Caporegime
It will because AMD simply flips the set celling locked to 1.25v switch in their AGESA code, and that's it.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
AMD Ryzen 7 7800X3D CPU Burns Up
- Thread starter Jamin280672
- Start date
It will because AMD simply flips the set celling locked to 1.25v switch in their AGESA code, and that's it.
Motherboard makers have been pushing voltages too hard for auto overclocking features as far back as I can remember. They've been called out on these shenanigens before. Those AI overclocking features they used to advertise were particular offenders. The difference here is that they've finally managed to damage some CPUs.
If anything good comes out of this it has shed light on the practice, and will hopefully provoke some introspection with themselves, and the way the media reviews these things.
If anything good comes out of this it has shed light on the practice, and will hopefully provoke some introspection with themselves, and the way the media reviews these things.
Associate
I went for a 7600 (non x) off the members market. I am really please with it as I have a RTX3080. Techpowerup tested with a 3080 and I game at 1440p.
It's nice seeing the chip boost to 5150Mhz and stay cool. I have a large CPU cooler on it that I have had for years (Thermalright Macho Rev B).
You will always see people say go for 8 core, but there still isn't much difference in Games. I have had 2600x / 3700x and still have 5600x / 5600G and a 5700x in spare PC's.
Also the 7000 series does have more L2 cache as standard:
Full technical specification of AMD Ryzen 5 7600 and Ryzen 5 5600
| L1 Cache | 64K (per core) | 64K (per core) |
| L2 Cache | 1MB (per core) | 512K (per core) |
| L3 Cache | 32MB (shared) | 32MB (shared) |
I might get a 7800X3D one day, but there will probably be a 9600 out next year that will be fast!
Might do that because surprise surprise OcUK declined my return request.
Last edited:
Soldato
Why are you stressing ? You have a 3 year warranty on the CPU so if it degrades within that time you can get a brand new one.
Nobody knows what voltages etc AMD actually put through the chips during binning and testing etc anyway.
Just crack on and use your kit and enjoy it.
Soldato
But if they weren't breaking that then what difference does it make?
I'm not even sure if such a thing is possible or not, but it would be a nice idea from AMD if they released a diagnostic app for users to download and run checks on their 7000 or 7000X3D CPU to see if it has or hasn't been damaged in any way, and if it has they will honour a replacement.
It probably won't happen. But, that's what I would do anyway.
If the CPU was an automobile this would be when they would issue a recall.
It probably won't happen. But, that's what I would do anyway.
If the CPU was an automobile this would be when they would issue a recall.
Last edited:
The usual suspects, though what goes on in here is mild compared to the utterly retarded rambling you see of some people on Facebook. Some of them can't even accept the 7800x3d is the fastest gaming cpu and its all a media conspiracy against intel. Makes some of the drivel you see on here actually seem rational.
Caporegime
I might have some sympathies for this if AMD paid dividend, which they haven't since 1997, there is absolutely 0 incentive for anyone to have an agenda like that, even if.... a conspiracy? What like they all got together and collectivity decided "lets do Intel in, because reasons" lol....
I beg to differ: IMHO it is a hardware fault. That it can be worked around by software is only a temporary fix.
Soldato
- Joined
- 14 Aug 2018
- Posts
- 3,548
Yes, everything should be sorted in a month or so, so it might be prudent to wait.
Something like this can happen to Intel or AMD and any board vendor when new platforms are introduced. The slightly disconcerting thing about this is that we're ~7 months into the release of AM5 motherboards and you would kind of hope that most of the gremlins would have been discovered and ironed out by now. There can also sometimes be the notion (maybe without good cause) that if there was an issue with this then what other yet to materialise issues could there be....
There is something to be said for getting well tried and tested platforms like AM4 or Z690.
Well it can take time to show things. Like it's pretty apparent the x3d chips are more sensitive to voltage/heat so they just showed it up easier. Where as in the same situation the non-x3ds might have been fine with those voltages. GN mentioned that the shut down temperatures seem based on the non-x3d values too, so it seems like its part laziness too.
Truthfully after my experience with Am4 I'm not sure I'd recommend AMD for a builder who isn't happy going into the bios tweaking things really. I got a x570 board about a month after that chipset launched, and I didn't have too many issues until the 3080 came out about a year later and showed up how completely buggy the PCIE Gen 4 implementation was. I couldn't use my 3080 at Gen 4 speeds for over half a year(to be fair it didn't change much performance wise) because it would just randomly cause my USB devices to drop connection. Since I was lucky enough to get a 3080 on launch I was one of the early people and at that stage it wasn't super clear what the cause was. I spent so much time troubleshooting, messing with voltages, even swapped my PSU out. Until eventually I found some very old posts from Navi card users which were technically the first Gen 4 cards having very similar issues, just the navi cards weren't that popular so the issues got glossed over. I submitted a report to AMD early on about the issue and they responded saying we'll look into it. I didn't get a response till over 6months later when they actually responded to that same age old ticket saying it was finally fixed in the new AGESA. I also had lots of issues getting ram stable via XMP, always had to manually tweak timings/voltages. Funnily enough EXPO seemed a lot better this time around in that sense as it just worked this time, just a shame about it supplying too much voltage, lol.
The hardware that pumps the voltage out is controlled by software so if voltage is clamped in bios then its a permanent fix.. well it should be.
I'm sure there are a lot of AAA game studios that would love him in PR, nah it's not badly coded it's a hardware fault 
There's a lot to complain about with Motherboard makers for this generation, making overengineered vastly expensive products that still somehow seem to lack some basic features at each tier. But there's no evidence to suggest fundamental hardware faults yet. Maybe some will turn up at some point and people can post with glee about them then.
There's a lot to complain about with Motherboard makers for this generation, making overengineered vastly expensive products that still somehow seem to lack some basic features at each tier. But there's no evidence to suggest fundamental hardware faults yet. Maybe some will turn up at some point and people can post with glee about them then.
Caporegime
The hardware is pretty dumb, it will sit and do literally nothing but wait on instructions from software and it will do whatever the software tells it to do, including slowly kill itself, because it does not know any better, it knows nothing at all. It is not able to act on its own.
To say its a hardware fault is one of the most idiotic things written on this forum.
To say its a hardware fault is one of the most idiotic things written on this forum.
Last edited:
Caporegime
I hope you properly tested those memory overclock settings I gave you.
Last edited:
Caporegime
Are you still running it at 6000 CL30? in that a manual tune or just EXPO?
They were working before but I’m back to XMP profile with lower Soc voltage. Seems 1.25v isn’t enough to for XMP be stable.I hope you properly tested those memory overclock settings I gave you.
Last edited: