Ever heard of a motherboard model that eats ram?

[moocow]

Over 1.5v is just fine. I've seen guys on some forums run their memory at 1.75v for daily use without any issues.

 

[Brando]

Over 1.5v is just fine. I've seen guys on some forums run their memory at 1.75v for daily use without any issues.

Idk man this thread is worrying
http://communities.intel.com/thread/30798

I'm thinking about selling the gskill when I get a replacement set and trying a different brand. This muskin looks nice and has far more positive reviews. Maybe my board and cpu just don't like doing lower latencies or something.

http://www.newegg.com/Product/Product.aspx?Item=N82E16820226245

I couldn't find anything that says this specific mushkin set is compatible with this motherboard but I'll try calling them tomorrow and see what they say. I also wouldn't mind changing to their brand because gskill is a little sheisty about cross shipping (money orders only, no credit cards) which is annoying and probably doubles rma time if I want to avoid down time. Mushkin seems to be easier to deal with in that department according to their policy on their site.

 

[moocow]

Yeah, and Intel also technically doesn't support overclocking.
Chips are much stronger than people give them credit for.
I've been running a 3770K since launch with memory at 1.65v for daily use.And I've even given memory 1.9v while benching with no degrading at all.

 

[Brando]

Yeah, and Intel also technically doesn't support overclocking.
Chips are much stronger than people give them credit for.
I've been running a 3770K since launch with memory at 1.65v for daily use.And I've even given memory 1.9v while benching with no degrading at all.

I guess if somebody is going to know about the limits of hardware it'd be somebody on the benching team. Touche'.

 

[Brando]

I think I'm finally getting some love from this ram. Bios 1406 that JUST came out has an xmp profile #2 that ironically increases stability by tightening up some sub timings. Got about 30 min in prime blend with max ram and no errors. Gonna run memtest overnight. This is the most I've ever had to tinker with anything to make it work. Mother of god.

 

[Brando]

One hour of prime95 blend and 8 hours of memtest 86 with no errors. Weird weird weird. Apparently it needed tighter sub timings. WTF I guess alls well that ends well. Thanks everybody. I'll try this for awhile and see if it holds.

 

[larrymoencurly]

By "under each screw hole" do you mean on the back of the motherboard between the plate and the mobo around the little peg that goes through the hole? The screws don't actually go through the motherboard because the pegs that that screws go in to stick through the motherboard a little.

Yes. I'd try increasing the minimum distance between the plate and motherboard.

One hour of prime95 blend and 8 hours of memtest 86 with no errors. Weird weird weird. Apparently it needed tighter sub timings. WTF I guess alls well that ends well. Thanks everybody. I'll try this for awhile and see if it holds.

Try Gold Memory. You really need to see how everything does when run with something different, and sometimes results have been very different between GM and MT86.

 

[Brando]

Yes. I'd try increasing the minimum distance between the plate and motherboard.


Try Gold Memory. You really need to see how everything does when run with something different, and sometimes results have been very different between GM and MT86.

You were right. Gold memory failed on the first pass. I'm ordering the mushkin to see if my board likes it more.

EDIT: It's weird. Goldmemory runs without errors through test 289 and then as soon as it gets to test 290 the errors start adding up rapid fire by the thousands. The errors kept piling up at light speed into test 291 when I turned it off. Are the different tests checking individual chips on the stick or what?

 

[larrymoencurly]

Gold memory failed on the first pass. I'm ordering the mushkin to see if my board likes it more.

EDIT: It's weird. Goldmemory runs without errors through test 289 and then as soon as it gets to test 290 the errors start adding up rapid fire by the thousands. The errors kept piling up at light speed into test 291 when I turned it off. Are the different tests checking individual chips on the stick or what?

Each of its tests checks all the chips, but I've also seen certain tests show tons of errors while others never found anything with certain memory modules. I kind of remember that happening with MemTest86 ver. 3.xx and its first 3-4 tests, so to save time I just skip to the other tests, and any modules that pass them then get screened again by running all the tests on them.

You may want to check these reviews of memory diagnostics: Part 1 Part 2

The articles are over a decade old, but apparently they remain the newest such reviews available. Notice how badly SIMMtester.com's DocMem performed; it's kind of famous for passing memory that other diagnostics say is really bad. And despite how well Gold Memory performed in that review, I don't trust any single diagnostic because I've had modules that passed one of them for hours and hours but failed another fairly quickly.

I wouldn't substitute G.Skill with Mushkin, unless the Mushkin modules had name brand chips on them and no heatsinks. I haven't bought Mushkin in a long time, but apparently it's not better than most retail brands, meaning it's not very trustworthy. I think the only way to always get top quality chips is by buying Samsung or no-heatsink Crucial modules.

 

[Brando]

Each of its tests checks all the chips, but I've also seen certain tests show tons of errors while others never found anything with certain memory modules. I kind of remember that happening with MemTest86 ver. 3.xx and its first 3-4 tests, so to save time I just skip to the other tests, and any modules that pass them then get screened again by running all the tests on them.

You may want to check these reviews of memory diagnostics: Part 1 Part 2

The articles are over a decade old, but apparently they remain the newest such reviews available. Notice how badly SIMMtester.com's DocMem performed; it's kind of famous for passing memory that other diagnostics say is really bad. And despite how well Gold Memory performed in that review, I don't trust any single diagnostic because I've had modules that passed one of them for hours and hours but failed another fairly quickly.

I wouldn't substitute G.Skill with Mushkin, unless the Mushkin modules had name brand chips on them and no heatsinks. I haven't bought Mushkin in a long time, but apparently it's not better than most retail brands, meaning it's not very trustworthy. I think the only way to always get top quality chips is by buying Samsung or no-heatsink Crucial modules.

Ah well it's too late now, the mushkin is on the way. I don't care if it's the best as long as it works at it's rated speed with stability. I just want to try another brand and it's the only thing on newegg that runs at 1.5v at a desirable speed that doesn't have massive heatspreaders that block heatsinks. I do kind of like the looks of that low profile samsung ram that everyone loves and may try it if the mushkin doesn't work out. I prefer the simplicity of xmp (if it would actually work). Prolimatech is sending me a universal backplate like the ones that come with the newer megahalems and they're sending it from taiwan for free so that's pretty cool of them. I'll be running the mushkin with a stock 1155 heat sink in the mean time. I'll check out that link you posted in a minute. Btw my ram passed goldmemory at 1600mhz, even test 290. I feel like if turning down the ram makes me stable it's a pretty safe bet that the cpu and the ram slots are ok. Unless I got a 3570k with a gimpy memory controller, but what are the odds of that? Then again I guess I'd believe anything at this point.

 

[Brando]

1.1 vccsa didn't work and neither did 1.56v ram voltage. Apparently my board doesn't have separate vccio, so vccsa is what I'm supposed to use according to the manual. I think I've officially tried everything with this ram now unless I'm forgetting something.

 

[Brando]

The mushkin 2133 ram booted right up with xmp profile # 1 without changing any other settings and made it through a pass of goldmemory on the first try. I'm running hci memtest right now to test it all at the same time and get a second perspective. It looks like my board just doesn't get along with the gskill because I can't see getting 2 bad sets in a row and never hearing a bad thing about that model anywhere.

EDIT: Got through half an hour of hci memory test and an hour of prime 95 blend with maximum memory. Seems legit. Gonna do a few more passes of goldmemory to make sure.
EDIT2:Another hour of hci memory and a pass of memtest86 good to go.
EDIT3:Made it through 2 more passes of goldmemory. I think it's safe to say I'm good to go. Awesome. Wtf Gskill(asus?).

 

[Brando]

Newegg is being nicer than I expected I guess. They agreed to give me store credit for the gskill a couple months after I originally bought it after being warrantied once and after me admitting that I probably didn't need to replacement motherboard they sent me. I'm not about to complain. Also, I forgot to say thank you to everyone that helped, especially larrymoencurly for pointing me towards goldmemory. Without that post I might never know it was ram and kept changing out other stuff. Much appreciated. I'll spread the good word.

 

[Brando]

Welp, it's been a month with the new mushkin memory and the same questionable prolimatech back plate. The mushkin is still working fine.

DAMN YOU GSKIIILLLLLLLLLLLLLLLLL!!!!!!!!!!!!