General > General Technical Chat

Linus Sebastion Turned Down $100M Offer

<< < (17/35) > >>

hans:

--- Quote from: wraper on August 16, 2023, 05:44:43 pm ---
--- Quote from: hans on August 16, 2023, 05:35:43 pm ---no mention or grasp of SCR latch-up

--- End quote ---
And why should it be mentioned when there is no any evidence of it being related to the issue?

--- End quote ---

Because it shows all the symptoms of one. Overvolting a voltage rail, then seeing it draw tons of current regardless of operational state, then see the chip go burn itself up in a thermal runaway.
Documentation of the runaway is clear but won't reveal anything. That's why their lab gave a listen of a few dozen reasons. The actual value in the post mortem is a bit exaggerated IMO, as a simple latch-up would be a prime suspect. But the comments section is full of praise of "next level testing" and "raising the bar". C'mon. It's cool to see the pictures, but I still don't know why the chip failed.


--- Quote from: coromonadalix on August 16, 2023, 06:32:00 pm ---aren't we all humans ??

--- End quote ---
Yes.

But that's not how social media works I think

wraper:

--- Quote from: Veteran68 on August 16, 2023, 06:28:03 pm ---Like when they really stretched the truth saying that Asus claimed updating BIOS fixed the voltage issues, but also voided warranty, thereby leaving users in a catch-22 situation. No mobo company has EVER claimed that updating BIOS voids their warranty. GN/Steve walked back the criticism after Asus carefully spelled out that warranties would not be voided.

--- End quote ---
Asus released it as beta bios. And it came with terms are that using it can void the warranty. I somewhat rolled my eyes too because in practice it won't be very likely that they will actually void the warranty if motherboard fails, however GN was correct at least technically.

wraper:

--- Quote from: hans on August 16, 2023, 07:20:34 pm ---
--- Quote from: wraper on August 16, 2023, 05:44:43 pm ---
--- Quote from: hans on August 16, 2023, 05:35:43 pm ---no mention or grasp of SCR latch-up

--- End quote ---
And why should it be mentioned when there is no any evidence of it being related to the issue?

--- End quote ---

Because it shows all the symptoms of one. Overvolting a voltage rail, then seeing it draw tons of current regardless of operational state, then see the chip go burn itself up in a thermal runaway.
Documentation of the runaway is clear but won't reveal anything. That's why their lab gave a listen of a few dozen reasons. The actual value in the post mortem is a bit exaggerated IMO, as a simple latch-up would be a prime suspect. But the comments section is full of praise of "next level testing" and "raising the bar". C'mon. It's cool to see the pictures, but I still don't know why the chip failed.

--- End quote ---
SCR latch-up does not happen because you run the chip at voltage somewhat higher than normal. It generally happens with large transients on I/O, like ESD discharge. Not due to running the chip at 0.15V higher than its safe maximum or exceeding its thermal limits as was with faulty BIOSes. You cannot fix an SCR latch-up susceptibility by applying a BIOS update.

coppice:

--- Quote from: wraper on August 16, 2023, 07:32:21 pm ---SCR latch-up does not happen because you run the chip at voltage somewhat higher than normal. It generally happens with large transients on I/O, like ESD discharge. Not due to running the chip at 0.15V higher than its safe maximum or exceeding its thermal limits as was with faulty BIOSes. You cannot fix an SCR latch-up susceptibility by applying a BIOS update.

--- End quote ---
That's a rather old school view of SCR latchup. On modern devices with multiple voltage rails you can find some fairly complex pin protection circuitry that may be possible to get into a latched up state if various voltage levels sequence in an unfortunate manner. I'm not saying that is what happened here, but its an ongoing design topic for making things like fine geometry MCUs as robust as possible.

Veteran68:

--- Quote from: wraper on August 16, 2023, 07:22:25 pm ---Asus released it as beta bios. And it came with terms are that using it can void the warranty. I somewhat rolled my eyes too because in practice it won't be very likely that they will actually void the warranty if motherboard fails, however GN was correct at least technically.

--- End quote ---

Steve was correct for questioning it and seeking clarification from Asus, but the whole pearl-clutching way it was played for dramatic effect was over the top. Asus released the BIOS as an emergency response to the overvoltage issue, which was killing brand new boards+CPUs costing $1000+. The warranty voiding verbiage was likely either some overlooked boilerplate or dropped in by an overzealous product manager already gunshy over possible legal liabilities from this issue. But it makes little sense to believe they'd void the warranty for an emergency fix to prevent a destructive fault that would have resulted in a warranty repair. Knowing what the problem was, they had to know they'd be opening themselves up to class action litigation, especially in the US, if they played such games. And I say that as someone who's had issues with Asus over the years.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod