Originally Posted by psyside
Anyone, ever experienced issues with running memory on lower freq then before? it was @ 2133mhz, and i read something about that high memory freq limiting the oc, and put stree on the IMC, so i decided to lower it 1866, but as fast as i booted up like 1 min after, my pc just freez
So long story in short, lower ferq causing instabillity isnt this super strange? also in past i was able to use higher ddr/vccio voltage, now it becomes really unstable if i touch any of this, (vccio auto) (dram 1.5) i really got no idea whats going on lol!
Not really, the IMC on the SB can withstand quite a bit, and I personally don't think that i limits the overclock, I mean look at some of the submissions that have 16gb, some have hit 5ghz stable. The only reason I can think of is that maybe the voltage needs to be lowered when you take the frequency down with it, sometimes too much voltage can cause instability.
Originally Posted by Overclocker.Monster
Hi guys, I was having a doubt, cuz I setted up my Vcore in 1.330 and when I was testing it with prime, it drops like to 1.310, when it's IDLE is about 1.328, 1.320... is it normal? is there anyway I could set it to mantain the Vcore to what I set up on BIOS?
the drop in voltage under load is called vdroop, it can be eliminated or reduced when using LLC. When the voltage is higher than what is set in the bios at idle that is known as vrise and not really important. The voltage under load is what is important, it's usually recommend to use LLC to make sure that the vdroop is at a minimum but doesn't cause vrise under load which obviously creates voltage spikes that can be dangerous when using high voltage.
Originally Posted by DEEBS808
I get this message when loading prime95.Is this normal?Will this affect my results?How can I change it if it does?Thanks
Not sure bud, it should however get through the testing just fine as long as the cpu is under 100% load.
Originally Posted by slimex
here are finally some Sisoftware Sandra benchmark results:
I think each result should be compared separately. There should be ~snip.....
I'll add the scores in a sec, I didn't really put much effort in the scoring system, maybe it could be improved but the benching part is just a bit of fun, however I'll try and see what I can do.
Better yet PM me with some kind of scoring system and we'll see how we can make it work
Originally Posted by leonredbone
If speedstep is disabled, should the processor stop cycling through different clockspeeds? I may be misunderstanding this tech. Also, should it matter if I use speedstep? It is going back and forth between 1600 and 4500 alot
On a sidenote, my system is STILL not stable with BSOD 124 occuring usually within an hour of being powered on. I don't know what else to do, I have tried going up in the VTT and PLL, andI have tried going down....no luck.
Vcore is now at 1.32 so I don't think the CPU isn't getting enough juice.
VTT 1.15 (Tried higher and lower than this)
PLL 1.83 (Tried lower than this)
C1E and Speedstep allow the multiplier to drop, the C3 and C6 states afaik are related to sleep and other safety procedures. THe should be left on auto when using manual voltage, however, with diffrent mobo's the BIOS settings are a little different so it may take some time to figure it out. If your mobo is not an asus 1155 mobo, contact some of the members here that have submitted on the spreadsheet that might be able to give you some insight. Alternatively check out the BIOS templates available to see how the Cstates should be configured.
I've also created a thread here just for BSOD 124 and the Random Idle BPSD problem: Solving / Fixing BSOD 124 on sandybridge. READ OP FIRST!!
Originally Posted by phinos
I have a question, I passed 30 mins run of 1344, 1792 FFT tests. and Doing 7th hour of blend test atm. but I just noticed that I have RealTemp 3.69.1 instead 3.67... Should I re-run the test with 3.67 to be accepted?
Edit: I just decided to re-run the test with 3.67.. But can't help wondering if there's any problem with RealTemp higher than 3.67.
Anything above 3.67 is fine. Thanks for reminding me, rules have been amended. Apologies I didn't make it clearer from the start.
Originally Posted by jdip
Hi guys, I just built a new PC a couple of weeks ago and am new to overclocking.
I am having trouble overclocking my i5 2500k and keeping it stable. My goal is 4.5 GHz, I have gotten this high and stress tested with Prime 95, where my system will BSOD and reboot at 12 hours or so. From what I understand, to overclock these SandyBridge chips, we just have to adjust the multiplier and then add voltage until it is stable, right?
The maximum Vcore that I have used so far is 1.36V, and it still crashed in 10 hours or more of Prime 95 blend testing. From what I have seen, many people are getting 4.5GHz stable at around 1.3V, so this seems kind of odd to me. Also, my temperatures seem a bit high. I have a CoolerMaster Hyper 212+ and at 4.5 GHz and 1.36V I was getting max temps of around 79C on one of the cores. Doesn't this seem a bit high?
I have also tried overclocking to 4.0 GHz and leaving Vcore at auto. CPU-Z showed that at 4.0 GHz and Vcore at auto, the computer was assigning a Vcore value changing between 1.288V and 1.328V. The strange thing is, with 4.0 GHz and Vcore on auto, my computer was still failing (bluescreen + reboot) the Prime 95 blend test at 12+ hours.
I don't really know what to do from here, any help would be greatly appreciated.
My hardware is as follows:
MSI P67A-G43 Motherboard
G.Skill Ripjaws X 8GB DDR3-1600 9-9-9-24 RAM
A-Data S511 120GB SSD
WD Black 1TB
Powercolor Radeon HD 6950 2GB
Running Windows 7 Ultimate 64-bit
Compare temps and voltage with other in the spreadsheet. Your Mobo is not exactly the best for overclocking due to the 4phase VRM design. Have a look at the guide I posted in the OP.Edited by munaim1 - 9/25/11 at 4:36pm