Getting my 4790K past 4.5 on MSI Z97 Gaming 7

nails666

nails666

nails666

Associate
Joined
17 Jan 2007
Posts
1,142
Location
Kent, UK
Hi, I've finally got myself a new 4790K for my rig and I've started playing with a bit of overclocking but I've not got very far.
Now first up I'm not going for any benchmarking records, I don't need every last MHz out of the chip, it's more just something fun to try but at the end of day I need the system rock solid stable and to not do any damage to the chip :)

Anyway, so I previously had a Pentium K in the rig and got that to 4.6GHz simply by setting the multiplier to 46 and the VCore to Manual at 1.25v, everything else left on auto. So I thought I'd try the same this time and have got 4.5 with 1.24v (shows in CPU-Z as about 1.27 under full load) and again, everything else at default or auto. I'm using RealBench stress test to check stability and it's run for 2+ hours fine with a max temp of 74 degrees (I have a Corsair H75 with SP120 fans in push/pull for cooling) as reported by Intel XTU. So next I tried going to 4.6 and it boots to windows but BSODs within 10 seconds of RealBench. I've tried upping the voltage to 1.265 (ends up about 1.285 in CPU-z) but with the same result and I'm not sure how much further it's safe to push the voltage? In the few seconds of running I see the temps go to about 75 but then it crashes.
So what are my options now, push the voltage further but if so how far should I go?
Are there other settings in the MSI bios that I should set manually rather than leaving on auto?

Many Thanks,

Ross
 
What's the code for the blue screen error. Could give us an idea on what may need adjusting.
 
Hmm, didn't take a note last time I saw it, I'll try and have a camera on hand next time I try some tweaking to catch it and report back.
 
Cheers, a very handy tool. According to that all of my BSODs have been for the same reason.

On Fri 03/10/2014 13:21:39 GMT your computer crashed
crash dump file: C:\Windows\Minidump\100314-6770-01.dmp
This was probably caused by the following module: hal.dll (hal+0x12A3B)
Bugcheck code: 0x124 (0x0, 0xFFFFFA800744C028, 0xBF800000, 0x124)
Error: WHEA_UNCORRECTABLE_ERROR
file path: C:\Windows\system32\hal.dll
product: Microsoft® Windows® Operating System
company: Microsoft Corporation
description: Hardware Abstraction Layer DLL
Bug check description: This bug check indicates that a fatal hardware error has occurred. This bug check uses the error data that is provided by the Windows Hardware Error Architecture (WHEA).
This is likely to be caused by a hardware problem problem. This problem might be caused by a thermal issue.
The crash took place in a standard Microsoft module. Your system configuration may be incorrect. Possibly this problem is caused by another driver on your system that cannot be identified at this time.
 
0x124 can be vcore related, or chipset voltages. Also make sure cpu input voltage is set at least 0.5v higher than vcore.
 
not the same board but WHEA_UNCORRECTABLE_ERROR and CLOCK_WATCHDOG_TIMEOUT were the two errors I was fighting against looking for stability. Stability for me means getting through at least 10 hours of intensive x264 encoding of my gameplay videos maxing out all the cores the entire time. having the system crash after 9 hours or so as the file is almost complete is heart breaking ;)

I takes 1.29v (1.308) to stabilise 4600. 1.300v (1.320) for 4700. LCC set to 'Turbo' Which keeps VRIN a constant 1.800 as default in bios under load. (1.744 idling)

for 4800 I've had to use 1.330v (1.344) and VRIN set to 2.000 (although I may be able to bring this down, needs more testing). cpu temp is around 80-82 with my H105 after a few hours of encoding.

uncore has happily stayed synced although my board wont let my change vring voltage in case I needed it. Change one digit and I get a crash loop at post. So it has stayed at the default 1.050.

I haven't really looked at other voltages. XMP profile could be setting some of those.
 
Last edited:
hmm, what voltages are regarded as safe for long term, 24/7 operation on these chips. I thought with the 4770k 1.3v was suggested as max?
 
Ok well I'm trying 4.6 now with 1.28v (1.304v in CPU-z) and VDROOP offset set to 100% and it is now running Realbench stress test so it's an improvement at least. Temps are nudging 77 degrees so still just about acceptable. Have to give it a more extensive run before I can claim full stability though.
 
Whats your cpu input voltage, (aka VRIN) at? Found this to be pretty important with haswell, ideally it should be at least 0.5v above vcore. For 4.7ghz im on 1.300 vcore, 1.800 input.
 
1.25 is too low for VRIN, its normally 1.7 ono at stock. With vcore at 1.200 try 1.800 on VRIN.
 
setter said:
Whats your cpu input voltage, (aka VRIN) at? Found this to be pretty important with haswell, ideally it should be at least 0.5v above vcore. For 4.7ghz im on 1.300 vcore, 1.800 input.
Click to expand...

Well 4.6 crashed after about 45 mins in RealBench so still a bit short of volts I suspect. Hopefully get a chance to try a little more tonight. My VRIN is still set to auto but I think it was reporting 1.8 in the bios, I'll have to check it later.
 
Ok, got it stable at 4.6 now with 1.285v, passed 2 hours of RealBench Stress test and a full Bluray conversion in Handbrake with temps topping out at 79 degrees.
I might give 4.7 a go but I suspect I'm going to hit the limits of my cooler now so 4.6 might be my lot.
 
You must log in or register to reply here.
Back
Top Bottom