Overclock.net banner

16121 - 16140 of 16305 Posts

·
Cheesebumps!!
Joined
·
2,420 Posts
Sorry for the off topic question, but I trust you guys.

I ordered a launch day 5900X from B&H, but assumed it would backorder and hustled to secure another one mid Nov.

I have only done some mild tuning on it, and it hasn't given me any issues after a BIOS update resolved the idle reboots.

B&H order just showed up, so I can either sell it unopened, or I could keep the new one and sell the used one.

Is there any evidence suggesting more recent CPUs perform better, or is it still just a lottery?

Sent from my Pixel 4 using Tapatalk
lottery..if you are curious you would have to try the new chip if it does any better in terms of IF Clock max range..higher should be better as it allows your Memory to run higher frequencies as well..(given no WHEA issues) else if you are happy then you can sell it for a higher price since its BNIB
 

·
Registered
Joined
·
192 Posts
Sorry for the off topic question, but I trust you guys.

I ordered a launch day 5900X from B&H, but assumed it would backorder and hustled to secure another one mid Nov.

I have only done some mild tuning on it, and it hasn't given me any issues after a BIOS update resolved the idle reboots.

B&H order just showed up, so I can either sell it unopened, or I could keep the new one and sell the used one.

Is there any evidence suggesting more recent CPUs perform better, or is it still just a lottery?

Sent from my Pixel 4 using Tapatalk
My RMA CPU from AMD is lot 2105 and old one was 2046. I can tell you for sure that the new one has lower boost clocks, this is how they "fixed" the idle reboots. So if you don't have stability problems with the old one, keep it, it gives better performance than the new ones. I have to use CO -10 to reach the same benchmark scores as the old one at stock. But yes, the new one is stable at CO -10 where the old one gave reboots due to a bad core that needed +10 to be stable. So in the end after tweaking you could say they reach the same performance point, but not at stock.

My whole RMA story here: Replaced 3950X with 5950X = WHEA and reboots
 

·
Registered
Joined
·
275 Posts
My RMA CPU from AMD is lot 2105 and old one was 2046. I can tell you for sure that the new one has lower boost clocks, this is how they "fixed" the idle reboots. So if you don't have stability problems with the old one, keep it, it gives better performance than the new ones. I have to use CO -10 to reach the same benchmark scores as the old one at stock. But yes, the new one is stable at CO -10 where the old one gave reboots due to a bad core that needed +10 to be stable. So in the end after tweaking you could say they reach the same performance point, but not at stock.

My whole RMA story here: Replaced 3950X with 5950X = WHEA and reboots
I 100% agree. I had a similar RMA experience with my 5900x and same issue with my #1 core needing +5 to +8 to pass single thread stress tests like prime95 and OCCT. Boost is a bit lower, but every core can handle some negative curve (ranges for -8 to -23). I always want a stable system at stock and then it's up to me to break it by tweaking :) . My max IF without WHEA warnings was and still is 1900 on both chips. First chip was produced in November 2020 and the new one is early January for reference.
 

·
Registered
Joined
·
356 Posts
Nice! Let us know how it goes.
Did you try the Dynamic OC?

I've updated my niece B550 TUF Gaming Plus over Easter break and I've to say is even more rock solid, BIOS is very mature.
Her 3600XT now can boot at FCLK 1800.

I love the original Noctua colors :D still have an NH-U14 on one old rig, more than 10 years?



My understanding is previous AGESA was only a partial fix, 1.2.0.2 fully fixes the USB issues (so they say).
Nah I didn't bother with the Dynamic OC as I like the software approach with CTR.
Audio issues are not gone btw, not even with 1.2.0.2. Already got a ticket open with Microsoft and AMD.
 

·
Registered
Joined
·
981 Posts
Anybody else experiencing the following issue, I'm on the latest BIOS Aorus Master: When I set curve optimiser to all core and the offset to whatever all is fine and I boost all core to 4.5+ but when I set curve optimiser to per core offsets even to the same value I use in all core it only boosts to about 4GHz, thinking it's a bug?

Edit: Was an issue in F33i BIOS, seems to be fixed in F33j.
 

·
Registered
Joined
·
1 Posts
Hey Guys,

I have come across some unexpected moneys these days :)D), and I'm looking to buy 64 or maybe even 128 GB RAM (although that may be a tall order). My requirements for it are very simple: they need to be 3600 CL18, and they need to run by activating XMP and doing not much else (I have not the time or interest to start manually tweaking timings). I have an X570 Aorus Master rev 1.2, and a 3950x (which I will maybe upgrade to 5950X or any Zen3+ thing that AMD cooks up on the AM4 platform, if that ever comes and is worth upgrading, of course)

My question is: do you guys have some experience with such high memory capacities? Any specific model recommendation? Any specific forum type resource where I might do some homework on these?

The QVL for the board seems to stop at 32GB capacities, so...

Thanks
I'm running 2xGSkill RipJaws V F4-3600C16-32GVK (so 64GB) with the same MB and a 5800X. Not got to tuning it yet, but it runs perfectly with the XMP config @ 16-22-22-42.

Hope this helps :cool:
 

·
Untitled.jpg
Joined
·
1,011 Posts
anyone used f33i on a Pro Wifi with a 5000 chip here yet?
Is it working properly?
My 5900X is due to arrive any day now, wondering if its worth updating from F32 which is working perfectly with my 3700X.
 

·
Registered
Joined
·
63 Posts
I just got my 5950x about 2 weeks ago. Now using F33i on a X570 Xtreme. I think I am finally stable after running many benchmarks as well as mprime and stressapptest which before caused machine check errors pretty quickly. Shadow of the Tomb Raider is also running fine which caused restarts before even though corecylcer/mprime/occt seemed stable.
I now settled on curve optimizer values
CCD1 0/-10/-10/0/-10/-10/-5/0
CCD2 -30
Compared to "No PBO, No XMP" I gain about 12% mixed performance across all those compile benchmarks (geekbench +3% ST, +30% MT)
Compared to "No PBO, XMP 3800cl14" I gain about 5% mixed performance (geekbench +0% ST, +10% MT).

I am really happy I have XMP 3800cl14/IF 1900 stable without fiddling too much. I set SOC Voltage to 1125, VDDP 900, VDDG CCD 950, VDDG IOD 1050. I could probably go lower on those and/or tighten XMP timings:
20210430-102154_swappy.png

Only have one issue regarding onboard Bluetooth which has very low signal strength. I have to move my headphones almost next to my case to not have sound interrupts.
 

·
Registered
Joined
·
1,587 Posts
Only have one issue regarding onboard Bluetooth which has very low signal strength. I have to move my headphones almost next to my case to not have sound interrupts.
The Bluetooth uses the Antenna for the WIFI, you need them installed for it to work properly.
You can then walk around the house without issue unless you have metal walls.
 

·
Registered
Joined
·
63 Posts
The Bluetooth uses the Antenna for the WIFI, you need them installed for it to work properly.
You can then walk around the house without issue unless you have metal walls.
Oh wow now I feel stupid. Thank you, that was it. Thought those antennas were for WiFi only. And you were not exaggerating with the range.
 

·
Registered
Joined
·
103 Posts
After 96+ hours of testing.

The bios F33i for the gigabyte X570 Xtreme REV 1.0 is ... error/bug free. (with a 3900X and a 5600X) (FINALLY ... xD)

No more bug with USB (before, i had random disconnection of keyboard/mouse and 2 external ssd/hdd) or sound-card, no more whea errors (with ram at 3600-3666-3733-3800)

With the F32 bios, i had WHEA errors, even at 3600mhz and 1800 FCLK ...

So far, it's the best bios released by Gigabyte/AMD.


The only "problem" that i have since the begining with ALL the bios :

ProcODT is wrong by default, it's impossible to boot with ram of 3600+ if i set ProcODT to default, with 4 dimm
. (i'm using 4*8GB 3200C14 (who are working at 3800C14 easily))
ProcODT at default value and ram of 3600-3666-3733-3800 and ~1.48V = no boot.
With a ProcODT of 60ohm and below (53.3, 48, 43.6, 40, 36.9) the pc is booting fine.

I'm using 43.6ohm with 4 dimm a 3600-3800C14 and everyting is fine. (is there any benefit to lower it more ?)
 
  • Rep+
Reactions: Nighthog

·
Untitled.jpg
Joined
·
1,011 Posts

·
Registered
Joined
·
1,587 Posts
After 96+ hours of testing.

The bios F33i for the gigabyte X570 Xtreme REV 1.0 is ... error/bug free. (with a 3900X and a 5600X) (FINALLY ... xD)

No more bug with USB (before, i had random disconnection of keyboard/mouse and 2 external ssd/hdd) or sound-card, no more whea errors (with ram at 3600-3666-3733-3800)

With the F32 bios, i had WHEA errors, even at 3600mhz and 1800 FCLK ...

So far, it's the best bios released by Gigabyte/AMD.


The only "problem" that i have since the begining with ALL the bios :

ProcODT is wrong by default, it's impossible to boot with ram of 3600+ if i set ProcODT to default, with 4 dimm
. (i'm using 4*8GB 3200C14 (who are working at 3800C14 easily))
ProcODT at default value and ram of 3600-3666-3733-3800 and ~1.48V = no boot.
With a ProcODT of 60ohm and below (53.3, 48, 43.6, 40, 36.9) the pc is booting fine.

I'm using 43.6ohm with 4 dimm a 3600-3800C14 and everyting is fine. (is there any benefit to lower it more ?)
Yeah F33i is working nicely thus far, no bug/issues encountered that are crippling.
Can chime in and agree on default procODT settings are a bit high for defaults [AUTO] with these new [F32-F33] releases, but choosing it manually is easy enough anyway.

I'm still a little disappointed PBO is locked for PRO processors & that they locked down AUTO-OC to +200MHz maximum for the new AGESA. (no more +500)
That extra +500 was useful and could be used to gain extra performance.

I'm quite sour AMD keeps locking away ways to gain extra performance with PBO. It's tightly controlled and "locked" to what AMD wants it to be. Extra's not available for long.
 

·
Registered
Joined
·
106 Posts
BIOS F33j does not fix the USB system freeze issues associated with X570 Aorus Masters. And honestly this crap is getting quite annoying. I use a USB switcher to frequently switch my mouse and keyboard from my personal and work PCs. This has been a pain ever since I built my system! So Gigabyte send F33j right back to the beta bin because it indeed IS NOT fixed contrary to your description statement.
 

·
Registered
Joined
·
526 Posts
As rumours thicken regarding a Zen3+ cancellation, I jumped the waiting boat and ordered a 5900x...prepare for a lot of tears coming...my silicon quality luck was always average at best. :p
 

·
Registered
Joined
·
343 Posts
F33j for Master on both systems since it was first release on Tweaktown.... No WHEA, no other issues of note other than minor item display problem in BIOS.

Maybe is coincidence but for some reason my older 3900X was never boosting past about 4550 ever since the very old Aegesa ABBA release... but on this BIOS it has touched 4650 quite a few times. Whether that is BIOS or moon/planets/stars aligning who knows 😆
 

·
Registered
Joined
·
112 Posts
Been running F33j for a few days but it’s been great so far, no real issues to speak of and the BIOS UI changes are very welcome and make a lot more sense compared to the previous menu’s.

Keep it up GB!
 

·
Registered
Joined
·
38 Posts
Also been running F33j for a few days. Running great! Both F33i and F33j is going well with my USB BT soundcard. Have had some problems before, but none seens F33i. And I also like the new UI in the BIOS, im just missing the VDDG CCD/IOD settings. I guess the VDDG sets both CCD/IOD at the same voltage?
Still can't post 1900mhz FCLK tho, and 1933mhz+ post, but gives WHEA.
 

·
Registered
Joined
·
320 Posts
I see f33i being the latest for ultra. Does that version comes with the same changes as the f33j on master?
 
16121 - 16140 of 16305 Posts
Top