Jump to content

OCZ 4000VX on NF2 Lan Party B


Guest Darien_merged

Recommended Posts

  • Replies 211
  • Created
  • Last Reply

Top Posters In This Topic

Guest Darien

OK guys,

 

this is kinda weird...

 

you've all heard how I can't prime for more than a minute with my VX... and that regardless of what RAM I use, I get serious bigtime SATA corruption over 250FSB...

 

well I just got my second DFI LP ultra B.... and guess what...

 

My VX RAM still fails memtest... still 191.8MB is the one that always fails.. still fails S&M at 250 as well... but I've been priming for about 2 hours at 250 x 10 2-2-2-2-11 1t 3-3-4-3-2-2-3 e-e-f-e auto (4-11) strength and slew settings...

and it's still going strong on my new board!

 

and

 

I can run upto about 258 FSB before I get SATA corruption

 

AND

 

I can do 257 x 10.5 where before I couldn't do 250 x 10.5, even though my chip can bench at over 2720 MHz...

 

STOKED!!!!! :D

 

better news is I talked with OCZ and they are primei95'n and running memtest86 on another pair of ocz 4000 vx for me... hand tested to make sure they are upto spec...

 

so I fully expect these pair of sticks to fail prime blend in under 8 hours because they don't pass memtest at 250, they are doing damn well!!!

 

so it looks like my:

1, SATA corruption

2. prime blend failure

3. inability to run any multi above 10 when @ 250 or higher

were all do to my motherboard!?!?!!!

 

the sticks still fail memtest and S&M, but if they can pass 8+ hours of prime blend and 3dmark testing, I may not have to send them back.... ofcourse that's a HUGE if...

 

i still expect the sticks to fail in prime under 8 hours... and even if they pass I'll have to make sure they can hold up to many other stress / benchies and most of all real world testing.

 

EDIT:

prime95 blend failed in 3 hours and 31 minutes... phew! now I feel justified in sending the VX Ram back... but confident that my mainboard was a large part of the problem.... from 0-1 minutes with the old LP B to 3 hours and 31 minutes with the new one is a HUGE improvement.... once I get my new 4000VX sticks I may have the dream!

Share this post


Link to post
Share on other sites

Sounds good Darien. On your 3hr 31min prime run, can you list all the relevent bios settings? At the 191.8mb failure in memtest, what test was it(5,6,7,8)? I ask that because there is a way to get round it if i know which test it is.

Share this post


Link to post
Share on other sites

Guest Darien
but I've been priming for about 2 hours at 250 x 10 2-2-2-2-11 1t 3-3-4-3-2-2-3 e-e-f-e auto (4-11) strength and slew settings...

 

I used 1.775 vcore, 3.37-9 vdimm, 1.9vdd... 191.8MB fails tests 2,3,4,5,6 and probably 7 and 8 if I checked... I switched the sticks( put slot 1 in 3, and slot 3 in 1) and it's the same 191.8MB alwasys fails

 

the thing is... 3dmark01 always fails, usually BSOD... weird thing though it started to pass S&M memory test... but every time I do the S&M cpu test it always fails and the computer reboots??? even at setting I know are stable?

 

it did pass 3dmork03....

 

I tried upto 3.57vdimm and 2-3-3-3-11 and auto alpha's and slews for the RAM but it doesn't matter always fails 191.8MB... Lots of error on 2,3 and 4... but always on 1... just one per pass and always the same error during test #5

Share this post


Link to post
Share on other sites

Guest Darien

Just got my new VX and was having some fun benching them...

 

keep in mind this is 1t.... try not to drool!

 

click the image below

 

attachment.php?attachmentid=3968&stc=1

 

:D

Share this post


Link to post
Share on other sites

Guest Darien
Looks good Darien :D Just a little warning though, the max image size is 800x800 so it would be a good idea to resize it.

Done any Primeing Yet?

whoops!.. I accidently switched the url's... corrected...

 

I'm priming right now at 257 x 10.... these bad boys were memtesting no errors 250 @ 3.24v.... I kicked it up to 255 and it got 40 errors a pass on test 5.... I bumped up to 3.36v... It passed NO ERRORS... over 30 minutes @ 255 at 3.36v no errors...

 

I kicked it up to 260 and 3.36v 2 error each pass on test 5, I bumped up to 3.47v and it didn't matter... probably a chipset/bios limitation for 1t...

 

it was doing ok in memtest @ 258... didn't test it thoroughly though... may do 259...

 

I was really wanting to get the 10.5 multi.... but I think the vcore flux is the issue.. cuz 255, 257, and 258 @ 10.5multi works better with 1.85 vcore than it does with anything up to 1.950 vcore... higher vcore = less stable... I think it's because the higher the vcore the higher the flux... temps may play a role there too...

 

I think I'm gonna hold off on the 10.5 multi for now... I'll lap my TR copper HSF really good and put that on... once I do that it will bring down the temps... and I'll put the tornado back on too, then I'll try the 10.5....

 

the good thing is the 10.5 x 255-258 doesn't throw errors, but it will freeze or BSOD... so may just be a vcore or temp issue...

 

I'd be thrilled with 257 x 10.... but if I could get 257 x 10.5 I'd die happy cuz that's xtreme as all hell... 257 x 10.5, 2-2-2-2-11, 1T = 2.7GHz... that's more of a pipe dream...

 

In reality, I'll probably find myself somewhere around 255 x 10... or maybe 255 x 10.5 if I get really lucky... but I'd be really happy with those too.

Share this post


Link to post
Share on other sites

Yes, realistickly the most i get out of mine is 255. Prime wise though in my case at 255 12hrs or so is as far as it goes. If your memtesting at 258 then it looks like you might have a better pair than mine, so 255 looks like it's on the cards :D

Share this post


Link to post
Share on other sites

Guest Darien
Yes, realistickly the most i get out of mine is 255. Prime wise though in my case at 255 12hrs or so is as far as it goes. If your memtesting at 258 then it looks like you might have a better pair than mine, so 255 looks like it's on the cards :D

well... now I don't know kit...

 

@ 257 x 10 prime blend failed in 1h 17min... not too bad for a start anyways... but now the things is crapping out on me...

 

failing 256 in one minute... failing 255 by bsod... now 256 and 255... were starting to freeze in memtest... where as before memtest was solid and prime did much better...

 

also S&M memory test was failing or freezing within 2-5 minutes... at 255!!

 

I found something strange though... I think I found out why... for some odd reason... 1.775vcore is plenty to do 260-265 with my plat rev 2 @ cas2.5 2T...

 

but doing 255 x 10, cas 2 1t with my VX... I need more vcore... once I bumped up the vcore to 1.825 in the bios... memtest didn't freeze, went over 30 minutes no error or freeze.... and now S&M long test has been running for over 2 hours @ 255 x 10 without error... it's like over %80 done...

 

so i dunno... just looks like I need (or now need) a higher vcore when using my VX ram for the same or even lower MHz, as compared to using the tcc5 @ 2t...

 

thoughts?

 

(btw... if you ran 255 prime for 12 hours... why didn't you update your toaster?)

Share this post


Link to post
Share on other sites

(btw... if you ran 255 prime for 12 hours... why didn't you update your toaster?)

 

Because it would bomb out of 3d03 within a min or so and past experience told me that at least one stick of ram was the cause. I had no intention of killing another set of VX when it was fine at 250. Don't forget "toaster's"should be a sign of a good stable rig that can be used 24/7, it's not just a willy waving exersise.

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now

×
×
  • Create New...