i2500k Memory Controller Failure?

[Phunky]

So was testing a returned mobo that I picked up cheap on eBay (Asus p8z77-v deluxe for £37!) and during testing I figured out that it was the first PCIe controller that was failing but then I started to get Q-CODE 55 errors (Memory not installed).

I took everything out and attempted to boot with a single stick of RAM in DIMM_A1, then A2 and still got the error. Test DIMM_B1 and B2 and system posted fine, which I found really odd.

So thought mobo much just be really shot and started to rebuild in my old board, only to then get the same error in the same DIMMs. Really had me confused as couldn't see any damage to sockets on either board and every HAD been fine prior to the swap.

Went out and bought a new PSU and CPU (i3570k) and tested it in both boards and they both POST with RAM in any slot. This has led me to thinking the memory controller might be shot.

Now the ASUS was overclocking the 2500k by default with some of the bios/mobo switchs and I did notice it pumping a bit more volts than it needed to, but when it did I would restart and turn it down - it certainly wasn't running for long but CPU-Z did show 1.6 at one point and i'm thinking this might have done it, even if it was only for a min or two.

Has anyone ever had anything similar? Should I RMA the chip back to Intel?

 

[dandragonrage]

If your board had your chip at 1.6v for even a split second, regardless of whether you told it to or not, returning it for an RMA is unethical. But it does sound like that CPU is toast. No modern CPU should be run at over 1.4Vcore. Some motherboards do pump a ton of voltage in auto mode and that is absolutely unacceptable - but it's not Intel's fault whatsoever.

Also, what voltage is your memory rated for, and what were you running it at? 1.65v is the absolute maximum for modern Intel CPUs, but even then I highly recommend using 1.6v tops if possible. I typically recommend people get rid of 1.65v memory if it won't work at <=1.6v. The memory controllers on these chips can be damaged at higher levels and may be what happened. And if so, again, it's unethical to RMA it to Intel. Whether they would replace it if you RMA'd it anyway, I don't know.

Personally, if I could prove that the Asus mobo ran that kind of voltage in auto mode (which might be difficult), I would be going after Asus for this. I would possibly go as far as speaking to a lawyer about it, though even that's not guaranteed to get you anything. I doubt they'll accept responsibility for anything, especially since you got it used. But if their motherboard ran 1.6v in auto mode, there's no doubt that it caused issues from poorly written BIOS/EFI code.

 

[leeleatherwood]

Reseat your CPU, it might not be making good contact with all pins.

I thought I blew the IMC on my 3570k because I had my ram at 1.6v and then my ram would stop being recognized, I swapped the ram to the other slots and it worked fine.

About a month ago I was doing some PC maintenance and reseated my CPU, now all of my ram slots work.