FranzB, can you post your serial number so we can compare?
Which serial master? The CPU’s one?
Franz,
When you have the prime95 fail, is it only 1 or 2 workers or several more?
Another thing to note, we’re running similar power supplies, you the hx1000i, me the hx1200i.
1 Like
In the past people have posted the lid serial number. That helps determine the batch I think.
Many workers (>= 8), haven’t let it run long enough to see if they stop dying at some point (I would assume so).
Can I get the lid serial number without removing the heatsink?
I think it is on the Box. Otherwise Like On my system I have to remove the heatsink and thermal paste.
Ok, I’m indeed seeing a part number on the box. Sending it to you in PM.
Thanks
I am also having this issue and discovered this post. I noticed that there was a new firmware released by GIGABYTE (F4B) on 3/4/19. Has this fixed it for anyone? I haven’t tried it yet.
Luckily I bought my parts from Amazon so I am thinking I am going to be doing a lot of returns and going with the new 2nd gen Intel Xeon Cascade Lake processor and board now that Intel has lowered their prices significantly.
I’m a bit skeptical of your authenticity. Can you post more details of your build (possibly pictures too?) and what exactly you use it for?
I saw F5B for for the Aorus Master, I haven’t tried yet. I was hoping to today but life got in the way. Hopefully by the weekend I can try.
Aside from this one specific test, my system is otherwise stable.
LOL… What do you want to see a picture of exactly? My receipt from Amazon? A picture of the motherboard in the case? A screenshot of Prime 95 failing?
I have a 3960X with a GIGABYTE TRX40 DESIGNARE mobo.
I don’t have any real workload yet as I was simply doing a burn-in test with Prime 95 and had the error. I haven’t even had a chance to really use it yet. I did some googling and stumbled upon this discussion, originally found it on “Hacker News / YCombinator” (Can’t link to it from here) but there is ton of discussion if you find the post by googling:
“hacker news prime 95 3960x avx2”
It’s probably something on Gigabyte’s end which should be fixable by bios updates or worst case changing Mobo vendors. I don’t think AMD would’ve released a flawed processer, especially something that shares a lot with Epyc.
If you don’t really have a use case on mind, I don’t see why you’d wanna return it all back asap for an inferior platform.
1 Like
Aours Extreme has now F4c ( 2020/03/04 ). It seems that they decided to no longer make available previous F4 releases for that board.
… and its gone too.
Not sure if they have issues with website or the BIOS.
Is it safe to buy a TRX40 Aorus Xtreme and 3960X now based on the assumption it will be fixed?
I would not make any assumptions that they would (or even could) fix this with a BIOS update. Instead, I would hold off on making any purchases until someone here says that its fixed.
They are under no obligation to fix this problem
I downloaded F5B, but I will test F5C.
There is now a F4d.
But
lol
On the real glad to see there is some progress being made and seems to be an issue with the mobo.