The I7 2700K BSOD occurred after 30 minutes using Prime95.
The I7 2700K BSOD occurred after 30 minutes using Prime95.
Hi
I just swapped out my old board after it failed, but I’m struggling to match the overclock I used before. I was stable at 4.3 GHz with 1.28V, yet now achieving a 4.2 GHz overclock for more than 30 minutes at 1.3V isn’t possible, even in prime mode. Is this a significant difference or could another factor be involved?
Also, I switched from a Corsair H70 liquid cooler to a Hyper Evo 212 setup because the new board’s back plate doesn’t fit (mini ITX to micro ATX). Idle temps are high in the 20s to 30s and it climbs to 60-70°C during prime. I haven’t seen any crashes during regular use, but Prime crashes are a concern.
Old board: EVGA Stinger Mini ITX
New board: ASUS P8Z77-M
CPU: I7 2700K
Thanks.
Where should I cease with the voltage? What's the upper limit I should reach (if it gets there)?
I'd suggest not exceeding around 1.35v. The board could face issues even before that. Some might think 1.4v is too high, but there are many who have surpassed it—though I view that as a bit excessive. I'm more of a cautious OCer. Your performance will likely be constrained by temperature depending on your cooling setup.
Tom handled an i7-2600k with those settings for reference; remember, results may differ. Aim for a stable clock using the lowest voltage necessary to keep the chip fully stable at that frequency.
It might be that the voltage system isn't quite as refined. Your temperature readings don't appear to be a major concern. Memory compatibility could also contribute, though I thought the CPU's memory controller would mitigate this. It might still help to confirm your memory configurations are identical. Testing with more relaxed timing settings could also rule out a memory-related problem.
Some users suggest certain CPU socket types perform better. If that's accurate, it could indicate your current setup doesn't provide optimal contact with the CPU pads. You might consider cleaning the pads with an eraser if they look dirty or corroded and retest.
Additionally, it's possible that higher clock speeds since the chip was installed have led to slight performance degradation, causing it to fail testing despite a stable voltage and temperature. I don't believe your voltage or temperatures are unusually high for a 2700k model.
I'm puzzled why an mATX board would restrict cooler installation.
From a brief inspection of both boards, the Evga model clearly stands out as the superior OCing choice. The Asus board falls short in this regard. It's curious how the backplate fits the mini-ITX unit but not the mATX one. Still, when using the Asus board, keep an eye on the VRM temperature—those reported variations are often linked to system-level voltages. The lack of strong voltage regulation compared to the Evga board seems to be a real issue you're encountering firsthand.
It seems the voltage system might not be as refined as expected. The temperatures appear to be fine, but memory compatibility could still be a factor. I believe double-checking your memory settings might help. It might also be worth trying looser timings during tests just to rule out a memory problem. Some users suggest that certain CPU sockets perform better, possibly due to better contact with the CPU pads. You could try cleaning the pads with an eraser if they look dirty or oxidized.
It's also possible that higher clock speeds since the chip was new have led to slight degradation, affecting performance. This doesn't seem likely given your voltage and temperatures are within normal ranges for a 2700k. I'm surprised an mATX board would restrict cooler installation.
Running XMP should position memory optimally, but mismatched RAM (like 2 x 4GB Corsair Ballistix Tactical and 1 x 8GB Corsair Value) might be the issue. The BSOD message about a clock interrupt not being received suggests instability in OC settings. My OC was stable at higher clocks on my old board, so this could explain the change.
I'm considering increasing the voltage to above 1.3V, though I've never tried that before. Cleaning and re-seating the CPU might be necessary if temperatures are abnormal. Regarding the liquid cooler bracket, it likely came with a micro ATX design; since I obtained it from someone else, I probably need a new one with a proper back plate.
The brackets and backplates must match for mini-ITX or mATX, as they depend on the socket type. This is what I've observed.
I was discussing cleaning the bottom of the CPU, not the heat spreader.
The brackets need to be identical. I have a mITX board and the cooler bracket functions on all boards that support sockets. I had to modify part of it due to backside components on my mITX board.
Based on a brief inspection of both boards, the Evga board stands out as the superior OC board. The Asus model isn't ideal for overclocking. It's curious that the backplate fits the mini-ITX but not the mATX. With the Asus board, keep an eye on the VRM temperature when overclocking—those temps are often given as system readings. It clearly lacks strong voltage regulation compared to the Evga board, which might explain your experience.
I'm sure you can understand that. The Stinger was perfect for me, and I'd have chosen another if available. The Z77-Z97 1155 MATX and Mini ITX boards are extremely rare and usually sold at low prices. I ended up getting this one because it was the last 1155 Z77 in stock. So I suspect I'll need to increase the voltage beyond 1.3v, right?
The plate fits perfectly, but it doesn't stay in place. It aligns just right and can be inserted, though it doesn't have any fasteners. With the Mini ITX board, it sticks without needing anything attached.
i was discussing cleaning the bottom of the cpu, not the heat spreader.
the brackets need to be identical.
my mITX board has a cooler bracket that fits all boards with socket support.
i had to modify part of it due to backside components on my board.
didn't we ever need to clean the bottom of a cpu because it's so delicate?
I definitely understand that. The Stinger worked well for me and I would have chosen another, but none are available anywhere. Z77-Z97 1155 MATX and Mini ITX boards are extremely rare anyway and end up being sold at unreasonable prices. I obtained this board because it was the 1155 Z77 available. I think I’ll need to increase the voltage beyond 1.3v, correct?
It seems like it requires more power. If you haven’t already, try using load-line calibration (LLC) to check if that helps.