So far no issues I can report on 3065, AGESA 1.2.0.2b.
I only started using this board a few days ago and am new to AM5, but it has been continuously used, even overnight.
My setup isn't typical. I didn't wish to pull apart my daily rig this close to Crimbo. So opted to create a tinkering setup .
Ryzen 5 9600X (Batch BY 2435PGE) (Waiting on a 9800X3D to show up)
TUF Gaming X670E Plus WiFi
Patriot Viper Venom DDR5 6200C40 2x16GB (Single rank, SK Hynix M die AFAIK)
Sapphire Radeon HD 5850 Toxic 1GB
Crucial MX500 2TB (SATA)
ThermalRight Archon IB-E X2
CoolerMaster V650 Rig photo
Currently using W10 Pro, I used ASUS TurboV Core v1.05.03_beta (download link in OP here), Ryzen Master and ZenTimings to get all the voltages at stock for "Auto".
Manually set voltages to stock as didn't want a UEFI "Auto rule" to bump them as tweaked RAM.
RAM had EXPO profile of 6000C36, set EXPO II in UEFI as didn't want ASUS tweaked EXPO, but the one from SPD.
Set tCL 30 tRC 112, then testing showed needed SOC 1.05V, VDDP 0.850V, VDD & VDDQ 1.3V, VDDG CCD & IOD left at 900mV.
Dropping to tCL 28 has been a little hit and miss for stability. I had passed RT ~1hr, then a rerun failed sooner. Playing with that again today. Not done any PBO/CO tweaking yet.
In the past I have seen beta UEFI become stable release. Even when checking with hex editor only perhaps a version string changed and checksum. Only very occasionally a beta has been bad.
Also got some Patriot Viper Venom DDR5 7000C32 kit. Not tried it yet.
*** edit ***
ASUS TurboV Core version added to post and correct link.
I've just edited post 25 in regard to ASUS TurboV Core. I had been using an old image of OS on TUF. This didn't have ASUS TurboV Core v1.10.20, but v1.05.03_beta. Today when I did a fresh OS to test some multiboot setup, I found out which version works with TUF X670E.
ASUS TurboV Core v1.05.03_beta gives all info on voltages on TUF X670E, link to screenie, where as v1.10.20 is only this, link to screenie.
I have sent a private message to Shamino on ROG forum to ask if there is a new version of TurboV Core, don't know if will get a reply.
ASUS TUF GAMING X670E-PLUS is an AMD AM5 (Ryzen™ 7000) ATX Gaming motherboard with 16 teamed power stages, PCIe® 5.0, DDR5 RAM support, four M.2 slots, USB 3.2 Gen 2x2 Type-C®, and Aura Sync RGB lighting.
ASUS TUF GAMING X670E-PLUS WIFI is an AMD AM5 (Ryzen™ 7000) ATX Gaming motherboard with 16 teamed power stages, PCIe® 5.0, DDR5 RAM support, four M.2 slots, USB 3.2 Gen 2x2 Type-C®, and Aura Sync RGB lighting.
No worries , my basic 6000C30 profile passed 16k% in Kahru RAM Test, did also Y-Cruncher v0.8.5.9543 ~1.5hrs. Gonna try the basic 6000C28.
I can understand if it's not broken don't fix. But just be aware the UEFI (BIOS) is a combo package of AMD and ODM (ASUS) FW, fixes/patches/security updates from both sides. The change logs rarely state all changes.
This has been discussed for so many times I've already lost the count.
Unless you want the software, firmware and hardware engineers to spend their time in testing and then "translating" the made changes to something associatable with the real-world behavior, instead of doing their actual jobs, then most likely not.
I get that it would be nice to know how the behavior might change after a new AGESA version however, in cases where there has been a change to the stock behavior of the CPUs, AMD has announced that (e.g., fTPM fix in 1.2.0.7).
As I've said previously: The change logs that exist consist of comments made by HW / SW engineers and designers. For instance: "[FWDEV-1234] Poll for heartbeat until PMFW is ready" might be a valid change listed in the change log and it would obviously make perfect sense to people who work on it.
The issue is, that there is no way for the end-user to associate such a change to some phenomenon / behavior that can actually be observed (correctly, especially). In my opinion, it would only benefit the competition or other parties interested in digging up any filth or developing potential exploits.
Source link , The Stilt aka Roger Tolppola, known for overclocking, but has "connections" to industry (his RAM profiles were in AM4 Crosshair UEFIs) and really good technical knowledge from what I have read of his posts and contact on various forums in threads.
ASUS TUF GAMING X670E-PLUS is an AMD AM5 (Ryzen™ 7000) ATX Gaming motherboard with 16 teamed power stages, PCIe® 5.0, DDR5 RAM support, four M.2 slots, USB 3.2 Gen 2x2 Type-C®, and Aura Sync RGB lighting.
ASUS TUF GAMING X670E-PLUS is an AMD AM5 (Ryzen™ 7000) ATX Gaming motherboard with 16 teamed power stages, PCIe® 5.0, DDR5 RAM support, four M.2 slots, USB 3.2 Gen 2x2 Type-C®, and Aura Sync RGB lighting.
*Improves performance for 2CCD Raphael and Granite Ridge CPUs
AGESA 1.2.0.3 has an option App Compatibility Database, I believe benefits 2CCD CPU.
Latest beta UEFI 3202 with AGESA 1.2.0.3a was posted on ROG forum, link. This has SMU FW update for Granite Ridge CPU which UEFI 3201 Beta (AGESA 1.2.0.3) currently on TUF page doesn't have.
"1.Updated AGESA to ComboAM5 PI 1.2.0.3. Please also update the chipset driver to version 7.01.08.129 or newer to enhance gaming performance for select games.
2.Improved system performance and fixed the PeCoffLoader memory overflow issue for security.
3.Addressed AMD Microcode Signature Verification Vulnerability.
So do we enable app compatability or leave on auto?
Unfortunately I don't have a 2CCD CPU, only have 1CCD CPU. From various forums I haunt, I have not seen information what the setting defaults to and what is optimal, try it and see and report back .
I had seen a post on a forum, I can't find it at present to link. IIRC someone looked further into the new driver and it contained references to some games/apps, so seems optimised for 2CDD CPUs.
The new driver is defo needed regardless of what CPU you have, as when a UEFI has AGESA 1.2.0.3 or newer, an unknown device will appear in device manager. Installing new driver solves it, old driver doesn't. The device isn't "new", if that make sense, forgotten what it was when I applied driver.
The UEFI you are stating is AGESA 1.2.0.3, will have option I posted, will need new driver. But UEFI on ROG forum is newer beta, has AGESA 1.2.0.3a.
9000 series SMU FW is the same on UEFI 3203, so is 7000 series (which didn't get an update in recent few UEFI I looked at).
Compared UEFI settings dumps between UEFI 3202 and 3203 and no new settings to play with.
Usually I will set a base profile which has minimal tweaks and then see what the Auto values default to in OS with apps we as users have access to.
Base Auto Profile had these changes:-
Ai Overclock Tuner [Manual]
Performance Bias [None]
Clock Spread Spectrum [Disabled]
Above 4G Decoding [Disabled] (as my test system does not have GPU with REBAR support)
Realtek LAN Controller [Disabled]
When system is in sleep, hibernate or soft off states [Stealth Mode]
Serial Port [Disabled]
Launch CSM [Enabled]
Setup Mode [Advanced Mode]
Download & Install ARMOURY CRATE app [Disabled]
No changes in things I looked at for Auto values, link.
My assumption is whatever tweaks/fixes are in UEFI 3203 vs 3202 are on ASUS FW side and not AMD AGESA related, as AGESA has no fresh patch, no SMU FW update, no CPU microcode update.
My 6000C28 1:1 profile worked as determined on UEFI 3065. Not had time to see if 6200C28 1:1 and 6400C28 1:1 works.
CO per core profile with PBO +200MHz works as on prior UEFI, 9600X in use, have a 9700X and 9800X3D to play with, but not had time to use.
So far the TUF X670E has also been fulfilling my needs for tinkering .
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
wccftech.com
www.asus.com
