to figure out how the 9900K imc compares to the 8700K. its definitely better than the 7700k though, i was able to dial in 3000 mhz far more quickly, got it tm5 error-free in 2-3 tries

how long did u let it train before it finally booted?

Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk

 

Finally home again. It was a long week with only my old ZBook. Really missed my desktop.

 

Okay, wtf! Windows 10 on the left, and Windows 7 on the right?? First time benching anything on Windows 10. I thought it was known that 10 ran firestrike better?

 

I used to get higher Firestrike scores on Windows 7 till I put the latest version(1809) and latest build on Windows 10. Now I get much better scores in 10.

 

Im using 1809 enterprise edition. Wonder if I should update it. I have Windows update disabled though.

 

Combined score is what tanked W10. Try killing all of the worthless services running in the background. 1809 has more background bloatware filth running than prior versions. Also disable Spectre and Meltdown mitigation garbage.

How hot did the CPU get in the Combined Test?

 

That looks like a pretty healthy overclock on the RTX 2080! +200Mhz on the core and nearly +1000Mhz (at least on the MSI Afterburner slider) on the VRAM. Where's that stand on RTX 2080 rankings? With chilled water?

 

Actually.....That's like 1400 on vram.

Not sure, but the A51M is up to 30.2k@200W TDP

 

1400 on VRAM? You're multiplying the actual frequency increase by 4 then (which is the biggest way of describing the VRAM offset). I thought MSI Afterburner slider operated in "x2 mode" when it comes to VRAM frequency? EDIT: yes, I just checked, MSI afterburner slider operates in "x2 mode" when it comes to VRAM slider - so that's the value I was referring to. You're referring to the "x4 mode" of describing VRAM frequency, in which case the +1400 on the VRAM would stand correct in that situation.

Ahhhh, it's a laptop! Wondered why you were mentioning the A51M. Well that's darn good for a laptop.

 

It shows the correct speed in the screen shot. 2113Mhz x 4 = 8452 Mhz

 

Indeed.

 

Also to note. That's 150W TDP
And 1400+ on MSI afterburner = 2133 Mhz
1500+ = 8500 Mhz

 

MSI Afterburner slider has changed then with Turing in contrast to Pascal then? Because with Pascal it's definitely working on the "x2 method" when it comes to the slider value for VRAM. You're saying with your Turing card that the slider is operating on "x4 method"? I don't think that's the case.

EDIT: don't be nebulous, say what you mean, I think we communicate in a different language!

EDIT #2: answered my own question, like I expected the VRAM slider in MSI Afterburner operates on "x2 method" (just the same as with Pascal) as can be seen here: https://www.guru3d.com/articles_pages/geforce_rtx_2080_founders_review,37.html
I don't know specifically to what you're referring to in that case then, but it doesn't matter, we have a history of miscommunication.

 

Hey guys, just sharing some new PR's with the current set up (8086K) before I install the 9900K this coming week into my daily driver gaming desktop.

This 8086K is rock solid at 5.4GHz, but I haven't had much time to tune 55x yet...

My goal was to see what the desktop could do in its daily driver form and to see just how far I could push it with it being "Stable."

I'll be transferring this 2080Ti to the test bench for sure as it is a strong performer.

New PR for Cinebench R20: 4364cb


New PR for Cinebench R15: 1799cb


---

Time Spy: 15196 (New PR w/ 8086K+2080Ti. Stable game settings.)
Graphics: 16755
CPU: 9950


https://www.3dmark.com/spy/6550393

@Mr. Fox , welcome back from your travels. I know how it feels being away from the desktop for too long.

 

https://hwbot.org/submission/4098117_
Strange that I get a higher CB15 score in win 10.