Hello all, Long time lurker and many of you have helped me without knowing it already, so thank you Hopefully someone can come to my rescue with my new build? I've just put the below together, hit power and then encountered a boot loop. Everything spins up fine, runs for a few seconds and then dies. A few seconds later it will try to boot again and then die again. Repeat ad infinitum Nothing shows on screen at any point. This isn't my first build and I have double checked and re-seated everything as far as I can tell. Anybody know what might be causing this? i5 2500k MSI P67A-GD53 B3 2x2gb Corsair XMS3 1600mhz 1.5v 1TB Spinpoint F3 Sapphire HD 5870 Be Quiet! Dark Rock Advanced Be Quiet! Dark Power Pro 650W
First thing I'd do is pull all of your RAM sticks and boot. If you have an internal speaker on your motherboard (not sure how many do these days) you should hear some beeps, confirming that the board is working as it should. If you don't have an internal speaker, have a root around in the bits bag that came with your board, MSI may have bundled one in there. If either your board beeps or you don't have any kind of speaker, try all RAM sticks in the same slot, one after the other. Once you've tried all four sticks, add your RAM back one stick at a time. If your machine refuses to boot at any stage, repeat the last step you did (swap to another stick and back again) three times to confirm. Also, check your EPS power connections, and unhook any fans. If you have a spare PCI-E GPU, try that - even if it's an old banger.
Whilst you 'should' have been fine since it's relatively bog standard memory, i know with my Asus SB build i had to first boot with a single stick (2nd slot from the CPU) -> alter the settings in the bios --> & then switch off & install the other stick(s). Simply that, whilst there could be other causes, this appears to be a common issue/solution across manufacturers.
Dead stick of RAM it seems. Swapped them round but didn't think to try with one at a time for some reason! Probelm solved and I'm posting from my new rig - idling at 23-27°C across the cores no less! Thanks for the responses guys. Anybody know what Scan are like with dead on arrival sticks?
Glad the problem's sorted Returning really depends on how you bought them - individually or as a pair. If you bought them as a pair, you'll need to send both sticks back, which is a pain but necessary as Scan will need to RMA them with Corsair as a pair.
I could always order another pair and wait for them to turn up before returning I suppose. Either that or I'm without my PC for a while. Annoying to be stuck with 2gb, seems so sloooww
ring scan and tell them the problems and how you need your computer working. im sure like most companies they can send memory out and collect via a courier service
You could order another 4GB pair, then send the faulty pair back. When you get the RMA replacements, you can either keep the extra 2GB or sell it on eBay or the BT Marketplace (you've got a while to go before you get access though). I would personally keep then and run 8GB
i had a ram issue an when i spoke to scan they said to me all they can do is for me to return it then wait for another set to come in, or i can buy another set of ram and send the not working one back and wait for a refund. as i bough the vengance kit i would have had to send it all back and the pc would have been down for about a week.
Well, called into Scan today and it's not a dead stick. Both booted fine in their test rig. Taking my mobo back in tomorrow for them to test that. Bit of a ballache having to go back but I'll get a replacement board if they can't get it to boot. Either that or I'll ask to try a different kit of RAM and see if that works. I would be interested to know if anybody has the same mobo and what RAM they are currently using? EDIT: Re-read that and it could come across as moaning at Scan. I wasn't as the fella was perfectly helpful. I should've thought on and taken everything in the first place.
So it seems these boards aren't too keen on these: http://www.scan.co.uk/products/4gb-...0-(1600)-non-ecc-unbuffered-cas-9-9-9-24-150v I could have been doing something wrong but I've just picked up these from Aria and am now running fine in dual channel: http://www.aria.co.uk/Products/Comp...2x4GB)+DDR3+1600MHz+9-9-9-24+?productId=43961
Now I'm up and running I'd appreciate some help with a couple of things. -I'm using the guide on here as a base for hitting an oc of 4.5ghz. One of my cores will stop at around a minute under prime95 at this clock. Am I correct in thinking I just need to edge my vcore up slowly to fix? I've bumped up to 1.32 so far and don't want to go any further without being sure. I know it's nowhere near long enough but at 4.4ghz I've just ran prime95 fine for 20mins. Temps maxed out at 59 degrees. - Do I need to worry about a 5 degree temp variance across my cores? Or is this fairly normal? Obviously I would like all of them to show the lower temp but the cooler would be annoying to refit if I don't have to.
the varience shown is due to the position of the thermal paste and contact with the cooler. resiting it may give less varience but 5c is not alot in varience terms. depends on how picky you are. its very common to have some varience. for everyday use the max temp should be around 60-65c so you can increase the volts until its stable at 4.5ghz. but weigh up if having the cpu at 100mhz faster is worth those degrees ? find a sweet spot which seems to be 4.4ghz for you and keep safe within margins or upwards. i think 73c is the limit for a i7 2600k. also test using prime for abit longer then 20 mins in your gaming/working atmos. so door closed and window open. leave for an hour then come back. the room will heat up and so will the cpu. so some le-way should be left incase of something like this.
if you refit the cooler you 'could' end up with even worse results - or much the same results but a different higher core... With all of the builds & rebuilds i've done then it's only my current SB build where, very unusually, all the cores peak at the same temp with ~99.9% load, & personally wouldn't bother for 5C. As to the voltage, personally i'd be happy going up to ~1.45V (when i've got the time to carry on playing) if there was a significant gain above the very quick 4.8 i got... Though that's with my specific 2600K, & where it's my money that was spent on it/would have to replace it. Now, the problem comes that just because your processor is a 2500K, it doesn't mean that if someone's got an o/c of X at a certain voltage on their one then yours will; as each chip is different. Well, (a) there's no idea what its default voltage is (& this can very significantly) so this alters things somewhat, (b) each cpu will then have maximums at intermediary voltages & (c), whilst you 'could' stick (comparatively) very large voltages through 'if' you had some über extreme cooling solution to try & get it that little bit further, the specific cpu itself will have a more general max clock despite the voltage. So really it's 100% your call as to how far you want to try & push it - but, whilst i'm not responsible for your processor (as again it will vary from other ones & yours 'could' be unusually prone to failure), i really don't think you've got any reason to worry about upping the voltage a bit to get stability.
Thanks guys. With the temps it was more size of the gap, so I'll live variance if it isn't wide enough to worry about. Just used to my old E6750 matching on both cores. I had a max temp of 65c in mind when starting out from reading around, so it seems I have a little bit of headroom to play with to hit 4.5 stable. I'll up a couple of notches and see where it leaves me. If not I'm more than happy with 4.4 - although 4.5 is such a nice number to satisfy my OCD tendencies I just needed to know I was on the right track and it wasn't something else causing a core to stop.
