Ibt or p95 ver 28.5 have hit temps in the mid 90's on both 4790k's ive had, stock speed with voltage of around 1.26 at load. This is why i just use real bench, the latter are just way too sore on temps. Also real bench reflects real tasks more than prime etc.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
5ghz
- Thread starter Guest2
- Start date
I have just re-seated the HSF thinking it was that (chip not touching the pad properly) but still the same.
Temps initially go to ~65 then jump to 100 after about a minute and I notice the voltage on auto also jumped to 1.285v at the same time. This is with stock auto settings
Using IBT v2.54.
If my config is corret and temps are 100c then Intel or the people who created IBT need to put a disclaimer on their software as it could potentially easily break a lot of CPUs
Will try real bench now
Thanks
Temps initially go to ~65 then jump to 100 after about a minute and I notice the voltage on auto also jumped to 1.285v at the same time. This is with stock auto settings
Using IBT v2.54.
If my config is corret and temps are 100c then Intel or the people who created IBT need to put a disclaimer on their software as it could potentially easily break a lot of CPUs
Will try real bench now
Thanks
Associate
- Joined
- 15 Feb 2014
- Posts
- 778
- Location
- Peterboghorror
How many people saying not to use IBT does it take before you stop? Just in case I will add my name to it, use Realbench, seriously.
People want to know their chip can take the worst beating possible.
When I did the same thing (same board/bios/cpu) I found the vcore was ramping up when the load was. It would get near 1.3v and temps were also pushing 90+ with a Corsair H110GT. That was even with just the 4.4Ghz standard 'boost' that it tries.
I set the CPU to 4.4 manually and the vcore voltage manually in the BIOS and tested, temps dropped.
Then just worked backwards on the vcore dropping it bit by bit and running stress test and the drop in temp was pretty huge each time. Finally settled on 1.15 which was fine and I got bored of running tests.
It was stable on it for months, I had a tiny hiccup on planetside 2 and I bumped the vcore to 1.17 and that went away as well.
Long story short, auto volt increase really ramps up the heat.
Tbh, when you run something like ibt/p95. The cpu is more than likely to start throttling, so doesnt give a great indication of stability.
Below is a screenshot from one of the 4790k's ive owned. Attempt at 4.7ghz on 1.2600v set in bios.
The temps were after a few seconds of p95, (similair at stock too). That chip ended up needing 1.31v to complete real bench without issue. At that voltage p95 would have generated higher temps. The one in sig was slightly hotter but more even across the cores at stock.
The temps were after a few seconds of p95, (similair at stock too). That chip ended up needing 1.31v to complete real bench without issue. At that voltage p95 would have generated higher temps. The one in sig was slightly hotter but more even across the cores at stock.
Tighter than the difference between mine. Mine show about a 10c spread, which is pretty typical from what I can gather.
I'd recommend a nice long realbench stress test. Its failed for me after 4 hours or so. I understand a lot of people dont care about that as long as its stable in what they do, just my 2c, not saying my way is the right way.
I'd recommend a nice long realbench stress test. Its failed for me after 4 hours or so. I understand a lot of people dont care about that as long as its stable in what they do, just my 2c, not saying my way is the right way.
Is it normal for vcore to auto to 1.236 under load and 79c when left at stock settings? (4ghz turbo to 4.4ghz) Seems a bit high, especially as I am using a noctua nh-u14s. If I left the Intel stock cooler installed it would get roasting hot and probably throttle when running RealBench
Lapping might tighten up the range. If the contact plate is curved due to sloppy manufacturing or Q&A.
Not that bothered to be honest. I'm now perfectly content with my 4.7GHz overclock @ 1.214v. I don't care about more as I just don't see the point. Voltage goes up loads, temps follow. Could delid, but 4.8 is all I would get under 1.3v, so why bother?
FYI, I lapped my e6600 for the lols back in the day.
Last edited: