The Notebook Review forums were hosted by TechTarget, who shut down them down on January 31, 2022. This static read-only archive was pulled by NBR forum users between January 20 and January 31, 2022, in an effort to make sure that the valuable technical information that had been posted on the forums is preserved. For current discussions, many NBR forum users moved over to NotebookTalk.net after the shutdown.
Problems? See this thread at archive.org.
← Previous pageNext page →

    *** Official Clevo P870KM1/P870KM1-G/Sager NP9876 Owner's Lounge! - Phoenix 3.0 ***

    Discussion in 'Sager/Clevo Reviews & Owners' Lounges' started by Spartan@HIDevolution, Jan 5, 2017.

  1. MageTank

    MageTank Notebook Consultant

    Reputations:
    92
    Messages:
    123
    Likes Received:
    129
    Trophy Points:
    56
    Looks like it's only primary and secondary timings. Either way, this is far more than any other laptop I've seen. The fact that they give you access to tREFI and tRFC means tweaking latency for sub 40ns might be possible. As for your XMP woes, what kit are you using? Your IMC will respond differently to different IC's. There are various Samsung IC's (B-Die, D-Die, E-Die, etc) along with Micron, and SK Hynix (AFR/MFR). Each of these different IC's (along with their individual variations) prefer very specific tertiary timings, that are automatically trained by both your board and IMC. The fact that you cannot access them, means making them stable is more difficult. Especially if your kits are multi-rank, and you lack access to _DR sub-timings. Multi-rank kits, while benefiting from Rank Interleaving, require more tweaking to stabilize than their single-rank counterparts.

    From what I can tell with laptops, the G-Skill Samsung kits seem to be superior than the Corsair Hynix kits as far as plug and play stability goes. I have not seen any research on which clocks higher when manually tuned, as the memory overclocking community is niche as is, let alone finding enthusiasts (and the BIOS/hardware itself to do it) is difficult to say the least. That being said, I'd love to take a crack at the Premabios with as many different SO-DIMM kits that I can find. Part of the reason why I don't mind slowing my CPU down, is that I know I can somewhat make up for the additional CPU overhead that I add, by using faster memory. I'm just glad that people like @Prema seem to understand that enough to include these features in their BIOS. Here's to hoping we get some RTL/IO-L love in the future (if possible).
     
  2. D2 Ultima

    D2 Ultima Livestreaming Master

    Reputations:
    4,335
    Messages:
    11,803
    Likes Received:
    9,751
    Trophy Points:
    931
    You may get more timings available when using Intel XTU, but don't quote me on this. It was like that on my last laptop but I have not touched XTU on this one.

    I am using the G.Skill kits, but even setting 2400MHz using 100MHz * 12 instead of 133MHz * 9 wouldn't boot so I am sure I need to use less sticks until I get it to boot. I am an unusual case.
     
  3. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,710
    Messages:
    29,843
    Likes Received:
    59,629
    Trophy Points:
    931
    dm477, TBoneSan and Donald@Paladin44 like this.
  4. ole!!!

    ole!!! Notebook Prophet

    Reputations:
    2,879
    Messages:
    5,952
    Likes Received:
    3,982
    Trophy Points:
    431
    Papusan likes this.
  5. maxcrysis

    maxcrysis Notebook Enthusiast

    Reputations:
    5
    Messages:
    33
    Likes Received:
    22
    Trophy Points:
    16
    Capture+_2017-04-20-07-16-41.png

    The Clevo 1070 is not 115W.
     
  6. dm477

    dm477 Notebook Evangelist

    Reputations:
    56
    Messages:
    343
    Likes Received:
    220
    Trophy Points:
    56
    Just when I am getting the KM1! Well I need a machine now so it can't be helped.



    Sent from my iPhone using Tapatalk
     
    Papusan likes this.
  7. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,710
    Messages:
    29,843
    Likes Received:
    59,629
    Trophy Points:
    931
    Hit up Eurocom Or READ HERE:rolleyes:
    upload_2017-4-20_9-25-42.png
     
    Last edited: Apr 20, 2017
    TomJGX and maxcrysis like this.
  8. D2 Ultima

    D2 Ultima Livestreaming Master

    Reputations:
    4,335
    Messages:
    11,803
    Likes Received:
    9,751
    Trophy Points:
    931
  9. MageTank

    MageTank Notebook Consultant

    Reputations:
    92
    Messages:
    123
    Likes Received:
    129
    Trophy Points:
    56
    I believe it tends to differ per manufacturer. Nvidia themselves have stated that each individual board partner can have custom MXM's with different power delivery and even different memory configurations. The easiest way to know what the average power limit is on these cards, is for someone to actually load the card, then pay attention to the power limit in something like HWInfo64 using the GPU Power Sensor. Oddly enough, that image you posted came from Eurocom themselves, and as @Papusan pointed out, Eurocom also lists the 1070 at 120w. http://www.eurocom.com/ec/release(350)ec
     
  10. maxcrysis

    maxcrysis Notebook Enthusiast

    Reputations:
    5
    Messages:
    33
    Likes Received:
    22
    Trophy Points:
    16
    Sorry my mistake. I'm waiting for my laptop p870dm3 with two 1070s and I'll check it out, but once they say it's 120w it's probably going to be the case. This is very bad for me :( .
     
  11. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,710
    Messages:
    29,843
    Likes Received:
    59,629
    Trophy Points:
    931
    I would opt for either single or dual 1080 in the P870DM3 chassis. 1070 Sli doesn't make sense to me. But that's me.
     
  12. maxcrysis

    maxcrysis Notebook Enthusiast

    Reputations:
    5
    Messages:
    33
    Likes Received:
    22
    Trophy Points:
    16
    And I'm against sli and I would have taken a 1080 but got it at a very good price just for that.1800$ .I think it's worth it.
     
  13. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,710
    Messages:
    29,843
    Likes Received:
    59,629
    Trophy Points:
    931
    I will expect you would get back some money for the 1080 in resale if you later upgraded to the next from Nvidia. The 1070 version graphics from Clevo is a problematic graphics card. Aka won't fit in all models of laptops.
     
  14. cj_miranda23

    cj_miranda23 Notebook Evangelist

    Reputations:
    334
    Messages:
    567
    Likes Received:
    537
    Trophy Points:
    106
    Anyone here using NZXT cam software https://camwebapp.com/? Did you experience a delay or freezing on cursor when you use the track pad while the app is open?
     
  15. Meaker@Sager

    Meaker@Sager Company Representative

    Reputations:
    9,431
    Messages:
    58,189
    Likes Received:
    17,900
    Trophy Points:
    931
    It's polling is likely causing an interrupt issue, possibly with something else open?
     
  16. cj_miranda23

    cj_miranda23 Notebook Evangelist

    Reputations:
    334
    Messages:
    567
    Likes Received:
    537
    Trophy Points:
    106
    The issue occurs as long as the cam software is open, if it's close everything is fine. It doesn't matter which other software is open. Also if i used a separate mouse(wireless) the cursor is working fine! Can I modify the polling rate of the track pad?
     
  17. Rahego

    Rahego Notebook Consultant

    Reputations:
    77
    Messages:
    267
    Likes Received:
    136
    Trophy Points:
    56
    I have problem with new CCC acquired by obisidan tools - 5.0001.1.17. It does not run CPU_DRAM_OC (i did trick with changing dot in region settings). It does not start at system start, also - i am not able to run it by using exe in CCC folder - it shows errror right away. Anyone with similar problem?

    @John@OBSIDIAN-PC ?


    PS. I found same version on CLEVO.com - it also pops error, but at windows start it shows icon of CPU_DRAM working in background - description is CPU Memory OC OFF.
     
    Last edited: Apr 24, 2017
  18. John@OBSIDIAN-PC

    John@OBSIDIAN-PC Company Representative

    Reputations:
    754
    Messages:
    2,265
    Likes Received:
    2,862
    Trophy Points:
    181
    The OC thing was removed on purpose.
    We don´t recommend using it, we recommend the use of Throttlestop instead.
    Don´t ask for support for our app here, there is a thread just for it:
    http://forum.notebookreview.com/threads/clevo-drivers-update-utility-by-obsidian-pc.801464/
     
    dm477 and Spartan@HIDevolution like this.
  19. ole!!!

    ole!!! Notebook Prophet

    Reputations:
    2,879
    Messages:
    5,952
    Likes Received:
    3,982
    Trophy Points:
    431
    Papusan and dm477 like this.
  20. MageTank

    MageTank Notebook Consultant

    Reputations:
    92
    Messages:
    123
    Likes Received:
    129
    Trophy Points:
    56
    Honestly, it would depend. I imagine these are going to be clocked lower than their quad core Kaby counterparts. I also imagine the die is going to be physically larger, which should allow for soldering of the IHS to the die. In theory, it shouldn't run much hotter than their quad-core counterparts, but I wouldn't expect it to be a better alternative from a strictly gaming standpoint either. If you use your PC for things outside of gaming that would require a workhorse CPU, it'd be great, but most people looking for a gaming laptop will likely stick to the 7700k. Granted, newer titles are taking advantage of more cores, and it does aid especially well with minimum framerates, so I could be wrong and people might want them en masse. Either way, if these do work on Z270, it'd be a pretty great boon for people that wish for their laptop to maintain a decent resell value.
     
    Papusan and ole!!! like this.
  21. ole!!!

    ole!!! Notebook Prophet

    Reputations:
    2,879
    Messages:
    5,952
    Likes Received:
    3,982
    Trophy Points:
    431
    imho even for people who has 7700k should still upgrade just for the sake of future proof unless they can do 5.3/5.4ghz like johnksssssssssssss. 2 more cores make a lot of difference at least with what i do.

    i'd call that it wont get soldered IHS because intel mainstream they forgo solder IHS yrs ago since 3770k ivy era. i say it'll run hotter as we know 6c will need a lot more voltage and power, so heat/voltage both go up became a factor, though it is a 14nm++ so possibly even better silicon quality than 7700k kaby though it prob won't get any much better tbh.

    yea if it works with z270 then great, time to grab a p870km, or i can get one if clevo making a new one with a better cpu cooling heatsink, the current one for 7700k is a disappointment.
     
  22. dm477

    dm477 Notebook Evangelist

    Reputations:
    56
    Messages:
    343
    Likes Received:
    220
    Trophy Points:
    56
    Cooling competence will be the key. If these new hexa core processors are a challenge to cool, then I can't imagine many people upgrading to it. I made the mistake of upgrading 6700K to 7700K in my P775 chassis, without any delid, and had a difficult time keeping temperatures reasonable.
     
    ole!!! likes this.
  23. Jon Webb

    Jon Webb Notebook Evangelist

    Reputations:
    148
    Messages:
    315
    Likes Received:
    455
    Trophy Points:
    76
    My BIGGEST mistake was saving the money and getting the 1070. I'm waiting for the Prema bios to be in the KM1 and selling my machine and upgrading to a KM1 with the 1080. I figured installing a 1080 in my machine will cost me about $1300.00. Buying a KM1 with the 1080 and selling my machine, I should still be into the whole thing for about $1300.00
     
    Papusan likes this.
  24. MageTank

    MageTank Notebook Consultant

    Reputations:
    92
    Messages:
    123
    Likes Received:
    129
    Trophy Points:
    56
    The reason they've foregone soldering the current mainstream platform SKU's is due to the area size of the die. With a smaller surface area, you are subjected to higher chances of solder cracks forming from thermal shock. With more cores (and subsequently a larger die) you have a much better chance at soldering the die with lower risk of cracks ruining it. I believe Der8uer has a guide on this somewhere. Once I get home, I'll link it here as it's a good read.
     
    ole!!! likes this.
  25. Meaker@Sager

    Meaker@Sager Company Representative

    Reputations:
    9,431
    Messages:
    58,189
    Likes Received:
    17,900
    Trophy Points:
    931
    That same die is in the 7700HQ which is soldered and the package is the part that solders to the board and package size is not always directly linked to die size. (also the CPU die itself is soldered onto the package to give you an idea of what high end soldering can do).
     
  26. MageTank

    MageTank Notebook Consultant

    Reputations:
    92
    Messages:
    123
    Likes Received:
    129
    Trophy Points:
    56
    The 7700HQ's die isn't soldered (to a heatsink/IHS). While it's true that the BGA SKU's are surface-mounted to their respective boards, the dies themselves are not soldered to their heatsink (or in the case of desktop SKU's, their IHS). Unless you are referring to the process in which the die itself is placed on to the substrate, but that still wouldn't apply here as the heat isn't being produced at the contacts themselves, but through the top of the die itself. As for the reading material I previously promised to the other member, here it is: http://overclocking.guide/the-truth-about-cpu-soldering/

    It's definitely an interesting read for those with the time to spare.

    EDIT: I should really be more specific. I didn't mean that the contacts under the IHS won't get hot, I meant that the focal point of the heat itself will spread towards the surface of the die, as that is where the heatspreader/heatsink will be to take the heat away. The context in which we are speaking, are why Intel no longer solders the surface of the die, to their heatspreaders. The primary reason (aside from cost) is longevity. Cracks from thermal cycling/shock is one of the biggest reasons as to why they avoid soldering the surface of smaller dies to their IHS. On larger dies, the voids that form are less of an issue, as the total surface area is larger, to the potential for these voids to form a crack with one another is lower than a smaller surface area, as the chances will increase dramatically.

    Sorry for my poor wording.
     
    Last edited: Apr 24, 2017
  27. Meaker@Sager

    Meaker@Sager Company Representative

    Reputations:
    9,431
    Messages:
    58,189
    Likes Received:
    17,900
    Trophy Points:
    931
    I misread and thought you meant soldered as in BGA.

    It's true that a small die makes soldering a core to an IHS much harder.
     
    MageTank likes this.
  28. ole!!!

    ole!!! Notebook Prophet

    Reputations:
    2,879
    Messages:
    5,952
    Likes Received:
    3,982
    Trophy Points:
    431
    hmscott likes this.
  29. MageTank

    MageTank Notebook Consultant

    Reputations:
    92
    Messages:
    123
    Likes Received:
    129
    Trophy Points:
    56
    Who needs optane when these bad boys support 64GB of ram? Ram's caching drives > Optane. Would rather keep my M.2 slots for raw SSD space (since I don't really do heavy sequential writes) and use my ram itself to defer writing to the SSD's via Primocache. Part of the reason why I can't wait to get my hands on one of these. I use an ITX desktop board that is limited to 64GB of ram. To me, these laptops would be an upgrade, lol.
     
    ole!!! and hmscott like this.
  30. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,710
    Messages:
    29,843
    Likes Received:
    59,629
    Trophy Points:
    931
    I'm sure We can expect soldered die to IHS on mainstream chips this time. Intel wouldn't want to be worse than AMD regarding TiM for 6 cores chips :D And we know Intel wouldn't push out new chips that can't compete properly :rolleyes: Better temp means overclocking. Hence, expect it this time.
     
    ole!!!, TBoneSan and hmscott like this.
  31. ole!!!

    ole!!! Notebook Prophet

    Reputations:
    2,879
    Messages:
    5,952
    Likes Received:
    3,982
    Trophy Points:
    431
    naw, ram cache has obvious draw back than an actual storage bootable device thats close to ram speed level. cold boot you can't benefit from ram caching, plenty of software out there dont even take advantage of using ram to cache, its a horrible idea in general imho, better to just get optane storage device.

    besides, ram disk isn't on the same level of ram speed because it goes through software, meaning after going through software the level of performance we'd expect drops significantly, though still faster than optane but generally pointless. too many drawback for too little boost.

    with all that been said, i do use primo software to create ramdisk and cache bunch of my garbage files, wipe it all clean as soon as i do a restart its pretty nice, too bad i only got 32GB right now i want 64GB soon when i upgrade... soon...


    u never know, its intel :D it just might come with yet another crappy TIM. i'll still go for intel due to IPC crown by a 5-7% over ryzen and their 14nm++ better silicon quality than first gen 14nm glofo is junk. 14nm++ from intel mean a possible chance to see 6c 5ghz with decent voltage for once.
     
    Papusan likes this.
  32. Mobius 1

    Mobius 1 Notebook Nobel Laureate

    Reputations:
    3,447
    Messages:
    9,069
    Likes Received:
    6,376
    Trophy Points:
    681
    480w power limit affecting gpu slave

    https://i.imgur.com/IUHxqNx.png

    With this, CPU is limited to 55w no matter what (if 2 gpu is loaded), if you overclock is stable 24/7 (based on stress without gpu load) it will not be stable under combined stress test due to the cpu trying to reduce voltage to match 55w

    even running superpi 32m while doing fs ultra loop will force the cpu to use such low voltage that bsod occurs
     
    Mr. Fox likes this.
  33. dm477

    dm477 Notebook Evangelist

    Reputations:
    56
    Messages:
    343
    Likes Received:
    220
    Trophy Points:
    56
    5gh on 6 cores...now that will be something. Though apart from diminishing returns issue, I wonder if Intel will actually have stock clocks around 4.0 ish for people to be able to overclock it to 5.0ghz!


    Sent from my iPhone using Tapatalk
     
    ole!!! likes this.
  34. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,710
    Messages:
    29,843
    Likes Received:
    59,629
    Trophy Points:
    931
    What FS - 3DM11 score? And where did you buy Your machine?
     
  35. Mobius 1

    Mobius 1 Notebook Nobel Laureate

    Reputations:
    3,447
    Messages:
    9,069
    Likes Received:
    6,376
    Trophy Points:
    681
    Papusan likes this.
  36. MageTank

    MageTank Notebook Consultant

    Reputations:
    92
    Messages:
    123
    Likes Received:
    129
    Trophy Points:
    56
    The only obvious drawback I see, is the volatile state of ram. If your PC crashes, everything within the ramdisk won't be saved. As for "bootable devices close to ram speed level", I'd have to disagree. Ram, even with the speed penalty imposed by software level caching, is magnitudes faster than the fastest NVMe SSD's. We are talking over 10x faster in terms of raw bandwidth alone, not even counting ram's extremely superior latency advantages. My ram as of right now, hits 55GB/s writes, 50GB/s reads in Aida64 with a latency of 37ns. Using it as a ramdisk, I still manage to hit 40GB/s in reads/writes. That's still over 10x faster than the 960 Pro at it's peak performance. Granted, nothing I do ever requires this speed, and I am bottlenecked by literally every other device (network card, other storage solutions, etc).

    As for boot times, we've pretty much peaked with standard SATA SSD's. NVMe drives are no faster when booting, and in my personal experience with the MyDigitalSSD BPX, it boots slightly slower than my old 850 Evo M.2. As for "software not taking advantage of ram to cache", I've had no issue with software not recognizing my Primocache setup. It literally intercepts I/O data from your disks, for your OS.
    [​IMG]

    Not trying to act as some sort of spokesperson for Primocache, as I am not paid to do so, I just personally use it and know it works, lol. Either way, I don't see Optane as an impressive alternative to what we can already do, especially since we have to sacrifice an M.2 slot for it. I'd much rather use extra ram, or an extra SSD to cache my spinners. If you already have a Z270 chipset and didn't already have copious amounts of ram, or a large SSD to pair with your spinners, then by all means, Optane might be a cheap solution for that problem, but it's certainly not a feature I would upgrade platform/chipset for.
     
    CaerCadarn likes this.
  37. ole!!!

    ole!!! Notebook Prophet

    Reputations:
    2,879
    Messages:
    5,952
    Likes Received:
    3,982
    Trophy Points:
    431
    thats why i linked the review, boot time is faster with optane by a bit. not only that make computer completely snappy at pretty much highest level at that point.
     
  38. MageTank

    MageTank Notebook Consultant

    Reputations:
    92
    Messages:
    123
    Likes Received:
    129
    Trophy Points:
    56
    The review didn't show stand-alone boot times with just the SSD. It compared Optane + HDD boot time (14 seconds) against Optane + SSD boot time (10 seconds). This makes it difficult to see whether or not Optane is helping the SSD boot faster, as we don't have the information as to how fast the SSD is capable of booting on it's own. I'd say it's not helping it by much at all. For anyone with an SSD, Optane just doesn't serve a purpose for them.

    I would say the best overall setup is: Large HDD's for mass storage, single SSD to cache the drives, and use ram to defer mundane writes from hitting the SSD to improve it's lifespan. I don't really see Optane fitting in. I am also curious about the longevity of Optane, since it's bound to have the same finite lifespan of SSD's when it comes to constant writes. I suppose we will get that answer in the future.
     
    CaerCadarn likes this.
  39. ole!!!

    ole!!! Notebook Prophet

    Reputations:
    2,879
    Messages:
    5,952
    Likes Received:
    3,982
    Trophy Points:
    431
    no.. read again lol, its 10.5 sec with optane, theres a few video, one of them is OS on optane alone.
     
  40. ole!!!

    ole!!! Notebook Prophet

    Reputations:
    2,879
    Messages:
    5,952
    Likes Received:
    3,982
    Trophy Points:
    431
    also heres the ramdisk created by software, its not even close to the actual ram performance due to latency. https://www.pcper.com/image/view/81097?return=node/67578

    heres another optane m.2 cache, m.2 cache + HDD boot time is almost as fast as nvme SSD and 2x 2.5" SATA SSD. pretty good yo, too bad i wont be using it as cache, i'll be using it as just a boot device and watch computer fly.

    lolol take a look at this
    https://www.pcper.com/image/view/81229?return=node/67603
    optane only 32gb.. 14.6 latency while Samsung 960 evo NVME at 95 latency.. thats 1/6th of NVME's latency, so damn snappy in 4k QD1 read
     
    Last edited: Apr 25, 2017
  41. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,710
    Messages:
    29,843
    Likes Received:
    59,629
    Trophy Points:
    931
    hmscott likes this.
  42. Mobius 1

    Mobius 1 Notebook Nobel Laureate

    Reputations:
    3,447
    Messages:
    9,069
    Likes Received:
    6,376
    Trophy Points:
    681
    With stock EC and bios, it is limited to 480w


    The slave card cannot fully utilize the TDP, meaning that it's throttled.


    The DM3 in that score is running premamod, which removes the 480w limitation.
     
    hmscott likes this.
  43. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,710
    Messages:
    29,843
    Likes Received:
    59,629
    Trophy Points:
    931
    I know. See Schenker XMG U727 (Clevo P870DM3-G) 6700K + 1080 Sli push 490w. But this is done from different load than FS or 3DM11. Test with 3DM11 if you not already have done it. See what you can push.
     
    hmscott likes this.
  44. Mobius 1

    Mobius 1 Notebook Nobel Laureate

    Reputations:
    3,447
    Messages:
    9,069
    Likes Received:
    6,376
    Trophy Points:
    681
    furmark and occt don't push past 480w, only spikes to 500w

    I can test 3dm11 later
     
    hmscott and Papusan like this.
  45. MageTank

    MageTank Notebook Consultant

    Reputations:
    92
    Messages:
    123
    Likes Received:
    129
    Trophy Points:
    56
    Read what again? I just said myself that Optane + SSD = 10.5 seconds. This is your source, is it not? http://www.tweaktown.com/reviews/8157/intel-optane-memory-32gb-2-nvme-ssd-review/index3.html

    [​IMG]

    Take your own advice here, and read again. Nowhere on that page does it show a stand-alone SSD, against Optane. Optane, in it's current state, is essentially worthless to the vast majority of people. It can't even do write consolidation. As for your random screenshots of benchmarks without any data to test against, I'll stick to my own benching. In fact, I'll go ahead and run a few storage benchmarks with my ram under Primocache, to show you why Optane isn't even close. Be right back shortly.

    To the rest of you, sorry for derailing this thread with storage.

    EDIT: The results of my benching with Primocache level 1 config:
    [​IMG]
    [​IMG]
    I'd say Optane has a very long way to go before it measures up to ram in terms of raw bandwidth and latency. ASSSD is likely being bottlenecked by my CPU at this point (with it being uncompressed and all) with CDM being heavily compressible and slow. Either way, the results speak for themselves. Again, the risk with this method is, any hardware failure/crash will likely require a reinstall of windows, so a mirror backup needs to be performed often (have mine scripted to do so daily at specific times) and a backup battery (UPS) is also handy in this config. This is another reason why I am considering this P870KM1. It's one of the few laptops with 64GB ram support AND DDR4 3000 support. I assume with Prema, even more memory options will be available to us. Super excited to tinker when I get my hands on one.
     
    Last edited: Apr 25, 2017
  46. Mr. Fox

    Mr. Fox BGA Filth-Hating Elitist

    Reputations:
    37,225
    Messages:
    39,334
    Likes Received:
    70,637
    Trophy Points:
    931
    That's pretty darned sad. At this point I'd be inclined to suggest that nobody purchase one until @Prema firmware is available to correct that problem. Just put the money in the bank and wait until it is fixed and works properly. If it doesn't get fixed, use the money to build a desktop instead. Clevo should be ashamed of this kind of nonsense, but apparently they do not care or it never would have been released without gonads.
     
  47. Mobius 1

    Mobius 1 Notebook Nobel Laureate

    Reputations:
    3,447
    Messages:
    9,069
    Likes Received:
    6,376
    Trophy Points:
    681
    Though @Meaker@Sager said it was a good thing for regular users to not have unlocked power draw...

    Only "advanced" users would benefit from it.
     
    TomJGX, Papusan and hmscott like this.
  48. MageTank

    MageTank Notebook Consultant

    Reputations:
    92
    Messages:
    123
    Likes Received:
    129
    Trophy Points:
    56
    Buying a multi-GPU laptop with a desktop CPU in it kinda qualifies you as "advanced user" in my eyes. Can't imagine your average user being capable of dealing with the hassle of SLI in general, let alone trying to tame the thermals of a desktop CPU with relatively small heatsinks.

    I'd say, having a default power limit, with the means of unlocking it is a far superior option than outright locking it completely. I disagree with this mindset that manufacturers have that is basically them assuming all of their consumers are idiots, and that they have to protect them from themselves at all costs. There is a reason when you load overclocking software of any kind, you get a warning beforehand. Those who do not heed it, are on their own if they mess something up.
     
    Papusan and hmscott like this.
  49. Mobius 1

    Mobius 1 Notebook Nobel Laureate

    Reputations:
    3,447
    Messages:
    9,069
    Likes Received:
    6,376
    Trophy Points:
    681
    Yea, especially with the amount of money put into the SLi 1080 setup, it's just weird that the stock bios would castrate the performance.

    Though single card and SLi 1070s would probably still be ok, maybe...
     
    Papusan, hmscott and MageTank like this.
  50. Mr. Fox

    Mr. Fox BGA Filth-Hating Elitist

    Reputations:
    37,225
    Messages:
    39,334
    Likes Received:
    70,637
    Trophy Points:
    931
    Well, that's one person's opinion. I wonder if anyone that buys one thinking that it is going to function properly would actually agree with that. I kind of doubt it.

    I'm not sure how it can actually be considered a "good thing" for anyone, whether they overclock or not. For one thing it is probably accurate to say that "regular users" don't buy a beast like this, and Clevo castrating it is just totally dishonest. I mean, c'mon... if the CPU cannot even hold stock clock ratios and at least reach the stock 91W CPU TDP, that is sinful and probably even qualifies as fraud.

    Talking about "unlocked power draw" is one thing, but in this case we are talking about a product that is essentially incapacitated and is not even capable functioning normally. That sucks and I think even "regular users" would and should be livid about it. Unless or until @Prema fixes it or Clevo takes accountability for correcting their mistakes, the product can only be legitimately viewed as broken and defective, and that is totally unacceptable.

    It needs to be pulled back from all resellers, production halted, and no more units sold until it functions correctly. That's the message I think Brother @Meaker@Sager should be conveying to Sager and Clevo. But, that's just my opinion... for whatever that's worth.

    Maybe he can gang up with @Eurocom Support @Prostar Computer @Ted@HIDevolution @Donald@HIDevolution @pat@XOTICPC @mythlogic @Larry@LPC-Digital @John@OBSIDIAN-PC and @XMG to let Clevo know they are not going to be selling emasculated trash for them anymore. Unless it works 100% off the shelf it never should have been allowed to be put into production, much less allowing the product to leave the factory to end up in some unfortunate customer's lap. It's not a crippled BGA turdbook and should not perform like one. People that want broken garbage can buy that from half-assed outfits like Alienware, ASUS or Razer.
     
    Last edited: Apr 25, 2017
    Jon Webb, Papusan, bennyg and 2 others like this.
← Previous pageNext page →