We reported this problem a couple of days ago, here:
Impacts all motherboard makers and all Ryzen 7000 chips:
Multiple reports of Ryzen processors burning out have burst onto the internet over the last few days. The damaged chips have not only bulged out and overheated to the point they have become desoldered, but they have also done significant damage to the motherboards they are installed in. We reached out to our industry contacts and learned some new information about the nature of the problem and the scope of AMD's planned fix. Our information comes from multiple sources that wish to remain anonymous, but the info from our sources aligns on all key technical details. As with all unofficial information, we should take the finer details with a grain of salt until AMD issues an official statement.
First, we're told this condition can occur with both standard Ryzen 7000 models and the new Ryzen 7000X3D chips, though the latter is far more sensitive to the condition, and the root cause could be different between the two types of chips. AMD will issue a fix soon, but the timeline is unknown. We're told that failures have occurred with all motherboard brands, including Biostar, ASUS, MSI, Gigabyte, and ASRock.
According to our sources and seconded by an ASUS statement to Der8auer, the problem stems from SoC voltages being altered to unsafe higher levels. This can be imposed from either the pre-programmed voltages used in EXPO memory overclocking profiles or when a user manually adjusts the SoC voltages (a common practice to eke out a bit more memory overclocking headroom).
Our sources also added further details about the nature of the chip failures — in some cases, excessive SoC voltages destroy the chips' thermal sensors and thermal protection mechanisms, completely disabling its only means of detecting and protecting itself from overheating. As a result, the chip continues to operate without knowing its temperature or tripping the thermal protections.
AMD's modern chips often run at their thermal limits to squeeze out every last drop of performance within their safe thermal range — it isn't uncommon for them to run at 95C during normal operation — so they will automatically continue to draw more power until it dials back to remain within a safe temperature. In this case, the lack of temperature sensors and protection mechanisms allows the chip to receive more power beyond the recommended safe limits. This excessive power draw leads to overheating that eventually causes physical damage to the chip, like the bowing we've seen on the outside of several chip packages, or the desoldering reported by Der8auer.
The chip continues to receive excessive current through the motherboard socket during this death spiral of sorts, thus leading to the visible damage we can see in the socket to the vCore pins and the bulging on the chip's LGA pads. However, less visible damage also extends to the CPU SoC, CPU_VDDCR_SOC, and CPU VDD MISC rails/pins — they just don't pull enough current to leave visible scorching like we see with the vCore pins.
We do know that 1.25V is the recommended safe SoC voltage limit, and we're told that 1.4V and beyond definitely increases the likelihood of the condition occurring. To be clear, running beyond 1.4V doesn't ensure that your chip will burn out, but your odds will increase. Conversely, 1.35V appears to be "safe." Proceed at your own risk, though.
Our sources say that AMD is working on a fix that includes a voltage cap or lock in the firmware/SMU, which should prevent EXPO memory profiles and simple BIOS manipulations from exceeding an as-yet-undefined limit. We're also told that AMD can't completely prevent SoC voltage manipulations because the amount fed to the chip is dictated by the VRMs, leaving a means for crafty motherboard vendors to allow voltage changes despite AMD's lock (this would not be the first time motherboard vendors have circumvented limits to offer rare functionalities).
A few motherboard vendors, like ASUS and MSI, have already issued new BIOSes to correct some of the issues. However, we have confirmed that failures have also occurred on Biostar, ASRock, and Gigabyte boards, so all vendors are impacted to some degree.
As with all forms of overclocking, any damage from using an EXPO overclocking profile is not covered by your warranty, but given the situation, we don't think that AMD or the motherboard vendors would use the lack of warrantied EXPO support to invalidate warranties.
The advertised performance you get from an EXPO profile is also not guaranteed by the chipmaker. It's also noteworthy that AMD's purportedly planned SoC voltage cap could lead to lower stable memory overclocking frequencies. However, we don't think that will matter too much to most Ryzen 7000 owners, as the sweet spot DDR5-6000 should work just fine within the proposed limits. However, extreme overclockers and those pushing the very bleeding edge of performance could end up with lower overclocking limits. Time will tell.
« The FDA Would Like to Remind You Not to Put Amniotic Fluid in Your Eyes | Bad Medical News Causes Patients to Choose Brand Name Drugs Over Generics, Costing Billions »
Related Stories
Likely a motherboard BIOS issue:
High-performance microprocessors tend to get hot and, when overclocked without proper cooling or throttling, can literally burn out When an old CPU does so in an old PC because of dust and a worn out fan, there is nothing surprising about it. But when a new CPU breaks on a shiny new motherboard, that's surprising. This is what happened to an AMD Ryzen 7 7800X3D, one of the best CPUs, which belonged to a Reddit user. And, according to other users, he's not alone.
"The CPU pad is physically bulging," wrote Speedrookie, the owner of the burned-out processor. "I imagine there was just too much heat on the contacts causing the pad to expand. Not that the CPU has an internal component which exploded."
At least when it comes to AMD's Ryzen 7000X3D-series processors, it appears that this is by far not an isolated case. There are reports from other Reddit users who had the same experience with their Ryzen 7000X3D CPUs on motherboards from Asus and MSI.
Famous overclocker der8auer also had an issue with an AMD Ryzen 7000X3D chip burning out during some early overclocking tests a few weeks ago. As you can see in his video, he says "I did not expect this to happen so quickly and especially right out of the BIOS."
There are about a million of reasons why a modern processor can burn out. Defective sockets or a motherboards [sic] are likely causes and insufficient cooling can cause a similar result. A BIOS version that tends to automatically overclock CPUs too much could be a yet another reason for a processor failure. In fact, as noticed by HXL (@9550pro), Asus has just withdrew old BIOSes for many of its AMD X670-based motherboards, but for some of them old BIOSes are still available.
[...] Again, given that the information is insufficient to say the least, we cannot make any conclusions at this point. We'll reach out to our contacts at AMD and the motherboard vendors to see if they are aware of any issues. For now, we recommend that those with AMD's Ryzen 7000X3D processors keep a close eye on their CPU temps, use adequate cooling and keep their BIOSes up-to-date.
(Score: 2) by Rosco P. Coltrane on Wednesday April 26, @12:16PM
Excessive heat
(Score: 2) by Mojibake Tengu on Wednesday April 26, @11:46PM
This is the worst kind of failure reaction out of possible. The processor is already smart enough to fault or double fault itself when thermal control is lost, and it should halt itself, since we all know the late designs are so overengineered they can't operate without thermal.
The edge of 太玄 cannot be defined, for it is beyond every aspect of design
(Score: 2) by hendrikboom on Friday May 19, @02:34AM
Would the 7000 series include the 7040 series?
Because that's what framework is planning to use on one of their new processor boards.