Ryzen 5000 Series Discussion

Discussion in 'AMD x86 CPUs and chipsets' started by Skramit, Nov 7, 2019.

  1. RnR

    RnR Member

    Joined:
    Oct 9, 2002
    Messages:
    16,772
    Location:
    Brisbane
    The press release posted above shows it in action on both Intel and AMD platforms. They used a Threadripper for some reason though...

    You might have missed nCrypt 's golden 5800X that is quite happy with an IF of above 2000Mhz.
     
    Last edited: Jun 10, 2021
    SnooP-WiggleS likes this.
  2. nCrypt

    nCrypt Member

    Joined:
    Feb 12, 2010
    Messages:
    2,421
    Location:
    Melbourne
    Will work fine on Ryzen.

    The only issue is finding a CPU that will do 2000mhz IF allowing you to run the ram at 4000mhz, as snoop said below, want that 1.1 ration IF:DRAM

    im running my 4266c19 royals @ 4000 16-16-16 on my 5800X with IF 2000. can run IF 2133 and rams at 4266 but volts quite a bit higher then 2000.

    Runs beautifully,

    love this CPU :D

    my 5950x also boots 2000 IF, but havnt much more trouble getting that stable.
     
    SnooP-WiggleS likes this.
  3. RnR

    RnR Member

    Joined:
    Oct 9, 2002
    Messages:
    16,772
    Location:
    Brisbane
     
    mAJORD and MUTMAN like this.
  4. raincloudx

    raincloudx Member

    Joined:
    Dec 18, 2007
    Messages:
    1,962
    Location:
    Victoria
  5. mAJORD

    mAJORD Member

    Joined:
    Jun 4, 2002
    Messages:
    11,948
    Location:
    Griffin , Brisbane
  6. SKITZ0_000

    SKITZ0_000 Member

    Joined:
    Jul 19, 2003
    Messages:
    1,457
    Location:
    Tas
    I've had some experience with 4 x SR Micron e-die ballistix elites vs 2 x DR Micron Rev.e vanilla ballsitix. I think the board makes a difference in this case and mine was on the cheaper side, but I had way better results on the dual rank kit. Both kits as a pair were happy at 3800C16, but the four sticks were a nightmare to stabilise for daily and I ended up running them at 3600C16, I sold them in the end because of it. If you have an ultra high end board you might have more luck, I was literally on a $200 mATX board.
     
    mAJORD likes this.
  7. mAJORD

    mAJORD Member

    Joined:
    Jun 4, 2002
    Messages:
    11,948
    Location:
    Griffin , Brisbane
    Thanks for the feedback. That's the sort of thing I'd expect I guess.

    Problem is finding DR at W decent price.. er at least guaranteed to be DR
     
  8. MUTMAN

    MUTMAN Member

    Joined:
    Jun 27, 2001
    Messages:
    9,392
    Location:
    4109
    was that zen 3 SKITZ0_000 ??

    I'm sure I linked some vids on dual vs single rank over in the FAQ
     
    mAJORD likes this.
  9. kr0k

    kr0k Member

    Joined:
    May 9, 2002
    Messages:
    1,406
    OJR likes this.
  10. SKITZ0_000

    SKITZ0_000 Member

    Joined:
    Jul 19, 2003
    Messages:
    1,457
    Location:
    Tas
    Yeah on my 5950X agesa 1.2.0.1
     
    Last edited: Jun 14, 2021
    MUTMAN likes this.
  11. MUTMAN

    MUTMAN Member

    Joined:
    Jun 27, 2001
    Messages:
    9,392
    Location:
    4109
    funny cause I thought I read AMD had sprinkled magic beans on the zen 3 setups :confused:

    maybe it's like you say, the mobo plays a big part in this
    it'd be interesting to see something like the impact board in a single vs dual shootout
     
  12. nCrypt

    nCrypt Member

    Joined:
    Feb 12, 2010
    Messages:
    2,421
    Location:
    Melbourne
    wouldnt have a clue if my kits are SR or DR

    currently have a 5950x in the impact, with this kit some shitty 32gb 2x16 3200 c16-18-18.

    so dam tempted on the 32gb 4000 14-15-15 royals, but $800 for a DDR4 mem kit when DDR5 isnt that far away is tough!
     
    MUTMAN likes this.
  13. ska'

    ska' Member

    Joined:
    Dec 25, 2011
    Messages:
    1,768
    My G-Skill 32gb 3600c14 kit decided it now longer wants to run at XMP, has to be downclock to 3400mhz in order to not crash PC.

    Pain in the ass
     
  14. OJR

    OJR Member

    Joined:
    Jan 19, 2013
    Messages:
    5,534
    Location:
    Melbourne
    That doesn't sound right. That sounds more like a BIOS update screwing things up?
     
  15. RnR

    RnR Member

    Joined:
    Oct 9, 2002
    Messages:
    16,772
    Location:
    Brisbane
    Or hardware that needs to be reseated?
     
  16. nCrypt

    nCrypt Member

    Joined:
    Feb 12, 2010
    Messages:
    2,421
    Location:
    Melbourne
    Did you update bios?

    i would up the volts, dual sided sticks in 2x16 or 4x8gb configs even though rated 1.35v sometimes prefer 1.4 or up too 1.45v to run rated timings stable. this is all dependent on cpu/board configurations.
     
  17. ska'

    ska' Member

    Joined:
    Dec 25, 2011
    Messages:
    1,768
    Kit runs at 1.45 on XMP, same bios I've been running for a while..just started shitting itself whilst running BF4...
     
  18. Butcher9_9

    Butcher9_9 Member

    Joined:
    Aug 5, 2006
    Messages:
    2,522
    Location:
    Perth , St James
    Strange, what are your other voltages running at (SOC ect).
     
  19. netlord

    netlord Member

    Joined:
    Jun 27, 2001
    Messages:
    569
    Location:
    Melbourne
    Some updates...

    I tweaked my 3600 ram timings from 16-18-18-38 to 14-16-15*-38 at 1.4v
    *i think one reached 15

    Then i put them back to stock XMP but raised it to 3800 (1900 IF) and it worked ok. But I couldnt touch the timings and had to leave them at stock.
    Both performed within 1fps (margin of error) when doing Shadow of the Tomb Raider benchmark.

    Stock RAM before XMP = 149fps (but i think the GPU was stock too, all others GPU overclocked so maybe it wouldve hit ~155fps)
    XMP applied = 170fps
    with Tweaks = 177fps

    Now Im doing some undervolting and per core optimiser was successful at -30 on all of the cores. (couldve just set it to all cores in the beginning!)
    It only resulted in +1 fps in SoTR, but there were noticeable improvements in synthetic tests CPU-Z with Single thread = 686, Multi-thread = 9959 (before 670/8500 respectively). Cinebench r15/20 also seeing improvements too. I havent tried any other game benchmarks yet.
    Peak single core speed is 4.95Ghz, multicore between 4.5 to 4.55Ghz.

    I theres a couple more options i want to try to see if i can hit the magical 5Ghz single core speed, even though it will be relatively meaningless.
     
    MUTMAN likes this.
  20. Gibbon

    Gibbon grumpy old man

    Joined:
    Jun 30, 2001
    Messages:
    6,513
    Location:
    2650
    So ... finally (*cough* OJR) got my 5950X up and running over the long weekend. Only had the bits sitting around for about 2 months ... sigh.

    Anyway, fear my GT710 :leet:
    Can make a PC fairly neat when there aren't any power cables going to the GPU ....

    IMG_4695.jpg

    (will move the 6800 across from my old PC in the next day or two - it's still in use).

    Have whacked Win10 on it, and have been playing around a bit before I do a proper software install. Ran corecycler on it for about 2.5 iterations (about 4 hours) just now with no errors.
    All I've done in the BIOS so far is kick the RAM up to XMP (or whatever it's called in AMD land) 3600-16-16-16-36 @ 1.35V, enabled the PBO FMax Enhancer, and set the boost override to 200MHz. PBO is on auto (which I think is off on this board?). Everything else is on Auto.

    The best cores seem to be hitting 5.25/5.20GHz, and max VID is 1.494V (briefly, I guess) during the corecycler run.
    corecycler.PNG

    All core speed running OCCT seems to be around 4.6GHz at about 1.25V VID.
    OCCT.PNG

    What else should I do? I had a quick play by setting PBO to enabled instead of Auto, and it seems to drop the speeds a bit?? Weird.

    Cheers.
     
    Last edited: Jun 15, 2021
    BuzzPuppy, Perko, netlord and 6 others like this.

Share This Page

Advertisement: