Ryzen 7000X3d is burning out.

[Vishera]

That's why i am never a first adopter.

I typically buy products that have been around for at least a year, which also over time become cheaper and cheaper.

That's why i went with a Ryzen 7 5700X.

[Vishera]

47 minutes ago, HenrySalayne said:

Yes, you are absolutely right...

AMD brought this upon themselves - and i will explain why.

 

47 minutes ago, HenrySalayne said:

6800X dies cracking - only a dozen people affected, issue has nothing to do with AMD in particular, wordwide outrage

While what happened to those cards  is not AMDs fault, the reputation of their drivers is their fault.

Since i began building PCs i always went with ATi\AMD until severe driver issues began in March\April 2016 with my then six months old R9 380 which lingered for 2-3 years!

 

Even if AMD did fix their drivers, people are still skeptic - Especially those who were burned by them in the past.

 

47 minutes ago, HenrySalayne said:

7900XT(X) issues with the stock cooler - number of affected people unclear, issue resolved within a month, worldwide outrage

Anyone would be outraged if they purchased a defective product and the manufacturer declined their RMA requests.

It's only after the issue was covered by the media that AMD made changes.

 

47 minutes ago, HenrySalayne said:

AM5 boards are expensive - worldwide outrage

That's because AM5 boards are 2-3 times more expensive than AM4 boards...

In 2019 i bought a GIGABYTE B450 Gaming X for $70 and you could get cheaper B450 boards you for $50.

but with AM5 there is no chance to get a B650 board for less than $150.

 

Outrage justified.

 

47 minutes ago, HenrySalayne said:

Z790 boards are expensive - ... nothing.

People are outraged over current pricing of motherboards in general...

[HenrySalayne]

17 minutes ago, Vishera said:

AMD brought this upon themselves - and i will explain why.

Nobody cares...

These were just three recent examples to show that AMD get's negative reactions again and again and again and is not "treated lenient".

5 hours ago, Arika S said:

damn, people are are lenient when it comes to AMD...

5 hours ago, GoodBytes said:

People are waiting for info

5 hours ago, Arika S said:

exactly, something only AMD seems to be afforded in much more often than a certain blue themed company.

 

19 minutes ago, Vishera said:

Anyone would be outraged if they purchased a defective product and the manufacturer declined their RMA requests.

Yes, and rightfully though. And it's a good thing to make some noise to make an issue heard.

 

But after AMD acknowledged the issue people who were not affected still jumped on the outrage bandwagon and then these not affected people jumped onto the "there is no supply" outrage bandwagon - which is especially annoying for the affected people but certainly nobody's fault.

A screaming and screeching crowd can be a tool but it gets unnecessary and annoying very fast.

 

The "blister" on the PCB in conjunction with dying and overheating CPUs is now a known issue.

AMD is looking into it, Asus (and probably other mobo manufacturers) have taken precautions with changes to the BIOS, GamersNexus has acquired a CPU and sent it to a lab. The number of affected people is once again small.

We can stop the war drum and wait for the results.

[leadeater]

1 hour ago, Vishera said:

People are outraged over current pricing of motherboards in general...

Then why direct it mostly at AMD? Seems a little unbalanced. 

 

But put it this way, currently for a lot of people or recommendations is that AMD is the better option or would/was the better option either performance wise or value wise but now that board costs have gone up to be in compliance with PCIe 5.0 and DDR5 requirements the lowest price points simply cannot be as low.

 

Where Intel is able to offer lower minimum priced motherboards comes from not having any PCIe 5.0 or DDR5 support.

 

Not that minimum possible price options are that good to look at, sometimes those can just be downright bad options and are only able to be that cheap due to compromises that isn't worth the cost reduction for the consumer.

 

Price creep is certainly annoying but it's good to keep that grounded with it's not just one brand, company, product type etc. Problems don't become worse because you can compound different issues together and say an issue is bad because of this other one existing as well. That's more emotional reasoning than logical reasoning.

 

1 hour ago, Vishera said:

Even if AMD did fix their drivers, people are still skeptic - Especially those who were burned by them in the past.

Yet people are and have been WAY more forgiving of Nvidia simply because of brand reputation and willingness to put aside past problems to justify buying the best thing they can with their money. Not that it is a problem, it's just only afforded to Nvidia 

 

Nvidia changed the saying from "Can do no wrong" to "Nobody cares they do wrong".

 

At the end of the day people will buy what they want, reasoning doesn't even matter at all realistically. I just wish the whole elaborate reasoning/justifications wouldn't get put out there, it is actually good enough that a person felt whatever it was is the right choice for them. I like many could have owned a string of generations of AMD GPUs without ever having an issue, someone else could have nothing but problems. Someone also could have gotten an Nvidia GPU that caught on fire then after that a GPU with transient power spikes and system hard shutdowns. Literally anything is possible with the millions of GPU consumers in the world.

[Fred Castellum]

Is this mostly exclusive to X3D chips right now? I'm running a 7950X and am wondering If I should be checking over my settings and voltages. 

[porina]

Quote

Update 4/25/2022, 10:41 am PT: AMD has responded to the numerous reports of chip failures with a short statement acknowledging that claims do exist and that the company is investigating. The statement confirms that AMD is working with its ODM partners (motherboard makers) to ensure safe voltage settings are applied to its Ryzen 7000X3D CPUs, but doesn't name the specific actions that it is taking:

"We are aware of a limited number of reports online claiming that excess voltage while overclocking may have damaged the motherboard socket and pin pads. We are actively investigating the situation and are working with our ODM partners to ensure voltages applied to Ryzen 7000X3D CPUs via motherboard BIOS settings are within product specifications. Anyone whose CPU may have been impacted by this issue should contact AMD customer support." -- AMD Spokesperson to Tom's Hardware.

Notably, the statement does not acknowledge the multiple reports of failures with standard Ryzen 7000 processors. ASUS has also issued a statement, clarifying that it will issue firmwares that limit SoC voltage to 1.3V. We're following up for more detail with AMD and will update as needed. Our original coverage with deeper details about the issues follows:

https://www.tomshardware.com/news/amd-ryzen-7000-burning-out-root-cause-identified-expo-and-soc-voltages-to-blame

 

Tom's Hardware got the above statement out of AMD.

[GoodBytes]

35 minutes ago, MyInnerFred said:

Is this mostly exclusive to X3D chips right now? I'm running a 7950X and am wondering If I should be checking over my settings and voltages. 

Considering that there like 1 case of non-X3D that was reported. Well, 2, Der8auer's... but Der8auer we know he did extreme OC, and the other case... probably did as well, or played with voltages without knowing what he was doing.

 

So, in my opinion, you are fine. I would not really worry about it, especially if you are on the latest BIOS, and you keep things default, and temps looks normal (i.e.: not 90C idle, for example)

[Fred Castellum]

4 minutes ago, GoodBytes said:

Considering that there like 1 case of non-X3D that was reported. Well, 2, Der8auer's... but Der8auer we know he did extreme OC, and the other case... probably did as well, or played with voltages without knowing what he was doing.

 

So, in my opinion, you are fine. I would not really worry about it, especially if you are on the latest BIOS, and you keep things default, and temps looks normal (i.e.: not 90C idle, for example)

Gotcha, thanks for the info. That's reassuring, if anything I should be using less wattage. 

[WallacEngineering]

5 hours ago, porina said:

https://www.tomshardware.com/news/amd-ryzen-7000-burning-out-root-cause-identified-expo-and-soc-voltages-to-blame

 

Tom's Hardware got the above statement out of AMD.

Yep, I was just about to reply with that. Apparently its SOC voltage spikes related to EXPO RAM overclocking.

 

In the article it is clearly stated that this affects ALL MOTHERBOARDS AND ALL RYZEN 7000 SERIES CPUS including X, Non-X, and X-3D.

 

The funny thing is that this is basically nobody's fault. AMD can't be held responsible for overclocking and MoBo vendors just enabled enough voltage for EXPO to work and be rock solid stable.

 

In the article, AMD states that 1.30V is the absolute maximum safe SOC voltage and will be issuing updates to lock down SOC voltage ASAP.

[Bitter]

Even 1.3 volts seems high! Zen2 it's like 1.15 max for a safe lifetime right?

[HenrySalayne]

On 4/25/2023 at 9:18 PM, GoodBytes said:

Well, 2, Der8auer's... but Der8auer we know he did extreme OC, and the other case... probably did as well, or played with voltages without knowing what he was doing.

He got the CPU with a drop if Indium coming out of the IHS from a viewer. This CPU was not used in extreme OC.

 

[TheVaultboy]

On 4/23/2023 at 2:49 PM, GoodBytes said:

Looks like ASUS is removing old BIOS's from its site.

I have both the previous and the current BIOS for the Gigabyte X670E if anyone has the required understanding to figure out what has changed.

On 4/23/2023 at 2:49 PM, GoodBytes said:

Probably they had the bug that caused the issue and wants to be sure that people don't install these old versions.

 

[TheVaultboy]

I got very annoyed the other night after being shown a few things on AMDs website, you can follow through on the links in my E-Mail to them if you want to see why. The following is that E-Mail.

 

Okay, when I brought my CPU and motherboard it was on the basis that I could run DDR5 memory at 6000Mhz all day long 24/7/365 but now I am being hounded with doubts which AMD  is the cause of. You advertise EXPO as a feature of your systems allowing memory to run at 6000Mhz and in some cases beyond but then I get told about this on your webpage found here: https://www.amd.com/en/technologies/expo hidden away in the bottom left hand corner where it is TOTALLY obvious under footnotes with a triangle beside it.

"Overclocking and/or undervolting AMD processors and memory, including without limitation, altering clock frequencies / multipliers or memory timing / voltage, to operate outside of AMD’s published specifications will void any applicable AMD product warranty, even when enabled via AMD hardware and/or software. This may also void warranties offered by the system manufacturer or retailer. Users assume all risks and liabilities that may arise out of overclocking and/or undervolting AMD processors, including, without limitation, failure of or damage to hardware, reduced system performance and/or data loss, corruption or vulnerability. GD-106"

Being that EXPO is a feature that is helping you sell boards - I want to know EXACTLY what you mean by the statement "outside of AMD’s published specifications" and I would like you to EXPLICITLY state what those Published Specifications are... if they are what is mentioned on your website under:
https://www.amd.com/en/products/apu/amd-ryzen-9-7950x3d which states you don't support memory running above the 5200Mhz Jedec standard then I feel it only fair to point out in the bottom of this page under yet ANOTHER hidden footnotes triangle the following:

"Game testing as of 5 December, 2022, by AMD Performance Labs using the following hardware: AMD Socket AM5 Reference Motherboard with AMD Ryzen™ 9 7950X, Ryzen™ 9 7950X3D and G.Skill DDR5-6000C30 (F5-6000J3038F16GX2-TZ5N) with AMD EXPO™; ALL SYSTEMS configured with NXZT Kraken X63, open air test bench, GeForce 4090, Windows® 11, PCIe® Resizable Base Address Register (“ReBAR”) ON, Virtualization-Based Security (VBS) OFF. All games tested at 1920x1080 with HIGH in-game preset and the chronologically newest graphics industry API available within the game’s rendering engine (e.g. Vulkan® over OpenGL™, DirectX® 12 over DirectX® 11). Desktop configurations will vary, yielding different results. Games tested: CS:GO, League of Legends, Assassin's Creed: Valhalla, Grand Theft Auto V, Hitman 3 Dubai GPU, Wolfenstein Youngblood (LabX), Hitman 3 Dubai CPU, Cyberpunk 2077, Borderlands 3, DOTA 2, Middle Earth: Shadow of War, F1 2021, Far Cry 6, Final Fantasy XIV, Warhammer: Dawn of War III, Shadow of the Tomb Raider, Total War: Three Kingdoms Battle, Deus Ex: Mankind Divided, Watchdogs: Legion, Riftbreaker CPU, Red Dead Redemption 2. RPL-035"

Now this shows that you are running EXPO enabled memory and are showing it as a feature so one can naturally assume that having EXPO set is not running out of "published specifications" one could also assume that taking a 7000Mhz XMP kit and running it at 6000Mhz with settings <= that of the EXPO kit would also not be outside "published specifications" unless of course you are meaning again the specification of 5200Mhz as stated on the previous URL.. in which case I feel AMD are being disingenuous and quite misleading.

You are hiding details that go against what you publicly advertise as features while excluding warranty on CPUs if you use said features. I see this as contradictory and a MASSIVE problem..

Do I want you locking things down?. No. No I do not.

Do I want you to be absolutely explicit in what you are saying? HELL YES. you need to be specific to avoid miscommunication. If you are saying that overclocking using EXPO past what you specify on your webpage for the 7950X3D will invalidate warranty then don't be wishy washy about it - damn well explain all the limitations explicitly or - you can expect users to come back at you and say  - you didn't clearly and explicitly specify therefore contributory negligence.  

Now please clarify your position on "AMD’s published specifications" be explicit and verbose and leave no room for any doubt.

 

 

 

[starsmine]

36 minutes ago, TheVaultboy said:

I got very annoyed the other night after being shown a few things on AMDs website, you can follow through on the links in my E-Mail to them if you want to see why. The following is that E-Mail.

 

Okay, when I brought my CPU and motherboard it was on the basis that I could run DDR5 memory at 6000Mhz all day long 24/7/365 but now I am being hounded with doubts which AMD  is the cause of. You advertise EXPO as a feature of your systems allowing memory to run at 6000Mhz and in some cases beyond but then I get told about this on your webpage found here: https://www.amd.com/en/technologies/expo hidden away in the bottom left hand corner where it is TOTALLY obvious under footnotes with a triangle beside it.

"Overclocking and/or undervolting AMD processors and memory, including without limitation, altering clock frequencies / multipliers or memory timing / voltage, to operate outside of AMD’s published specifications will void any applicable AMD product warranty, even when enabled via AMD hardware and/or software. This may also void warranties offered by the system manufacturer or retailer. Users assume all risks and liabilities that may arise out of overclocking and/or undervolting AMD processors, including, without limitation, failure of or damage to hardware, reduced system performance and/or data loss, corruption or vulnerability. GD-106"

Being that EXPO is a feature that is helping you sell boards - I want to know EXACTLY what you mean by the statement "outside of AMD’s published specifications" and I would like you to EXPLICITLY state what those Published Specifications are... if they are what is mentioned on your website under:
https://www.amd.com/en/products/apu/amd-ryzen-9-7950x3d which states you don't support memory running above the 5200Mhz Jedec standard then I feel it only fair to point out in the bottom of this page under yet ANOTHER hidden footnotes triangle the following:

"Game testing as of 5 December, 2022, by AMD Performance Labs using the following hardware: AMD Socket AM5 Reference Motherboard with AMD Ryzen™ 9 7950X, Ryzen™ 9 7950X3D and G.Skill DDR5-6000C30 (F5-6000J3038F16GX2-TZ5N) with AMD EXPO™; ALL SYSTEMS configured with NXZT Kraken X63, open air test bench, GeForce 4090, Windows® 11, PCIe® Resizable Base Address Register (“ReBAR”) ON, Virtualization-Based Security (VBS) OFF. All games tested at 1920x1080 with HIGH in-game preset and the chronologically newest graphics industry API available within the game’s rendering engine (e.g. Vulkan® over OpenGL™, DirectX® 12 over DirectX® 11). Desktop configurations will vary, yielding different results. Games tested: CS:GO, League of Legends, Assassin's Creed: Valhalla, Grand Theft Auto V, Hitman 3 Dubai GPU, Wolfenstein Youngblood (LabX), Hitman 3 Dubai CPU, Cyberpunk 2077, Borderlands 3, DOTA 2, Middle Earth: Shadow of War, F1 2021, Far Cry 6, Final Fantasy XIV, Warhammer: Dawn of War III, Shadow of the Tomb Raider, Total War: Three Kingdoms Battle, Deus Ex: Mankind Divided, Watchdogs: Legion, Riftbreaker CPU, Red Dead Redemption 2. RPL-035"

Now this shows that you are running EXPO enabled memory and are showing it as a feature so one can naturally assume that having EXPO set is not running out of "published specifications" one could also assume that taking a 7000Mhz XMP kit and running it at 6000Mhz with settings <= that of the EXPO kit would also not be outside "published specifications" unless of course you are meaning again the specification of 5200Mhz as stated on the previous URL.. in which case I feel AMD are being disingenuous and quite misleading.

You are hiding details that go against what you publicly advertise as features while excluding warranty on CPUs if you use said features. I see this as contradictory and a MASSIVE problem..

Do I want you locking things down?. No. No I do not.

Do I want you to be absolutely explicit in what you are saying? HELL YES. you need to be specific to avoid miscommunication. If you are saying that overclocking using EXPO past what you specify on your webpage for the 7950X3D will invalidate warranty then don't be wishy washy about it - damn well explain all the limitations explicitly or - you can expect users to come back at you and say  - you didn't clearly and explicitly specify therefore contributory negligence.  

Now please clarify your position on "AMD’s published specifications" be explicit and verbose and leave no room for any doubt.

 

 

 

All EXPO and XMP are running out of spec.
JDEC is spec, deviation is out of spec.

Jdec is guaranteed, XMP/EXPO should work.

AMD is not being disingenuous or misleading.

Overclocking does not invalidate warrantee on AMD (doesn't on intel either) with exceptions, which are, YOU BROKE IT by doing something unreasonable. 
XMP/EXPO are considered reasonable overclocking to an extent. 

[goodtofufriday]

On 4/25/2023 at 4:36 AM, Vishera said:

That's why i am never a first adopter.

I typically buy products that have been around for at least a year, which also over time become cheaper and cheaper.

That's why i went with a Ryzen 7 5700X.

wouldnt the 7800x3d be 2nd gen? 

[WallacEngineering]

36 minutes ago, goodtofufriday said:

wouldnt the 7800x3d be 2nd gen? 

He is referring to the AM5 socket I believe.

[porina]

Quote

"We have root caused the issue and have already distributed a new AGESA that puts measures in place on certain power rails on AM5 motherboards to prevent the CPU from operating beyond its specification limits, including a cap on SOC voltage at 1.3V. None of these changes affect the ability of our Ryzen 7000 Series processors to overclock memory using EXPO or XMP kits or boost performance using PBO technology.

We expect all of our ODM partners to release new BIOS for their AM5 boards over the next few days. We recommend all users to check their motherboard manufacturers' websites and update their BIOS to ensure their system has the most up-to-date software for their processor.

Anyone whose CPU may have been impacted by this issue should contact AMD customer support. Our customer service team is aware of the situation and prioritizing these cases." — AMD Representative to Tom's Hardware.

https://www.tomshardware.com/news/amd-issues-follow-up-statement-on-ryzen-burnout-issues-limits-soc-voltages

 

Sounds like AMD will be accepting failures due to this.

 

Edit: thinking a bit more, the question then arises, why were higher voltages being used before? Was it the mobo makers setting it high to help ensure stability at higher speeds? It'll be interesting to see if this impacts higher speed kits in particular.

[pas008]

39 minutes ago, porina said:

https://www.tomshardware.com/news/amd-issues-follow-up-statement-on-ryzen-burnout-issues-limits-soc-voltages

 

Sounds like AMD will be accepting failures due to this.

 

Edit: thinking a bit more, the question then arises, why were higher voltages being used before? Was it the mobo makers setting it high to help ensure stability at higher speeds? It'll be interesting to see if this impacts higher speed kits in particular.

are they all mobo with phoenix or ami bios?

could this be bios company?

[Vishera]

1 hour ago, goodtofufriday said:

wouldnt the 7800x3d be 2nd gen? 

Nope, Zen 4 is new in the market.

[HenrySalayne]

2 hours ago, porina said:

Edit: thinking a bit more, the question then arises, why were higher voltages being used before? Was it the mobo makers setting it high to help ensure stability at higher speeds? It'll be interesting to see if this impacts higher speed kits in particular.

AMD mainboards take a quite a long time to boot because of "memory training". This is just speculation on my part, but maybe the SOC/IOD voltage is sometimes too aggressively increased to hit the required DDR frequency and latency. The training is done every time booting and I get slightly different CPU SOC voltages (If I'm not mistaken).  (I had 1.225V yesterday and 1.279V today with DDR6000 32-38-38-96*)

 

A combination of poor silicon (memory controller), a lack of temperature sensors in the vicinity and excessive voltage could result in a hotspot and eventually a short circuit.

 

Edit: *after a quick reboot it is still at 1.279V. I will check again tomorrow, when it's a cold start.

[Napper198]

2 minutes ago, HenrySalayne said:

AMD mainboards take a quite a long time to boot because of "memory training". This is just speculation on my part, but maybe the SOC/IOD voltage is sometimes too aggressively increased to hit the required DDR frequency and latency. The training is done every time booting and I get slightly different CPU SOC voltages (If I'm not mistaken).  (I had 1.225V yesterday  and 1.279V today with DDR6000 32-38-38-96)

 

A combination of poor silicon (memory controller), a lack of temperature sensors in the vicinity and excessive voltage could result in a hotspot and eventually a short circuit.

before turning off EXPO (for the time beien) I saw 1.38V on SoC. that’s what I get for leaving it on auto I guess. Also interesting to note that memory training happens on every boot, I just thought the server 10gig card took so long because of network stack shenanigans

[HenrySalayne]

9 minutes ago, Napper198 said:

before turning off EXPO (for the time beien) I saw 1.38V on SoC. that’s what I get for leaving it on auto I guess. Also interesting to note that memory training happens on every boot, I just thought the server 10gig card took so long because of network stack shenanigans

Just read that new AM5 AGESA 1006 from earlier this month should significantly reduce boot times. Since it is recommended to get the new BIOS version, this is a quality of life improvement that makes it worth the time.

[Napper198]

2 hours ago, HenrySalayne said:

Just read that new AM5 AGESA 1006 from earlier this month should significantly reduce boot times. Since it is recommended to get the new BIOS version, this is a quality of life improvement that makes it worth the time.

I'm already running 1.0.0.6 and it's still kinda slow but tbf I'm also always cold booting since te desk gets disconnected from power entirely when I'm not using it (0,45€/KWh do add up quite a bit over time)

[Kisai]

10 hours ago, starsmine said:

All EXPO and XMP are running out of spec.
JDEC is spec, deviation is out of spec.
 

No, no no. XMP is "in spec" because when you buy 3200Mhz DDR4 modules, they are only operating at 3200Mhz in XMP mode.

If that was not the case 3200Mhz modules would only be marketed as 2133Mhz.

 

Both the CPU and the RAM are validated to run at the speeds advertised. If you only run at 2/3rds the speed advertised to be safe, you have to buy much faster memory to operate at JDEC speeds that the CPU supposedly supports. Note the part number even includes the 3200C16 speed in it.

 

So if you're not allowed to use XMP/EXPO, because it might cook the CPU, then who's fault is that?

 

[leadeater]

10 hours ago, HenrySalayne said:

AMD mainboards take a quite a long time to boot because of "memory training". This is just speculation on my part, but maybe the SOC/IOD voltage is sometimes too aggressively increased to hit the required DDR frequency and latency. The training is done every time booting and I get slightly different CPU SOC voltages (If I'm not mistaken).  (I had 1.225V yesterday and 1.279V today with DDR6000 32-38-38-96*)

Wasn't there a recent AGESA version that addressed the memory training and made it faster? I could just be hallucinating though lol

 

Edit:

haha or just wait and finish reading

10 hours ago, HenrySalayne said:

Just read that new AM5 AGESA 1006 from earlier this month should significantly reduce boot times. Since it is recommended to get the new BIOS version, this is a quality of life improvement that makes it worth the time.