Ryzen Owners Zen Garden

[dont whant to set it"']

To delid or not to delid a 5800X from the point of CCD and IOD heights under the ihs, are these same height at top of insolator?

le: Anyhow, the deliding is in progress , patience due course and no bent pins(so far).
To be used with liquid metal alloy TIM if successful(direct die contact between heatsink and chips).

 

[freeagent]

I'm above every 5600x on techpowerup's cinebenchr23 list with 1,610 on single thread

freeagent`s Cinebench - R23 Single Core with BenchMate score: 1625 cb with a Ryzen 5 5600X

The Ryzen 5 5600X @ 4900.3MHzscores getScoreFormatted in the Cinebench - R23 Single Core with BenchMate benchmark. freeagentranks #null worldwide and #29 in the hardware class. Find out more at HWBOT.

hwbot.org

 

[oobymach]

Impressive, I just broke 1600 with 1.2.0.3c bios and pbo with a +200mhz offset, if I do a manual oc I can break 12k multi but single core drops to under 1500.

 

[RJARRRPCGP]

Impressive, I just broke 1600 with 1.2.0.3c bios and pbo with a +200mhz offset, if I do a manual oc I can break 12k multi but single core drops to under 1500.

View attachment 237492

I've been using CTR 2.1 RC5 and the only thing I see wrong, is that CPU-Z bench scores always drop, but always OK with Cinebench. Also, with RAM at only 3200 with low timing values with solely XMP, I get roughly the same as you with Cinebench single core.

 

[Mussels]

Noticed super odd thing recently. Normal boot, even with fast boot enabled was always around 12 seconds since forever. Switched CPU Vcore from "Normal" to "Auto" and boom, now booting in 7.6 seconds. That whole fan spin-up pre-post cycle is now like 2 seconds instead of 5.

Was it actually stable, or was it crashing out and ignoring the setting you chose?

 

[Chomiq]

Was it actually stable, or was it crashing out and ignoring the setting you chose?

It was stable. It's almost like fast boot didn't work until now.

 

[tabascosauz]

It was stable. It's almost like fast boot didn't work until now.

Wouldn't read too much into it, time spent pre-POST is largely memory training which is dependent on what you have. When I was on B450I Aorus and B550M TUF, it wasn't until around I switched to B-die that I ever saw any proper "fast boot" whether it was enabled/disabled in BIOS or Windows. 4Gb E-die takes a very long time to train, CJR not so much but not "fast".

And even afterwards it has never been consistent either. Give it a couple months and if it starts up quick every single day without exception, maybe there's something going on. Sometimes I'd change some red herring and temporarily believe that the problem was "fixed" by something else, but best not to read too much into this stuff.

 

[dont whant to set it"']

There , delided.

 

[Chomiq]

There , delided.

What's with bare copper showing up on the edges?

 

[Taraquin]

Impressive, I just broke 1600 with 1.2.0.3c bios and pbo with a +200mhz offset, if I do a manual oc I can break 12k multi but single core drops to under 1500.

View attachment 237492

With pbo and co you should break 12000 if cooler is good enough. Trick is to keep temp below 71-72C. I can do 12050points with +50 pbo, -30 allcore and 88W PPT (consumption never exceeds 84W at 4.7GHz). With stock 76W PPT allcore runs at 4.575-4.65GHz and I get 11800 points.

 

[oobymach]

With pbo and co you should break 12000 if cooler is good enough. Trick is to keep temp below 71-72C. I can do 12050points with +50 pbo, -30 allcore and 88W PPT (consumption never exceeds 84W at 4.7GHz). With stock 76W PPT allcore runs at 4.575-4.65GHz and I get 11800 points.

Breaking the 12k isn't the hard part, it's keeping it under that magic 72 degree thermal throttle. Good advice though, tried the +50 -30 but had isues, +50 and -25 though ran fine until the thermal ceiling so I'm still messing around with it. I hit 72 while gaming too (CODWW2) as well as p95 stress test.

 

[freeagent]

Throttle in the 70s, really?

 

[oobymach]

Throttle in the 70s, really?

Yeah when the 5600x hits 72-73 degrees it goes from stable 4.7ghz all core to random drops in frequency as soon as it hits that temp, stressing with p95 it's steady 4.7ghz on all cores with the +50 pbo and -25mv allcore until it gets warm.

 

[freeagent]

Hmm, I will have to try it sometime.. you sure it’s not the current bringing it down? What kind of cpu ppt and edc are you looking at when that happens?

 

[Final_Fighter]

 

[oobymach]

Hmm, I will have to try it sometime.. you sure it’s not the current bringing it down? What kind of cpu ppt and edc are you looking at when that happens?

I think they were on auto, using 95w 75a and 100a limits atm and temps are down a bit, if I go +200mhz offset it wants to use the full 95w in p95 right away and holds up around 4.75-4.8ghz all cores until all 12 workers pass the first round then it dropped to 4.4-4.5ghz range.

With the +50mhz with -25mv allcore it pulls about 80 watts and runs 4.7ghz all cores until second round of p95 then it jumps to 95w and starts dipping from 4.6-4.7ghz randomly but not below 4.6ghz. Still testing for maximum speed.

 

[Taraquin]

Throttle in the 70s, really?

If PPT is kept equal, temp alone will drop allcore speed by over 200MHz if temp goes from 60 to 80C.

A quick clockspeed comparison in CB23 with +50 pbo and - 30 CO, initial clockspeed first, then sustained after 3 rounds at 76W PPT:
Stock cooler: 80C, 4.55-4.4GHz 10900p
Bequiet dark rock slim: 73C, 4.625-4.55GHz 11550p
Noctua D15: 67C, 4.65-4.625GHz 11850p

Throttle in the 70s, really?

It's not actually throttle jefore close to 80C, but the warmer a CPU gets, the more power it uses. If kept at a PPT limit, clockspeed will drop a bit if temp rises. On my i5 12400F a simple swap of cooler lowered max consumption in CB from 78W to 72W due to temp dropping from 82 to 62C, this is at max boost of 4GHz allcore. Ryzen boosts allcore as high as it gets eithin PPT limit. If you use +200 pbo, CO and no limit and have a very good cooler you can get allcore 4.85GHz using like 140W on a 5600X if CO is so high that cores kept below voltage threshold

Using no limit and +50pbo consumption is 84W max pn my 5600X since 4.7GHz don't use more and can't boost higher. With +200pbo I can get above 90W, but temp gets above 75C and I get some slight actual throttling.

 

[The King]

I have gotten closer to getting my first gen Ryzen to run at DDR4000! I'm current in the BIOS @4000...

 

[freeagent]

Man that thing is a freak

 

[RJARRRPCGP]

View attachment 237757

Well, here's mine with CTR 2.1 RC5:

At 4.8 GHz:

At 4.9 GHz in the PX profile:

I lowered the Vcore for the max frequency in the PX profile:

 

[RJARRRPCGP]

I honestly think that boost is unstable, even when I got a higher score, because multiple times, I found a Chromium tab crashed with an access violation error, including the tab suddenly becoming crashed with an access violation when merely scrolling down a YouTube page!

So, I discarded the settings I have and let CTR 2.1 RC5 run the diagnostic function again: (A golden sample, but the Vcore gets set so low that a Prime95 worker appears to fail, it's usually thread#8 that fails, looks like I found the weakest core to be the third core or the fourth core. IIRC, the date code is 2143, making this a very-late Vermeer revision!

Here's me starting over again, because of my past boost apparently making Chromium unstable:

 

[oobymach]

I tried CTR 2.1, ran the auto detect and it went up to 5.2ghz before it crashed my system and forced a reboot, any settings I tried there didn't run nearly as well as the same settings in bios though I did get the single core up to 4.9ghz the performance was lackluster.

On a positive note 1900mhz fclk is stable for me now with lower soc voltage, I can game for hours now without crashes @ 1.125v soc where 1.5v had issues

 

[RJARRRPCGP]

I tried CTR 2.1, ran the auto detect and it went up to 5.2ghz before it crashed my system and forced a reboot, any settings I tried there didn't run nearly as well as the same settings in bios though I did get the single core up to 4.9ghz the performance was lackluster.

Did you increase the setting for CPU load line calibration in the BIOS? For me, Windows never crashed out during diagnose.

The crashed Chromium sessions, may have just been because I lowered the max Vcore for the max frequency boost to 1.2-something volts. It's set higher now, hopefully, won't put a nick in the chip.

What were the core temps? If it was 70 C+, it's probably too high and caused it to be unstable.

 

[oobymach]

It might have been a temp limit that forced the restart idk the frequency was going up pretty quick but it didn't run for more than a few seconds before the hard restart.

 

[RJARRRPCGP]

It might have been a temp limit that forced the restart idk the frequency was going up pretty quick but it didn't run for more than a few seconds before the hard restart.

Sounds like a stop error to me. It likely was a fatal WHEA for "Cache Hierarchy Error". Check the event log. Raise the Vcore, in CTR, it's expressed in millivolts, just like with video cards.