Jump to content
HWBOT Community Forums

Lucky_n00b

Members
  • Posts

    852
  • Joined

  • Last visited

  • Days Won

    11

Posts posted by Lucky_n00b

  1. How many 845 have you tested bro?

    I tested 3 and all have similar default Turbo VID at 1.475 - 1.4875V, Rather high compared to some other Kaveri athlons (except Godavari APU 7870k/7890k). Maybe these chips is already running close to its FMax? That would answer why AMD don't release K-version of these Athlon.

  2. Your last two posts are exactly why this will never happen. How do YOU propose to moderate those subs when you cannot even prove the method of cooling or temp?

    In a perfect honest world, great idea.

    The world is neither perfect nor honest.

     

    Sad but true.

    That being said, if there's a method (like a 'wrapper' Splave said) to correctly determine the maximum CPU or GPU temperature during the benchmark is ran, in which the security cannot be compromised, and the temp sensor can be trusted,

    it will open up a whole new possibility of an oc competition. (specific temperature-limited competition that's not only XTU for example)

  3. Agree with Pieter. I think a lot of people just know Daniel as 'the guy with the best CPU'. Well those CPU won't be running at its maximum speed by itself right? :P

     

    Even as someone who overclocks daily and doing oc competition as part of my job, I sometimes struggled to put out good scores out of certain hardware configuration. It really took dedication to be pushing the hardware to its limit every single time, both in online competition and live competition - and in that regard Daniel excels in both :)

     

    I got the honor of sitting beside him on the Galax GOC, so I knew that his skill in preparation, setup and handling system is better than most competitive overclocker I've met, should be on the same class as top live bencher like XA and Vivi.

     

    Keep pushing it, @Dancop!

  4. Maybe its logic reason, if I heard Bristol Ridge will be little different core than this FM2+ Excavator....

     

    Just read this (the AMD guide for Family 15h Model 60-6Fh): http://support.amd.com/TechDocs/50742_15h_Models_60h-6Fh_BKDG.pdf

     

    On page 85, I saw that on this models AMD might already given some DDR4-2400 support on its DRAM controllers(DCTs). This would pose a problem though, DDR4 has been evolved quite quickly these days, and DDR4-3000 should be found quite easily.

     

    Would've been a shame if AMD only have DRAM Ratios up to 2400 while its competitions can do 3600+ daily, really hope for the sake of their APUs AMD support higher DDR4 clocks..

  5. I was keep the turbo at "auto" and after raise the BCLK. Boot into the WIndows and after via TurboV. Did you had IDE mode or AHCI as at mine side?

     

    I see, I didn't try turboV since my version is not working. This is under AHCI mode, I've managed to boot straight from BIOS at around 114Mhz though, but with 35x multiplier.

     

    On XP IDE, Highest I've boot into OS was 120Mhz, but with lower CPU multi and low memory multi.

     

    http://techreport.com/news/29789/a10-7890k-and-athlon-x4-880k-take-top-spots-in-amd-fm2-lineup

     

    880K will be based on Excavator? Sounds like AMD Pi records are just around the corner if so.

     

    Based on that slide that said 16x PCIe 3.0 lane and 4MB total L2 Cache, it will be steamroller-based. Excavator only have 8x PCIe 3.0 and 2MB L2.

     

    I don't know if I had bad carrizo sample, but if hitting 4-4.1Ghz requires extra volts up to 1.55v, I might understand why AMD didn't make a K-series for Excavator athlons.

     

    This is like a mobile chip made into desktop, which not designed to hit big clocks. Not to mention this is similar 28nm process node as steamroller, further miniaturized by optimizing for density. This most likely hurts the overclockability and clock potential, the efficiency has a big price to pay it seems(not going to stop me from freezing this chip though :P).

  6. Silly AMD... can never get the NB right :rolleyes:

     

    Do you have unlocked control over this? Maybe these locked chips will consist of freezing to try to get NB up high.

     

    Both the CPU Multiplier and the NB are hard-locked, the only way you're increasing these is the BCLK, no other way.

     

    I found something strange though, in the BIOS version I tested on, the highest CPU Multiplier available is 35(not the 38x Turbo Multi). Since this would probably meant that the 38x is the Turbo state, I need to put AUTO in the CPU Multiplier, and OC BCLK from here. But larger than 109 Mhz BCLK, My CPU constantly dropped it's multi to 35x in OS, no matter what settings I put on. I even tried putting a pot and dropped the temp to 0C, no difference.

  7. Is that NB frequency correct or does it throttle?

     

    BIOS also noted it as 1.3Ghz on default, with 1.175 VID. Upper range on mine around 1400-1450Mhz is @ 1.35V VddNB, 1475-1500Mhz failed to boot.

     

    According to geekbench 3 memory scores, this NB affects memory performance greatly.

×
×
  • Create New...