It has been 2 weeks that I try to solve my issue. My PC does not boot on Windows 10, or Windows 7 or Linux, BSOD or linux equivalent everytime just after the motherboard logo. You will see that I tried a lot of things before getting the statement that the processor might be dead...
Here is my setup:
- MSI Z170A Krait Gaming motherboard
- Intel Skylake Processeur Core i5-6600K / 3.5 GHz
- Radiateur CPU Arctic Cooling Freezer 13
- Alim Corsair RM650x
- 2x8Go DDR4 2400MHz RipJaws (XMP set)
- MSI GTX 970 4GD5T OC
- SSD Samsung 256Go Pro (Windows 10 Pro installed)
- Seagate 2To Barracuda
There is no overclocking and the BIOS is updated to latest version.
The BSOD happened when I was playing a game, and from there the PC never started again and always does BSOD restart with all errors possible.
So I removed Graphic board, and hard drives. Here is the things I tried:
- Boot on Windows 10 usb key --> BSOD
- Remove RAM to let just One, changing slots, install brand new RAM --> BSOD
- Boot on Livekey Ubuntu --> Getting the kernel panic BSOD equivalent
- Boot on Windows 7 --> BSOD
- Reset of the motherboard, with default parameters, even tried to remove the battery
- As I first thought it was the motherboard, I installed a brand new Motherboard (of course same chipset, etc) --> BSOD
So, I see some posts the internet with my processor's generation related to BSOD events...
Do you think my processor could have a problem with what I mentioned?
Thank you very much !
Windows 10: Yes sometimes, the blue screen message is not even full
There is an issue with both Skylake and Kaby Lake processor generations related to Hyper-Threading. This can result in strange issues, including BSODs. The microcode fix for this issue was released a month or two ago. By now, your board manufacturer should have incorporated this fix into a BIOS update and made it available. Check to see if you have this BIOS update installed. If you don't, install it and see if this alleviates your problem...
Hope this helps,
P.S. If this isn't the issue, the next thing I would suspect is bad DRAM...
Thank you for your answer.
I updated with last update, in the releases, I don't see any message related to what you mentioned https://fr.msi.com/Motherboard/support/Z170A-KRAIT-GAMING.html https://fr.msi.com/Motherboard/support/Z170A-KRAIT-GAMING.html
But I asked MSI about it.
What to you mean by bad DRAM? You mean incompatible DRAM or dead DRAM ? Because I tried brand new ones from another brand (Corsair), and it made no changes.
Well, if you go by their notes regarding this provided for previous BIOS releases, it doesn't look like MSI has produced a BIOS with updated microcode within the correct timeframe. I would be contacting them to find out what is going on. At the same time, it is important to understand that this is only a possibility for what is causing your problem. It's not something to hang your hat on.
When I talk about memory, my first comment is with respect to memory speeds. The processor is designed and validated to support 2133MHz memory. Anything faster - like your 2400MHz memory - is technically overclocking the processor's memory busses. This can cause higher temperatures in portions of the processor die and can make memory transfers more susceptible to noise-related failures. The noise on the memory busses is a combination of that generated by the process, that generated by the DIMMs and that generated by the motherboard. If the noise threshold is too high, memory transfers can fail (noise can look like data). This can cause bus slowdown (and lockup), data errors (resulting in application GP faults and system BSODs), etc. Secondly, as circuits and silicon age, they produce more and more noise. Memory that works just fine when a system is fairly new may eventually start to fail as a result. The higher the bus speeds, the more susceptibility to noise and the sooner that this can happen.Thus, my first recommendation is always: if you think memory may be involved in your issues, down-clock your memory to 2133MHz or (significantly better) test with true 2133MHz memory. My second comment is related to applications like MemTest86 and MemTest86+. They are not infallible. I have seen memory pass with flying colors for many, many passes yet it is still bad and will cause problems in normal use. Bottom line, while they are a good indicator, you simply cannot say it's not the memory just because these tests pass. Oh, and speaking of the Windows Memory Diagnostic: useless WOT; I don't trust it to tell me anything. It sometimes can't even verify that bad memory is truly bad.
Ok, all that said, you have to look at three possibilities, (1) problem with the DIMMs, (2) problem with the board or (3) problem with the processor. Since you say you swapped motherboards and the problem remained, you likely have eliminated # 2 -- unless it is a compatibility issue. You also said you tried different memory. Was this memory 2133MHz? If not, it's not a valid eliminator in my mind; go back and test with true 2133MHz memory and I will accept this then. Now we hit the last one, namely the processor. Could you have damaged it (zapped it, for example)? The only way to tell is to test with another processor and see if this eliminates the issue.
That's enough to chew on for now...
Thank you for your answer.
Yesterday, I tried with brand new true 2133Mhz, with only one first and then two. But it does the same thing.
I never tried these applications, but anyway I cannot go to a desktop anymore, not even once.
I have no idea how I could have damaged it... I built the computer in January 2016, never had any troubles, I never tried to modify anything (overclock, etc).
The only thing I did is move the computer from a city to an other.
So I will buy a processor. If it works, I hope I can make the warranty work.
Oh, it's pretty easy to damage the board and/or processor when physically handing them. The # 1 cause is static. It's so incredibly easy to do so.
Moving a system from one city to another can cause all sorts of issues. The system is going to be subjected to all sorts of bumps and vibrations. Cables can come loose, cards (including the DIMMs!) can work their way out of the board slots, etc. and etc. When moving a system like this, you need to reseat every board and every DIMM and every cable connector.
From what you have said, you have eliminated the memory as a player in this issue. This doesn't necessarily mean that it's the processor that has failed; it could be the Power Supply, it could be the motherboard, it could be an add-in card. You're a long way from knowing, unfortunately...
Ok so I changed with the same processor, just bought it to test.
It directly worked...
So I guess I need to deal with the warranty now...
Thank you for all your comments and suggestions!
For warranty replacement, you need to contact Intel Customer Support directly. Here is contact information for doing so, separated out by geography:
http://www.intel.com/content/www/us/en/support/contact-support.html# @11 Intel Customer Support Contact Information for US and Canada
http://www.intel.com/content/www/us/en/support/contact-support/emea-contact.html Intel Customer Support Contact Information for Europe, Middle East and Africa
http://www.intel.com/content/www/us/en/support/contact-support/apac-contact.html Intel Customer Support Contact Information for Asia-Pacific
http://www.intel.la/content/www/xl/es/support/contact-support/lar-contact.html Intel Customer Support Contact Information for Latin America
You should mark this issue as resolved (such as it is)...