New Rig initial setup! MSI GAMING 7 4790K

Well I got bored of my old rig and was off work a couple weeks due to a broken shoulder, and I caught the upgrade bug.

started with a couple 970s, and didn't like the results on old sandy bridge win7 setup.

So here is new rig.

4790K
MSI Gaming 7 (liked board layout and cost).

Hyper 212 evo with Corsair Fan
16gb Corsair Vengeance 2400
500gb Samsung Evo
970s SLI (once rig changeout is complete).

I assembled rig onto my benching station with the 670s sli since they were pulled already.

I like keeping my main rig fully functional while working out the kinks.

Booted up first try no weirdness. 8.1 install is funky do not like much learning and getting win7 look feel right now.

Core Temp works better than Real temp for Reading Intel Core (2/celly/penty/i3/5/7) temps.
It's generally agreed on this forum that Prime 95 is the go-to stressing software to use. Run it for a good 6 hours and you'll be sure you're stable. Run "blend" mode.
How come you didn't go for a board with a PLX chip so you could run 16x/16x on the GPUs?

Try not to go nuts with furmark. It's bad for your GPU.

furmark was for cpu load, I don't like prime95 for testing right now due to weird overheating issues. the cooler was in the spare parts bin and not the greatest but I had it.... Realtemp is spot on with mobo temp readings, I have a readout on mobo for temps and its exactly what realtemp is showing..so...going to let benches be my stability testing right now.

So it seems above 4.6 voltage and heat goes up exponentially what are you guys running vcore and oc on these?

There are no overheating issues. Prime95 is simply using all features of your CPU and use it during all test while in real-life applications it's activated only during some random calculations so temps are much lower. If you are using better cooler then it should work without problems. Other test can be XTU which is using older instructions ( still base on Prime95 ) but its new version is finding errors faster even though it's not heating cpu so much.

I'm rarely running Prime95 28.5 for longer than ~30mins. If something isn't right then it's crashing in 5 mins or something near.
My first 4790K was running @4.6GHz 1.19V, 4.7GHz 1.228V and 4.8GHz 1.28V but it died and I still can't explain why. The one that I got from RMA is running @4.7GHz 1.32V but most benchmarks is passing at about 1.27-1.28V.

So it all boils down to the crappy IHS deal then? Why can't they make sure heat transfer is optimal, with the lid ON.

theELVISCERATOR said:

So it all boils down to the crappy IHS deal then? Why can't they make sure heat transfer is optimal, with the lid ON.

Click to expand...

Lots of reasons. IMO it's more than anything to level out SB-E/IB-E/HW-E Vs SB/IB/HW. SB could very often OC to 5Ghz on big air. SB-E couldn't. There was a disparity. For IB/IB-E they levelled out the playing field and made it so both CPUs could only OC to about 4.5-4.6Ghz on average "big prebuilt" cooling. To accomplish this they had to ditch the solder on the IB and go with TIM. They kept the solder on the big gun CPUs because they have a much higher TDP.

This is just my theory, but it makes sense. They don't want 5Ghz Haswell refreshes competing with 4.5Ghz Haswell-E chips because in a lot of apps that only use 4 cores, the HW will beat the HW-E and they don't want that happening. JM2C.

Got rig off the oc bench and into the old 1200, changed out psu for HX750 I had, had to go modular finally, system is up and running and running pretty good all the way round.
like the tons of usb 3.0 ports. Moved sb into new case and fresh installs!