Crucial Anniversary: The results

Associate
Joined
3 May 2006
Posts
614
Location
Leeds
All you crucial anniversary owners post your results here!

For ref, these are rated to 533, 3,3,3,8 timings and use Micron D9 GMH chips.
 
E6300_Prime_sized.JPG


450FSB, 4,4,3,8 memtest and orthos stable (don't have a screeny of memtest, can get it if anyone wants) 2.04v.
Mobo limited. To be tested in Asus P5B :)
 
Got these from elsewhere just before OCUK brought the lot!

Voltage wise, 2.5 is probably the max safe up to 2.6 for short periods shouldn't be a problem, just keep the chips cool. 24/7 up to ~ 2.3/2.4 *should* be ok but again make sure the modules are well cooled.
 
Ram (and CPU) moved over to P5B for further testing. Will keep this post updated with results as and when :)

Memtesting (windows) at 480 4,4,3,8 atm. 2.2 but can hopefully drop that, just CBA finding a screwdriver for the VR
 
The 965 boards are suited to lower multi chips as they'll hit much higher FSB.
Above 400 FSB they're a little slower than the 975's due to looser chipset timings (aka 1333 strap). E6600's seem on average to run ~ 3.6ghz 24/7 stable which is 400 FSB assuming you use the stock multi. It may have made sense to either plump for a lower mutli chip and use the high FSB the P5B's capable of, or go for a 975 board which will max FSB wise around where your 6600 will.
Memory wise, all it means is you'll be forced into a divider to get the max out of your memory.
 
495_ram.JPG


Voltage there was ~ 2.18 or so. Well within safe 24/7 limits

Also have a Pi 1m shot @ 534 FSB. Voltage was around 2.2, haven't done any stability testing here as CPU's on the edge temps wise. Dropping mutli pushes the NBCC clock over 600 mhz and something craps out. Better cooling is en-route, so should have a definitive answer soon.

100_OC.JPG
 
Jokester said:
457-4-4-3-7-1T2.JPG


That's as far as the 1T goodness will go :( .

Bit of tweaking the other timings might see that over the 8GB/s mark but still a long way short of AM2 onchip memory controller goodness.

Jokester

What voltage mate?
 
Nelly said:
Kinda dumb really considering they bought all the stock at a set price, its just greed with no consideration to the customer. The worse thing is me speaking hypocritical it wont make any difference because people like me will still buy it anyway. lol :rolleyes:

No, its good business sense... Why make 190 quid per kit when there are people out there who will pay more?
 
fornowagain said:
Well I've got 2GB of G.Skill HZ, paid £170 for that, now look at it. Wish I'd bought a few extra sticks of that, I can tell ya. It never changes, I remember meeting some dodgy, cloak and dagger bloke for cheap 8Mb Simms back in the early 90's, stuff was so expensive you wouldn't believe.

Did the same, should have brought more! Wish I'd done the same with this crucial too, 186 quid inc was a bargain at the time!

Nice speed Jokester, still running 2.2v or did you need more juice?
 
Can't match Jokesters clocks on this mobo, not 1:1 at anyrate, dividers are next but seem to be a bit crappy on this board.
550 4,5,4,6 2.3v seems memtest stable but thats right on the edge of what the board will do. Need to drop voltage as there should be no probs running that speed with lower voltage.
 
PinkFloyd said:
Just reading through this thread has me very tempted......

I was looking at the ocuk value stuff to save money, but this looks like it could be very nice for me indeed.

Would be going in with a e6600, in a Asus P5B.

Is it worth the extra cost!

No :)
 
nightic said:
This all assumes you want to run 1:1, with a relatively low (400MHz or less) FSB due to the 9x multi on the E6600.

IIRC 'Hipro' mentioned in the XS forums that with 965 boards (specifically the P5B-Deluxe) there are much greater benefits in running a divider, pushing the RAM as high as it will go.
This is likely also the case (although perhaps to a lesser extent) with most other popular chipsets for Conroe CPUs.

This is true, Core2 seems to gain most from higher FSB on the whole. 680i gains more from lower FSB/1T running however this isn't possible on the 965/975 chipsets.
Some of the 965/975 boards do seem to have problems with certain memory dividers(or multipliers) which can make memory clocking a headache. The performance boost from running the memory at a higher speed isn't on the most part noticable (unless your checking FPS or pi times) and probably doesn't warrant the extra expenditure in this case.
Had the chip been a 6300 (or maybe 6400) then yes, the crucial maybe a better option, even then though for 1:1 running most boards (bar some of the 965's) are maxing ~450 anyway which is do-able for most 6400 with a timings drop/voltage boost.
If you've got a 680i board and pretty much any CPU, the unlinked memory clocking is highly useful, and the Crucial would certainly be a good choice.

PinkFloyd said:
While were on the discussion of high memory speeds being better, am I also correct in thinking that higher fsb = more performance because of bandwidth like with old AMD XP systems?
Had AMD64s for too long!

Yea, on the most part higher FSB = better performance than lower with tight timings. Exception to this is 1T running on 680i boards. 1T + low latency = better performance than higher FSB and looser timings.
 
Big.Wayne said:
I changed the timings from 4-4-4-12 (1.8vDimm) to 3-3-3-8 (also 1.8vDimm) still 400MHz (DDR800).

1.80v = No Boot
1.85v = No Boot
1.90v = No Boot
1.95v = No Boot
2.00v = No Boot
2.05v = No Boot
2.10v = Boot but freeze at BIOS splash screen
2.15v = Boot but errors in MemTest86 Fast!
2.25v = Boot + Run Memtest86 Overnight 100% Pass! :D

<image here>

Can you run windows memtest at that speed/voltage?
 
Marshall said:
Im having major issues with my 10th Anniversary, i've RMA'd it and the mobo (DS4), and the mobo was found faulty and replaced. but i was told the ram was fine. when i got it home even at a setting of 5-5-5-15 with 2.2v i cant get the ram to run at 333 (667) without it bringing up lots of errors in memtest. Does the ram sound dodgy even though it passed the RMA tests, is it possible my E6300 is damaged and thats whats causing the problems? :confused: :confused:

Test the ram in another board if you can. Where abouts in the UK are you mate, willing to test the ram/board/CPU out for you if you're close (Leeds) :)
 
Back
Top Bottom