Secure continuous overclocking and stable temperatures for the I7 4790K.
Secure continuous overclocking and stable temperatures for the I7 4790K.
The results are solid. Maintaining stability at 4.7 with 1.225 is confirmed. Given the need for a 40 to 50 millivolt increase to reach 100MHz, it seems reasonable to expect stability at 4.8 with 1.275. Core temperatures should remain below 80°C, and P95 v26.6 Small FFTs should operate smoothly at 22°C ambient. The motherboard power rating provided appears incorrect, so it's best to disregard that. Focus should remain on keeping voltage under 1.300 volts and temperature under 80°C.
I actually experimented with 1.25v and 4.8 Ghz, installed it in Windows and ran a partial stress test, but the temperatures rose to around 85 degrees and then I saw a blue screen.
I’ve observed that my VTT voltage is above 1.775, and I’ve seen references suggesting it shouldn’t exceed 1.4 (though I’m not sure if those details are accurate or current).
Do you have any advice on this? Or does anyone who has modified their i7 have any experience? I also have XMP enabled for my RAM, and I’ve read that VTT applies to the memory controller in the processor, but my RAM only requires 1.5v—so I’m uncertain about the correct VTT setting now.
Apologies for all the questions! I just want to be sure my system is functioning properly.
jordan1794 :
Reminds me of the kind of OC I might have had with a water cooler...
But perhaps it's wiser not to try anything, lol.
Your OC is limited by the vcore you can handle, not by the cooler.
Only the vcore above 1.3 needs extra cooling, and that's not a sustainable long-term solution.
geofelt :
jordan1794 :
Reminds me of the kind of OC I might have had with a water cooler...
But perhaps it's wiser not to try anything, hehe.
My OC is limited by the vcore I can handle, not by the cooler.
Only a vcore above 1.3 needs extra cooling, which isn't a sustainable long-term solution.
I think I could probably reach a clock speed of 4.9 with a vcore around 1.3 or higher.
But since I'm seeing temperatures hit 85 at 1.25v with 4.8GHz, pushing up to 1.3v would definitely get me into the 90+ range with my current cooler, which I'm not ready to risk.
If I were to go over that, I'd need to set up voltage offset so I wouldn't be stuck at that voltage constantly.
Anyway, I'm just reinforcing to myself that I shouldn't push it like this lol.
As discussed earlier, a rise of 40 to 50 millivolts would be required for a 100MHz increase, so I wouldn't anticipate stability at 4.8 with just a 25 millivolt change.
In my previous message, I mistakenly thought your core temperatures were near 70 instead of 80; I apologize for that error. I often follow temperature discussions, and it can be tough to remember everything after stepping away from the forums.
Nonetheless, it appears you were indeed achieving your optimal overclock at 4.7GHz. Remember, the gap between 4.8 and 4.7 is only 2.1%, which is negligible and hard to notice.
CompuTronix :
As previously discussed, a 40 to 50 millivolt increase is required for a 100MHz boost, so I wouldn’t expect stability at 4.8 with just a 25 millivolt rise.
In my earlier comments I mistakenly thought you were near the 70°C mark instead of 80, so I apologize for that error. I often follow temperature discussions and can struggle to remember details after stepping away from the forums.
Nonetheless, it appears you were indeed achieving your optimal overclock at 4.7GHz. Remember, the gap between 4.8 and 4.7 is only 2.1%, which is negligible.
jordan1794 :
I’ve observed that my VTT voltage exceeds 1.775, and I’ve seen notes suggesting it shouldn’t surpass 1.4. I have XMP enabled for my RAM, and I understand it’s meant for the memory controller in the processor. My RAM only requires 1.5V, so I’m unsure what my VTT should actually be.
About VTT – the termination voltage for memory – it’s usually best to keep it on “Auto.” Since I haven’t shared your motherboard model, I couldn’t verify the exact setting. However, going beyond a +0.25 increase over the default isn’t advisable.
Typically VTT should be half of VDIMM value, which in your case is 1.5, so it should be around 0.75. Adding a small margin would bring it to about 0.775.
Are you certain you meant 1.775 rather than 0.775?
This likely isn’t your motherboard, but you might want to check the manual:
GIGABYTE Z97X OC Guide
- http://www.overclock.net/t/1490835/the-g...king-guide
CT
VTT is shown as 1.775 in HWMonitor:
Here are my BIOS settings for CPU voltage, if that’s useful.
I doubt HWMonitor would be accurate. Could you confirm the Gigabyte motherboard you're using? VTT might appear in BIOS under memory controls instead of CPU settings. Are there any relevant voltages listed, like Memory Termination Voltage or VDIMM/VDDR Termination Voltage? Also, does your BIOS include an option for VTT?
Only alternative voltage configurations are managed by the chipset, specifically PCH Core and PCH I/O. Both auto and core settings were 1.090, while I/O was 1.5. I’ll review my "classic" BIOS and adjust this answer if more settings are found. (I have a "smart tweak" mode and a "classic" mode.)
In my regular BIOS, the voltages under PC Health Status showed VRIN at 1.774, which matches the VTT reported by HWMonitor—possibly because HWMonitor is displaying this value when my VTT isn’t available.
I’ll check the VDIMM voltage with a multimeter, though I don’t have one. Maybe I’ll get a cheaper one to test next week...
How likely do you think HWMonitor is showing my VRIN as my VTT?
jordan1794 :
This is essentially the same as what HWMonitor is reporting as my VTT, perhaps HWMonitor is accidentally reporting this voltage because my VTT voltage is not available?
How plausible do you think it is that HWMonitor is reporting my VRIN as my VTT?
As I mentioned in my previous post, I wouldn't count on HWMonitor being correct. Like your Gigabyte motherboard, my ASRock has no BIOS setting for VTT, so it's a non-issue.
What clock speed is your memory?