SparkysAdventure

https://www.dropbox.com/s/e2j64ya6r7h35qs/ROG-MAXIMUS-XIII-APEX-ASUS-0611.rar?dl=0 https://www.dropbox.com/s/hnsl0l50aop5qlp/ROG-MAXIMUS-XIII-EXTREME-ASUS-0610.rar?dl=0 https://www.dropbox.com/s/t8l30zudba0amrk/ROG-MAXIMUS-XIII-HERO-ASUS-0610.rar?dl=0 https://www.dropbox.com/s/igblkoucki7vygk/ROG-STRIX-Z590-E-GAMING-WIFI-ASUS-0610.rar?dl=0 credit: @shamino

https://www.dropbox.com/s/g2w9u6b79wumfz7/ROG-MAXIMUS-XII-APEX-ASUS-0031.rar?dl=0

@flanker no maxmem at all, 32bit OS does not need. still maybe worth trying, maxmem size is relevant for XP on Intel

Will Cezanne be legal for this competition? I feel it should be, as it is AMD. Edit: maybe not legal for the competition until the products are actually announced. To be fair, the ES versions of Cezanne have been available second-hand for quite a while now.

New version with much lower load Delta coming out? Saw the Facebook post

okay, that's looking better

at least it runs now

#2 would likely be a good candidate for Ryzen 1:1 fabric benching, maybe RKL too if what I've heard holds true

As the other thread said https://drive.google.com/file/d/1Wljq9ELaJtlkywUj_TdbYCjwREXyZPGr/view XOC bios, AGESA 1201, L3$ fix another user reported memory training to be easier

@I.nfraR.ed When I tried normal public release 1.2.2, switching between DIMMs didn't show any difference. I'll recheck again soon with the beta version.

I'm looking into AMD's equivalent of RTL and IOL, which are RDL and MRL. I remember @chew* and @I.nfraR.ed mentioning these were configurable in OS on K10. I have now found evidence of two sets of memory, when training the exact same configurable timings and voltages, reporting different RDL and MRL. I have video of this property in the in the 7-zip file included in the post. I have also uploaded both videos to YouTube. Here's a better picture of what the Ripjaws were running when passing benchmarks: Running mixed sticks would report different RDL/MRL in bios, depending on what stick is where, but Zentimings reported the same. Running only Neos would not enter OS. I noticed that the set of Neos trained far more consistently, and I'm curious if RDL and MRL have anything to do with that. My friend at MSI pointed out a couple settings to me: DRAM Latency Enhance, I believe this is similar to having RTL enabled in bios to be configured when set to manual.l ARdPtrInitVal, what he referred to as DFI.RX.MP_ I believe MP1 is equivalent to IOL offset on Intel. It had a range from 0 to 127, but I could not measure what the default value is. MP0 also appeared to be equivalent to something on Intel. It had a range from 0 to 15. I remember there being another setting next to IOL offset on Intel with the same range, but on Intel changing it would not boot at all. I also cannot measure the default value of this. RxDatChnDly may be relevant as well. The only settings shown were 0 and 1. At jedec, bios would report 0, and at manual OC, it would set 1. Unfortunately, none of these settings in bios would change the reading of RDL and MRL. I think the settings are either non-functional, improperly implemented, or blocked by AMD. I would like to hear if anyone has more information or experience on this, finding a way to control these settings could be pretty huge. 598387889_thesesetsbootdifferentRTLandMRL.7z

@Midnight703

https://drive.google.com/drive/mobile/folders/1lFVdWK2mRZsjDDOkRAjai76VK4KWWVz_?usp=sharing agesa 1201

Pretty strong chips!

11k did in fact need some bclk.

Figured out how to push past the 1866fclk barrier, needed very specific vsoc on this chip. 11k in sight, will need some bclk

Some more bioses for the various Unify AMD boards. https://drive.google.com/drive/mobile/folders/1weetsVwOcIp7Sj4evlUKOLRM5ebLVEVN According to the filenames, I believe these are 1.2.0.0 based and supposed to address something WHEA related.

@ManniX Probably, don't have it yet tho. @buildzoid @Wizerty Can y'all ask your contacts?

@noliso PMing on Discord, nice score!

Still learning new tricks with this platform

Should try throwing a 9800x at this bench

Not daily, Samsung B-Die.

https://hwbot.org/submission/4619903 5.50.391 by @rivaldokfc He's done even lower but crashed before proper screen

@sabishiihito

@Mr.Scott Can you link me to some of this guy's odd submissions?