Clevo + Coffee Lake: Status?

I still don't understand this niche market. It is not like you are taking these beasts to Starbucks. I am doing a mini-ITX rig next time, kind of regret my laptop, its great and all, but yeah. and don't create lies, I know 90% of you in this thread sittin in your bedroom or office, and that laptop moves maybe twice a year. lulz

 

the purpose is to have less wires, battery back up all in small form factor, easy to carry in a backpack. all of these isnt something mini ITX can offer.

 

to carry where? to class? better off with a cheap netbook for note taking. less risk to your expensive gaming machine. less cables? get a bluetooth keyboard and mouse, and one power cable and 1 hdmi 2.0 cable. your talking one more cable, thats it. well two if you count that monitor power cable.

 

its one of those things when asked, you dont have the answer to it until situation comes up, i could probably think of a few but there are more to list. similar to asking you to list all the words you know starts with A, you can't list all the ones you know.

but to answer rest of your remark:
- display connector
- no backup battery
- power cord - computer
- power cord - display
- bulk UPS for battery
- poorer wifi performance
- hard to carry

some people would be okay with having a netbook + primary laptop/ desktop. for some its a hassle having to switch between the two, if files needs to be transfered, wanted to do something thats not capable of doing on netbook etc.

not saying theres no pros with mini ITX computer, for ones not willing to sacrifice for the above would need to get a laptop.

 

I put it here if you haven’t seen it. Trust on reviewers is... Yeah, you know.

 

I wasn't expecting you on this forum, dangit you caught me slandering my mini-itx propaganda! I better lay low for awhile

 

Getting a laptop saved me a lot of money, the amount I was spending on different CPUs and motherboards for a time on desktops was kind of crazy

 

desktop is always cheaper isnt it? assuming very similar hardware specs

 

Not if you buy several motherboards and graphics cards over the course of a year because they are easy slot in upgrades

 

LOL no self control

 

I've taken my P870DM3, P570WM, M18x to starbucks, to school (with the 780W Eurocom ac adapter) and I always took them to work every Fri/Sat/Sun sooooooooo yeah..

When I get my next monster I will continue to do so. It's a LAPTOP for a reason, still don't understand why people don't get that.

I don't buy a laptop to have it sit on a desk connected to everything most of its life, I'd rather spend my money on a desktop at that point.

I use an Alienware Orion M17x backpack to carry it all WITH my books for school. 32lbs total last time I measured my backpack with my P870DM3 with books and AC Adapter.

 

Now I had to delete a bunch of posts in here. Argue the facts, and preferably one at a time so you aren't running circles around each other.

Charles

 

continue off the story. since high end BGA and LGA cost similar, LGA has much more benefits when comes to repair, repair cost, performance. what is the point paying similar for 15-20% lower in terms of performance. @Danishblunt

 

Both CPU's have their own benefits. LGA has more performance and your point about repairability is a fair point, however I've never heard of a case where a CPU was broken. The only point you could argue for are those where the mainboard has also a soldered GPU on it, since it seems boards with a BGA CPU or LGA boards are pretty much the same price if no BGA GPU is involved.

BGA chips require less cooling, hence a thinner case and less noise.

The main issue I see with the desktop CPU's in notebook are 2 things:
- Insufficient cooling
- Require a very thick and beefy case for cooling.

Since it was already hard enough to run the 6700k / 7700K on mid 4ghz I simply don't know if the coffee lake is actually able to maintain cool in a clevo notebook. That's my only concern.
For instance:


This is mr.fox gameplay on 7700K OC @ 4.7ghz. You can see his temperatures hover around 87-90, which in itself is already high, when you note that his CPU usage is around 65%ish CPU usage and throttling down to 4.2ghz most of the time.
This is what mr.fox did to his notebook:

Meanwhile this is a desktop PC:

5ghz throughout the whole gameplay mid 70s temp on average.

Now tell me, how on earth is a clevo notebook with apparently the same case and a hopefully modded cooling system going to handle a stock 8700K let alone a OC'ed 8700K while people like mr.fox even struggle to run a 7700K with 2 less cores on mid 4ghz? To top it all off, people even struggling on desktop to cool down the 8700K beast, once OCed a very beefy watercooled setup is needed. You expect me to believe that 5ghz on a 8700K will be a thing when using it for a longer period of time?

I'd rather have less performance than a machine that throttles while gaming and causing massive heat even after modding the heatsink heavily.

The thing is, you should stop getting fooled by mr.fox and others, just because you see that the CPU's can run at 5.2ghz on desktop or for a short benchmark, doesn't mean it's actually usable. Here is a quote from a smart person from overclock.net after seeing mr.fox 5.2ghz screenshot.

Be smart, question what you're seeing, do research. Then you don't get fooled by people who try to sell you sand on a beach.

Here is the screenshot:
https://imgur.com/2yPL4nN

Normal use, meaning browsing the web etc. 79c package. Meanwhile others play 5ghz crysis on lower temps.

Go ahead, tell me how 8700K will run at 5ghz on a notebook.

 

sorry man i stopped reading the rest of it once i got to this sentence. since it seems rest of your comment is derived from that statement I outta stop there and clear things up as thats where we start to disagree.

tbh most BGA chips are clocked lower due to intel purposely gave them inferior silicon knowing that if its clocked higher they'll run super hot, hence they would require *less cooling per se due to running slower.

if you were to clock them vs desktop at similar frequency between say 3.5 to 4.5ghz (a normal usage frequency), BGA will use much more power and requires more cooling. ie 7820HK at 4.2ghz and 7700k at 4.2ghz, LGA becomes the winner in terms of power consumption.

as for your statement about clevo, let me also clear something up for you. I am not fooled by anyone, and one of the reason i did not purchase P870DM laptop is because i hated its cooling design just by looking at it i know its got issues for me. Mr.fox and few others are hardcore enthusiast on not just CPU but also GPU, on the other hand i only really care about CPU and storage. in clevo's laptop originally its 1 fan per GPU rated at 100w for 980m. once it get to 180w-200w for a 980 and then 1080 its obvious that its not enough. clevo decided to direct GPU heat towards CPU heatsink so of course both GPU/CPU will run hot, its inevitable.

 

If you don't bother reading an entire post but only a sentence, then don't bother replying at all. thanks.

The rest of the post is actually explaining as to why desktop CPU's in notebooks are a problem, irconicly even showcased by mr.fox.

My point is being, desktop CPU's in clevo notebooks run way to hot and the performance difference u'd get from a BGA vs a LGA is not worth having a furnace of a notebook, not to mention that it won't run stable. If you're going to undervolt and underclock the desktop CPU to BGA levels, then why bother getting a desktop CPU for the extra money in the first place?

We can both agree that LGA notebooks are better, but are they worth the price, that's the other issue. In general a stock desktop CPU will run hotter than their BGA counterpart, you could tinker around with it but u'll only end up having BGA performance on a Desktop CPU with acceptable temps. As showcased, even with liquid metal, delid, bottom case of, fans blowing in the case for airflow, mr.fox reached late 80s to 90s temp on 4.2ghz with his 7700k on a game like crysis 3, now imagine something like rendering 4k or even 8k footage, so why on earth would I pay the extra premium for an LGA notebook?

Also the most interesting part, which you don't even talk about is coffee lake, if you see the struggle with kaby lake, how is a coffee lake going to run on a notebook? that's my main concern here and what I actually wanted to talk about. I don't see a kabylake running anywhere neat acceptable, do you?

I know it's yet to showcase if it does ot not, and I'm sure people here are excited about it, but my question is:
"Will coffee lake be worth it on a notebook?"
Looking at the previous kaby and skylake I'm downright confused as to how clevo will try to tame this beast of a CPU on a notebook.

 

This is from your posting Need OC help

Post #4


post #8


I don't think you should talk abot laptops you don't know about or own by yourself. Almost Every post you put up in the forum is full of wrong info. Why bother with it?

 

@Danishblunt
Wrong. Not all Clevo heatsinks are the same. Also it's all about prep work. I think this is video you have put up is one of Mr Foxs old videos before he discovered the cpu heatsink gap (correct if i'm wrong). This video here (my video that I have linked) the P870KM1-G 7700K CPU reaches 68c tops at 1440p. When I get home I will do the same settings with the same bench method but this time with fans on full including vardar mod. I bet I get 58c CPU max.
It's all about prep work.

 

Interesting developments in the 8700k reviews with reviewers caught accidentally OC'ing the Coffee Lake CPU's at stock because the motherboard BIOS defaults enabled Multi-Core OC + All Core Sync, running all the cores at 4.7ghz instead of stock boost.

Dropping the boost clocks down to stock changes the results dramatically, and Ryzen 7 1700x jumps ahead of the 8700K, with Ryzen 5 1600x much closer than original benchmarked.

This appears to be done to pump up the "stock" scores of the 8700k because the "real stock" 8700k scores are below the Ryzen 7 1700x without the OC boost.

AdoredTV video explaining the issue
http://forum.notebookreview.com/thr...coffee-lake-z370.809268/page-17#post-10614718

Jayz2cents explaining the mistake, and retesting the 8700k at "real stock" boost speeds.
http://forum.notebookreview.com/thr...coffee-lake-z370.809268/page-17#post-10614749

 

When doing the benchmarks please also include the speeds as well into the afterburner, other than that interesting vid, keep us updated.

 

Intel Will no Longer Disclose Multi-Core Turbo Boost Frequencies
"Intel from now on will only list the base and (1) and Single thread (3) turbo"

The main point is still what bench scores you will get with all 6 cores loaded.... 4.3-4.4-4.5-4.6-4.7-4.8-4.9-5.0-5.1-5.2-5.3-5.4-5.5 and 5.6GHz on all 6 cores. It's an unlocked cpu!! This is what's count.

 

As the AdoredTV video points out, this is a motherboard setting that controls MCE, it's not controlled by the CPU.

What counts is knowing what your motherboard is supposed to be doing is what it is actually doing, and they aren't.

Thinking you are running stock and stock voltages when you are running 400mhz higher on all 4 cores with higher voltages and thermal output isn't good.

Watch the video by Adored TV, and Jay's 3 videos, for the details and wait to comment until you've seen all 3.

I'm expecting retests to come through from many reviewers.

And, there is still a discrepancy between Asus and other motherboards 8700k scores, with Jay getting 1400cb now with others getting 1200cb.

I'm assuming there are other "boost's" enabled on the Asus motherboard that aren't on the other motherboards.

The 8700k review scores on Asus motherboards at stock are all OC boosted baloney at this point, until this all shakes out.

At least Jay's retest after disabling MCE is a move in the right direction:

To compound matters, Jay figured out his Ryzen system OC tests were done at stock memory settings - the motherboard had reset his RAM to stock speeds and weren't running at 2933mhz... so the Ryzen OC scores jumped up too.

In the end Ryzen 7 1700x beat the 8700k at both "real stock" and "real OC" the 2nd time around

 

I understand the importance of a fair comparison. Stock should mean stock. But, the most important thing is whether or not Ryzen can keep up when you toss stock out the window. If overclocking is one of your objectives, and it is for many, what matters most is how far you can push it and which one is the real winner in that arena. They should retest and provide factual information when mistakes are made. Fix it, then move on to the juicy stuff (overclocking) so we can identify the most performance capable contender.

 

Well, actually Jay made 2 mistakes, one mistake made the 8700k faster at stock, and the 2nd mistake made the Ryzen 1700x OC score lower than the 8700k OC score.

Now that both mistakes have been "fixed", Ryzen 1700x is faster at stock *and* at OC:

And, I think there are other settings or a setting pushing up the stock 8700k CPU, or it's just a golden CPU, given all the much lower results from other reviewers.

AdoredTV found that the highest scores came from reviewers that got their CPU's directly from Intel, while the lower scoring reviewers got their CPU's elsewhere, coincidentally of course

And, last but not least, Jay put water cooling on the Intel 8700k and air-cooling on the Ryzen 1700x, even for OC...

 

Wow, that was silly. I expect better than that from Jay. Not just for comparison purposes, but dumb under any circumstances. Maybe he didn't have the parts handy for Ryzen. Would have been better to postpone the testing until he had what he needed for proper liquid cooling.

 

Actually, 3 other reviewers made the same 8700k stock settings mistake Jay made - not catching the 4 core 4.7ghz Turbo Boost from enabling MCE, see the highest 4 scores here:

The mid-range scores are assumed to be the "correct" scores, and the rest of the lower scores are assumed to be from CPU's that wouldn't sustain advertised "maximum" turbo, AdoredTV discusses this.

The lower scores could be mis-configured, bad BIOSs, or maybe the mid-range scores are benefiting from another as yet undiscovered "boost" setting and the low end scores are the real ones.

 

He says in the video he had both systems on air-cooling initially, but the Intel 8700k system was thermal throttling, so he had to swap in a water-cooling upgrade.

He said the Ryzen system wasn't experiencing thermal throttling even at full OC, but I think he was just to lazy to swap in a water-cooling setup, who knows how the Ryzen system would have performed with cooling operating temperatures.

It's not a fair comparison without both systems with the same cooling.

 

I mean all by now should know what the CBR-15 score should be for [email protected]. And all should know that Coffee have about same IPC as Skylake/Kaby. +- scores depending on what ram (speed+timings) you have.

 

The 8400 is real winner out of CFL1, right? I mean given the price to performance ratio, that's the CPU that is blowing Ryzens out of the water.

 

Might be OK for gaming or business/office productivity use, but it is a mainstream consumer CPU. It is missing hyperthreading and not unlocked, limited TDP, so performance potentially is essentially static and severely limited. Depends on your interests more than anything else. A lot of people would never consider it at all, but many will where a lower price is more important than higher performance.

Here is a side-by-side comparison: https://ark.intel.com/compare/126684,126687

 

You're correct. I was speaking practically not from an enthusiast perspective. Is the 8350K the overclocking variant of that?

 

It looks that way. That actually might be a better option, even if one is not going to overclock. It's nice to have abilities if you ever change your mind, or need to give it an extra nudge for something more taxing than the norm. It looks like a few bucks cheaper (which is odd). Stock non-turbo frequency looks a lot whole higher unless the specs are wrong, so just running it stock would stomp the snot out of the other CPU (4.0GHz vs 2.8GHz base clock is a huge difference).

Here is another side-by-side with that one added: https://ark.intel.com/compare/126689,126684,126687

 

i'd say 8700k and 1700x, once both chips overclocked to their "general" limit, which would be 4.9 to 5ghz for 8700k and 4 to 4.1ghz on 1700x, 8700k wins in overall application.

CB15 is very well optimized and 1700x at fully clocked was only barely able to beat 8700k by a tiny bit, this would mean other software not as well optimized will guarantee 8700k come out on top due to higher frequency, even in multi threaded application such as handbrake etc.

to know something is wrong usually we'll need a base reference point, especially in computer benchmark. either their 8700k is wrong, or the 8700k score they did compared to their 7700k is wrong. say for example if 7700k was scored much lower than 7700k should while 8700k scored normal, you'd think their 8700k review is wrong because % is over 50%.

the reviews picked out by adoreTV in the case, are mostly 8700k review being wrong thats because we know the general score of 7700k.

 

Let's add the Intel® Core™ i5-8600K, which could be the 'sweet spot' - https://ark.intel.com/compare/126685,126689,126684,126687

 

I would say oc'd 8350K will be better for today's gaming due higher clocks (oc'd). But will lose against 8400 in multi threaded workflow. The new LGA worlds BGA chips.

http://hwbot.org/submission/3672335_mackerel_cinebench___r15_core_i3_8350k_848_cb
http://hwbot.org/submission/3674224_ev0lv3_cinebench___r15_core_i5_8400_1012_cb

 

Absolutely! Anything less for mobile consumer grade will be sour grapes, not sweet.

The REAL sweet spot for LGA notebook enthusiast grade will be nothing less than 8700K.

Go big, or go home.

 

Extra ~$100 USD chump change for hyperthreading, 3MB larger cache, extra 400Mhz stock turbo and same TDP as the semi-crippled i5-8600K is a no-brainer. Skip having evening dinner for a week, drop a few pounds, and *BOOM* it's paid for.

 

You can not include the CPU (Ghz) speed on MSI afterburner. Every game is different. Unless you know something else, then please share.

You can see the CPU in-game speed percentage though. Also it's stated on some of the programs ran on that video. I hope that helps.

Link on forums about CPU speeds on MSI Afterburner:
https://www.reddit.com/r/overclocking/comments/6ulg32/how_to_show_cpu_clock_with_msi_afterburner/

 

Actually, you can show the CPU clock, or individual CPU core clocks, with newer versions of MSI Afterburner. I have no use for it since HWiNFO64 is a far superior option for managing on-screen display through RTSS.

 

Thank you @Mr. Fox I will check out the new version. I think mine is the old version. I have tried a lot of stuff and I can not get up the CPU GHz speed in-game. If anyone knows a tutorial, then please show me. I'm running out of patience with this MSI afterburner lol.

@Danishblunt Anyways, its kind of to late; as I have finished the video of the same run as last test on Battlefield: Hardline using the .925 Silver lid, but with "Full" fans on Clevo dash and full fans on the x3 U3+ Vardar mod cooler. The last test was on "Automatic" fans through the Clevo dash and 40% 1100RPM x3 U3+ Vardar mod cooler.

Last video (Auto Fans, 40% 1100RPM x3 Vardar cooler)
2x 1080 GPU: 57c max
CPU 7700K 4.5GHz: 68c max


This Video (Full Fans, 100% 3000RPM x3 Vardar cooler)
2x 1080 GPU: 49c max
CPU 7700K 4.5Ghz: 62c max

 

Just use HWiNFO64 and you can have any information from the sensor table you want on screen. You can control row and column layout, whether or not labels are used, etc. MUCH more control than MSI Afterburner. Like this example...


Open Sensors, click the configuration button and go to the RTSS tab. There are some images in this album linked below that will illustrate how to use that tab and control layout. Ping me if something doesn't make sense. I think you will pick up on it immediately. (They are old images and you will see some differences in the UI, but nothing major.)

https://imgur.com/a/UYC3s

The easiest way to see a realtime representation of your OSD layout is to use EVGA OCscanner utility and run the Julia or PhysX tests in a smaller window and as you add sensors and move their row, column or add labels you can see how it will look in realtime. Those tests are very light and won't really do much as far as stressing things. Do not use the Furmark test.

You can download the latest beta MSI Afterburner and RTSS from unwinder's posts in this thread:

https://forums.guru3d.com/threads/rtss-6-7-0-beta-1.412822/page-44

 

Sweeeett!
I never knew this. I'm going to do it on future benchmarks. Thank you for your time

 

You're welcome, bro. I added a few more details to the post above. EVGA OCscanner will make the layout super easy.

I even have it running while playing games. I like knowing everything about everything important no matter what I am doing.

 

Brother @Prema isn't alone in having his work stolen. Sadly, it is very common. The nerve and boldness of some of the retarded ass hats out there is dumbfounding. I do not know where some people find the gall to believe they have a right to claim things or insist on proprietary works being open source. A "sharing economy" mentality sucks six ways to Sunday.

See here: https://forums.guru3d.com/threads/rtss-6-7-0-beta-1.412822/page-44#post-5481163

 

Always hurts more when they go after the little guy :/

 

Absolutely. It is like a predator preying on the weakest animal in a herd because their chances of getting away with murder are greater. The bandits did the same thing to @svl7 too. Shameful. But, it's the same moral and legal dilemma even when the copyright owner is a big company like MSI.

 

Watch AdoredTV's take on the binning and poor silicon, Intel's new declaration of "best" Turbo and no additional information being supplied, he said he wouldn't trust the retail 8400's to hold best Turbo, like the 8700k with the wide variance in Cinebench results at the low end where a number of CPU samples didn't hold Turbo - and the 8700k is supposed to have the best silicon and therefore the best chance of holding Turbo speeds for a full benchmark, even a short one.

 

Edit my original post: I take my laptop in a bag with my Laptop and Surface Pro 4, 3DS, Switch, spare battery, surge protector, and a few other odds and ends. This bag goes with my twice a week to my local customers, and I take it through airports to customer sites remotely. The idea was that I needed a laptop to replace my desktop on the road as much as I could.

No the laptop isn't for everyone, but it suits my needs well.

 

You may have missed a few key details of the AdoredTV and Jay's reviews.

1) The motherboards are enabling MCE/All Core Sync giving 4.7ghz x 4 core at stock, instead of 4.3ghz, skewing stock scoring. Even after fixing this the wide range of CB scores 1200cb to 1400cb suggests Max Turbo isn't being held for the whole benchmark.

2) The Ryzen OC runs in the 1st review dropped RAM speeds, lowering OC score, other reviewers might have done the same if they weren't watching the RAM speed. For Ryzen lower RAM speeds tank the performance, you want near 3000mhz RAM speeds or greater.

3) The new Intel policy for Max Boost speeds, not declaring values except for Base and Max Boost shows Intel is hiding something, and given the wide variance of Cinebench scores across the reviewers, there are a good percentage of CPU's not holding Max Turbo throughout the whole benchmark.

The inability to hold max turbo might be motherboard or BIOS related, with additional tuning to those boards helping, but for now it's something to be aware of as being more likely CPU silicon quality related.

As better silicon is supposed to be in the top SKU's, then the bottom SKU's will likely be worse off. For the Clevo laptops this won't come into play so much unless someone wants a lower SKU than a 8700k.

It will be interesting to see how the "real" production Coffee Lake CPU's perform as they release in the coming months.

 

But we don't really know, there is too wide of a variance of results being reported.

Here is the range of stock supposedly 4.3ghz CB scores from a wide sampling of reviews, and the top 4 scores were actually run at 4.7ghz unbeknownst to the reviewer running the tests - so not stock as claimed, but instead a hefty overclock.

Taking out those top 4 scores, the rest of the scores range from 1200cb-1400cb, too wide a range to be considered reliable. Which is it 1200cb or 1400cb is the "real stock" result?

The overlay of results for the 7700k in comparison show the large variance in 8700k CB scores at "4.3ghz", those 7700k scores are tightly grouped, hardly any variance, much more reliable: