Scumbag ASUS: Overvolting CPUs & Screwing the Customer

AMD's X3D designs are more sensitive to overvolting and overclocking than Intel and non-X3D designs. We've known this since the first X3D chips came out.

Pretty sure the motherboard vendors have been playing fast and loose with the voltages for a LONG time on both AMD and Intel platforms, and just continued doing so for X3D chips, despite it being well known that they really can't tolerate being run out of spec by much.

Is that AMD's fault for designing a CPU with all the benefits of an X3D chip despite it being more sensitive? I'd argue no. New solutions are going to have varying degrees of voltage sensitivity.

The truth is that X3D chips absolutely need to be run in spec, and Asus and Gigabyte just didn't give a fuck. It has always worked before, so why should we care now?

The only solution here is to adhere to the spec like it is a goddamn religion, and it is clear the board makers haven't.
Oh, trust me, I know. What I'm saying it that ultimately AMD should have known and put some hard limits in both hardware and in writing to their partners. But they were too busy enjoying the lime light of amazeballs benchmark numbers and techtubers throating them hard for making something exciting and new.
 
Oh, trust me, I know. What I'm saying it that ultimately AMD should have known and put some hard limits in both hardware and in writing to their partners. But they were too busy enjoying the lime light of amazeballs benchmark numbers and techtubers throating them hard for making something exciting and new.
"should have known" how/why?

Vendors have boosted the voltages but always stayed within reasonable limits to not burn the CPU up by testing and not boosting to the sky. Now the MB vendors fucked up and burned up some CPUs so yeah - AMD is locking shit down and replacing the CPUs.

Seems like a reasonable response. AMD could have denied replacement and REALLY throwing the MB vendors (as well as themselves) under the bus. AMD made the right move here.

Please continue your tirade.
 
Last edited:
"should have known" how/why?

Vendors have boosted the voltages but always stayed within reasonable limits to not burn the CPU up by testing and not boosting to the sky. Now the MB vendors fucked up and burned up some CPUs so yeah - AMD is locking shit down and replacing the CPUs.

Seems like a reasonable response. AMD could have denied replacement and REALLY throwing the MB vendors (as well as themselves) under the bus. AMD made the right move here.

Please continue your tirade.
Not going to bother debating with you, you have made it abundantly obvious that you will die on the "defend AMD at all cost" hill.


Screenshot_20230302_144225_Samsung Internet.jpg
 
The next person to address the poster rather than the poster's argument will get excluded from this conversation.

The REPORT post button is your friend.
 
Last edited:
I stopped buying Asus motherboards long time ago because the quality of their motherboards they were releasing. I have found MSI board to be much better built and priced.

Now I will say, when I did buy Asus products years ago I never once had any issues with RMA. That being said I do think Asus products are VERY overpriced.
I had issues with the quality and the second RMA. Same problem. Board just failed. ASUS wouldn't cover it. It was a microATX ROG board. Had solid experiences with an MSI board overclocking.

Going against my own advice and using an ASUS board, because it's cheap and I am not overclocking it much.
 
I wonder if the OCP on the ASUS mobo was just configured to allow the highest expected current draw from whatever the most power hungry CPU that fits the socket? It should adjust with the installed CPU I would think.

As far as I know AMD is warrantying the CPU's and ASUS is warrantying the motherboards. But both have some responsibility in allowing this to occur.

If you calculate the Amps that these CPU's pull, it's pretty nuts. 125W TDP CPU at 1.25 volts = 100 Amps... And 125W TDP at the 1.1v which if I found the correct specification, is the default voltage, current is even higher at 113.6 Amps! Overvolting actually allows it to pull less current to reach the same total power.

There were failure reports from users who were running at stock too. Ultimately it's such a high current situation, that the overcurrent protection 'system' might need completely redesigned.
 
I wonder if the OCP on the ASUS mobo was just configured to allow the highest expected current draw from whatever the most power hungry CPU that fits the socket? It should adjust with the installed CPU I would think.

As far as I know AMD is warrantying the CPU's and ASUS is warrantying the motherboards. But both have some responsibility in allowing this to occur.

If you calculate the Amps that these CPU's pull, it's pretty nuts. 125W TDP CPU at 1.25 volts = 100 Amps... And 125W TDP at the 1.1v which if I found the correct specification, is the default voltage, current is even higher at 113.6 Amps! Overvolting actually allows it to pull less current to reach the same total power.

There were failure reports from users who were running at stock too. Ultimately it's such a high current situation, that the overcurrent protection 'system' might need completely redesigned.
These AMD cpus are DYNAMIC with the power, as most modern CPUs.

7000 CPU Power Envelope = 170w

Package power is 230w

"AMD would like to issue a correction to the socket power and TDP limits of the upcoming AMD Socket AM5. AMD Socket AM5 supports up to a 170 W TDP with a PPT up to 230 W. TDP*1.35 is the standard calculation for TDP v. PPT for AMD sockets in the "Zen" era, and the new 170 W TDP group is no exception (170*1.35=229.5).

A CPU is a collection of logic gates best modeled with capacitance and frequency.

P = (propto) *CV^2f

Where P is power C is Capacitance (total) and f is frequency.

propto or "constant of proportionality" is typically found through experimentation.

There are some more details that can affect the calculation but (propto) *CV^2f should get close.
 
Last edited:
That sounds like the maximum that the socket specification supports. And of course they pull less at idle, and different workloads will have different power needs.

It's still insane amounts of current. That likely makes having an "Over Current Protection" system, more difficult to design properly. Probably needs to monitor multiple sections within each chiplet, because a high drain that will cause damage to one section could easily stay under the max TDP overall.
That level of "detection" would probably require monitoring units spread through-out the silicon. Could also be possible thru multiple points in the socket.

i.e. I really do not think blame can squarely be placed on a single party, which seemed to be all of the thread arguing.
 
That sounds like the maximum that the socket specification supports. And of course they pull less at idle, and different workloads will have different power needs.

It's still insane amounts of current. That likely makes having an "Over Current Protection" system, more difficult to design properly. Probably needs to monitor multiple sections within each chiplet, because a high drain that will cause damage to one section could easily stay under the max TDP overall.
That level of "detection" would probably require monitoring units spread through-out the silicon. Could also be possible thru multiple points in the socket.

i.e. I really do not think blame can squarely be placed on a single party, which seemed to be all of the thread arguing.

Power dissipation of the cpu package defines the package power. The power delivery envelope is a product of the motherboard.

You want to hear about INSANE power, Intel wants to talk to you.
 
So glad I went with the non X, runs so cool.
The X is easily put into eco mode @ 65w and becomes a non-x. I'd rather have the flexibility of the X. Non-X are good for budget boards that have no hope of ever supplying stable power anywhere near the 7000's power envelope of 170w.
 
Wraith Spire or?
ngl, i do like the aesthetics of that Wraith Spire tbh
That does look kind of cool, and I'm sure is more than sufficient cooling for any non-x 7000 series CPU too. But the question is between X and non-X it's usually like only $20 difference in price, and hell I've seen discounts on the X version that even bring it cheaper.
 
  • Like
Reactions: erek
like this
That does look kind of cool, and I'm sure is more than sufficient cooling for any non-x 7000 series CPU too. But the question is between X and non-X it's usually like only $20 difference in price, and hell I've seen discounts on the X version that even bring it cheaper.
aren't the X3Ds even hotter for some reason beyond the standard Xs?
 
aren't the X3Ds even hotter for some reason beyond the standard Xs?
The V-cache is stacked on top of the cores, which increases the distance between the IHS and the cores. The distance decreases the heat transference from the cores to the IHS to the cooler.
 
I don’t anticipate having any issues on the latest bios with my AM5 setup regarding this. However If it dies it dies— I’ll build something different. Out of curiosity has anyone heard of people having his happen and being denied a warranty claim?
 
ASUS' actions relating to the Exploding Ryzen CPU debacle are disgraceful and abrasive to the trust that the brand has earned. ASUS has demonstrated clearly it wishes to not only avoid supporting users, but actively engineers ways to abandon them. ASUS' updates haven't even fixed the problems, yet they posture as if they have while simultaneously suggesting that users 'just run defaults' on their $700 motherboards, as if that makes any sense whatsoever. So, to accommodate ASUS' request, we ran defaults and re-benchmarked the Ryzen 7000 series. It sucks. Big surprise. They also don't support their own BIOSes for the ASUS ROG boards.



I would also hold AMD partially responsible for this debacle. They have enough power, and certainly obligation, to influence their partners' behavior and practices when it directly compromises AMD's reputation and business. Intel is, for better or worse, much better in that regard. Asus and its customer relation is standard MO for them.

We joined Leo of Kitguru to talk about ASUS' reviewer relationships, its advertising ethics, warranty behaviors, and what we need to see from ASUS to know it's improving. This is just meant to be a casual discussion between two critical technical outlets about where ASUS is pointed.
 
We joined Leo of Kitguru to talk about ASUS' reviewer relationships, its advertising ethics, warranty behaviors, and what we need to see from ASUS to know it's improving. This is just meant to be a casual discussion between two critical technical outlets about where ASUS is pointed.

It's going to take a while for Asus to get some respect back in the community after this. Leo from Kitguru does not hide his emotions too well it seems and this video shows how irritated at Asus he really is right now.
 
Has anyone found this extending to laptops? I see some pretty high voltages on zen3+ over 1.4v....
 
  • Like
Reactions: erek
like this
1.4v on the cpu can be normal 1.4v on the soc is not. which are you seeing?
I haven't looked closely at the SOC just the voltage the cores are requesting. See them calling for as high as 1.45v and they seem to get a little over 1.4v

I'll keep an eye on the SOC voltages going forward...
 
I haven't looked closely at the SOC just the voltage the cores are requesting. See them calling for as high as 1.45v and they seem to get a little over 1.4v

I'll keep an eye on the SOC voltages going forward...
that sounds normal. pretty sure this is an am5 only issue though.
 
7800X3D + X670E Tomahawk
OP says the system never successfully powered on. Wonder if it's a returned CPU.

https://www.reddit.com/r/pcmasterra...d_help_with_my_new_build_possibly_broken_cpu/

The situation: I've decided to build my own pc and ordered the parts. After everything's installed I noticed that nothing was powering on. I tried to find out what the faulty component is by disassembling the build piece by piece. I noticed that I can't start the pc and I'll get no power if the cpu is installed in the motherboard. When I first removed the cpu from the socket, I noticed there was some black markings in the back of the cpu. I suspected that it might be thermal compound, so I used 99% isopropyl alcohol and toilet paper to clean it. You can see how it looked before cleaning it in the picture below. After cleaning it, the back of the cpu looks like it should and there are no markings or anything. But, even after the cleaning, I get no power when I install the cpu in the socket.

So, I removed the cpu and wanted to test the motherboard if it's working. I plugged in nothing but the motherboard power cable. The motherboard would get power this way and I could install the latest bios using msi bios flash.

To give complete info, the cpu is a 7800x3d and the motherboard is a msi x670e tomahawk. I have flashed the latest beta bios, but I still get no power when the cpu is installed and nothing else. I touched the back of the cpu while cleaning as well, which I learned was a mistake later.

No power in this context = no fans, no leds, plain nothing

7800X3D_melt.png
 
Last edited:
ASUS' actions relating to the Exploding Ryzen CPU debacle are disgraceful and abrasive to the trust that the brand has earned. ASUS has demonstrated clearly it wishes to not only avoid supporting users, but actively engineers ways to abandon them. ASUS' updates haven't even fixed the problems, yet they posture as if they have while simultaneously suggesting that users 'just run defaults' on their $700 motherboards, as if that makes any sense whatsoever. So, to accommodate ASUS' request, we ran defaults and re-benchmarked the Ryzen 7000 series. It sucks. Big surprise. They also don't support their own BIOSes for the ASUS ROG boards.



I would also hold AMD partially responsible for this debacle. They have enough power, and certainly obligation, to influence their partners' behavior and practices when it directly compromises AMD's reputation and business. Intel is, for better or worse, much better in that regard. Asus and its customer relation is standard MO for them.

GIGABYTE Announces Enhanced Security Measures for Motherboard Products in UEFI BIOS Firmware​

by Hilbert Hagedoorn on: 06/06/2023 08:36 AM | source: | 3 comment(s)

The company that somehow always has to put out fires (sometimes literally), GIGABYTE, has announced security enhancements for UEFI BIOS firmware in their motherboard products.
Based on information provided by third-party research institutions, GIGABYTE has strengthened the security of UEFI BIOS. Revised measures have been implemented to conduct stricter security checks during the OS boot process. These enhancements are designed to detect and prevent the activities of malicious programs as much as possible.
The verification process for downloaded update files from remote servers has been bolstered. By introducing electronic signature verification, the integrity and legitimacy of the downloaded file content are ensured, thus preventing the insertion of malicious code. Additionally, privilege access limitations have been activated to enable standard encryption verification of remote server certificates. This restricts the download of update files, ensuring they are sourced only from servers with trusted certificates.
These security measures apply to motherboards with the "APP Center Download & Install" feature embedded within the UEFI BIOS. Currently, updates to UEFI BIOS for applicable motherboard products are being carried out, and they will be progressively released on each product's download page. Until the UEFI BIOS is updated, users can enhance security by disabling the "APP Center Download & Install" feature in the UEFI BIOS settings.
 
That's what we get for replacing BIOS with a shitty successor.
 

GIGABYTE Announces Enhanced Security Measures for Motherboard Products in UEFI BIOS Firmware​

by Hilbert Hagedoorn on: 06/06/2023 08:36 AM | source: | 3 comment(s)

The company that somehow always has to put out fires (sometimes literally), GIGABYTE, has announced security enhancements for UEFI BIOS firmware in their motherboard products.
Based on information provided by third-party research institutions, GIGABYTE has strengthened the security of UEFI BIOS. Revised measures have been implemented to conduct stricter security checks during the OS boot process. These enhancements are designed to detect and prevent the activities of malicious programs as much as possible.
The verification process for downloaded update files from remote servers has been bolstered. By introducing electronic signature verification, the integrity and legitimacy of the downloaded file content are ensured, thus preventing the insertion of malicious code. Additionally, privilege access limitations have been activated to enable standard encryption verification of remote server certificates. This restricts the download of update files, ensuring they are sourced only from servers with trusted certificates.
These security measures apply to motherboards with the "APP Center Download & Install" feature embedded within the UEFI BIOS. Currently, updates to UEFI BIOS for applicable motherboard products are being carried out, and they will be progressively released on each product's download page. Until the UEFI BIOS is updated, users can enhance security by disabling the "APP Center Download & Install" feature in the UEFI BIOS settings.
Motherboard makers should leave their crapware utilities where they belong: on optical discs, as those carry a low risk of installation. The UEFI specification already defines methods for updating firmware via "capsules". If Gigabyte wants to make automatic firmware updates available to end users, it should use the existing distribution channels such as Microsoft/Windows Update and LFVS (Linux). I don't know whether any of the "enthusiast" consumer motherboard manufacturers currently offer updates that way.
 
Last edited:
Motherboard makers should leave their crapware utilities where they belong: on optical discs, as those carry a low risk of installation. The UEFI specification already defines methods for updating firmware via "capsules". If Gigabyte wants to make automatic firmware updates available to end users, it should use the existing distribution channels such as Microsoft/Windows Update and LFVS (Linux). I don't know whether any of the "enthusiast" consumer motherboard manufacturers currently offer updates that way.
Have you seen the bugs with windows updates doing that?? Or the security issues that have happened?
 
Have you seen the bugs with windows updates doing that?? Or the security issues that have happened?
What I'm suggesting is that a standard, OS-agnostic approach to handling firmware updates is preferable to the alternative of motherboard vendors attempting to reinvent the wheel by creating proprietary update mechanisms that are more buggy and less secure, such as what Gigabyte did. I don't see how simply making the updates available through established distribution channels would create new bugs or insecurities. It would do the opposite by making it easier for users to patch security vulnerabilities. I'd be pleased if more hardware vendors published firmware updates to LVFS. They can be applied at the user's discretion without any form of automation necessary.

I'm not sure what Window Update bug(s) you're referring to as I seldom (as in, not really at all) use Windows anymore, but according to Microsoft's documentation, "system firmware updates for UEFI-based systems [are] deployed as device driver packages", and I think that disabling automatic driver updates should prevent them from being installed without the user's knowledge or consent (assuming that's the issue). That seems less than ideal, but that's a Windows issue. Depending on firmware updates to remain unavailable from Windows Update doesn't seem like a viable long-term strategy. Motherboards that support UEFI capsule updates should have an option to disable them in the firmware settings, so it should also be possible to block them at the firmware level.

It's critical that end users retain explicit control over the installation of firmware updates. Motherboards and OSes need to respect that by providing options for users to define their own update policies. In terms of security, the track record of gaming-type motherboard makers troubles me much more than Microsoft's. The issues I have with Microsoft are different, but that's another topic.
 
Back
Top