Crucial Ballistix DDR3-1600 / PC3-12800

[saaya]

isnt that pretty much just saying the same thing 3 different ways?

i said i dont like the heatspreaders and then elaborated why.
whats your point?

Ahh he's just spoiled for choice because his company gets to pimp theirs in big fancy aluminum ones

:P
well to be quite honest our heatspreaders are a bit too simple imo... theres not much profile, i hope i can get a more fancy heatspreader onto our sticks
like the dominator ones but they are too thin for my taste...

[*]Interesting behaviour of probably the board or memory is that Crucial don't like more than 2.1v vdimm if you manually set strap to 266FSB with 1T enabled - it gives more memtest86+ v1.70 errors as you push vdimm >2.1v. But set to 333FSB strap the memory can respond to >2.1v better.

did you try different slots?
this sounds exactly like what im getting on the orange slots of the p5k3, i didnt notice a corelation to the strap tho... very weird...

interesting results! thanks goerge!

Maybe your Blitz Extreme it is not so lucky it is one of the possibilities.

We tried with air and water cooling:
E6420 around 550 FSB in both motherboard (Blitz Extreme and P5K3)
E6850 around 540 FSB (no tests with P5K3)
E6600 560-570 FSB on DFI 965P but only 495 on Blitz Extreme.

So at the end in our tests Blitz Extreme performs well only with some CPU types. For our motherboard is not a trouble of FSB wall related to motherboard, because Blitz Extreme in our hands was able to reach 550 of FSB.

For these reasons I am thinking about some trouble with bios CPU support. I think you have to try with E6420 or E6850.

ciao vincenzo!
now thats very interesting!
thanks for sharing
did you try different straps too?

[v_parrello]

......
ciao vincenzo!
now thats very interesting!
thanks for sharing
did you try different straps too?

Hi Sascha,
if I remember well we did with 266 and 333 NB strap.

[eva2000]

Maybe your Blitz Extreme it is not so lucky it is one of the possibilities.

We tried with air and water cooling:
E6420 around 550 FSB in both motherboard (Blitz Extreme and P5K3)
E6850 around 540 FSB (no tests with P5K3)
E6600 560-570 FSB on DFI 965P but only 495 on Blitz Extreme.

So at the end in our tests Blitz Extreme performs well only with some CPU types. For our motherboard is not a trouble of FSB wall related to motherboard, because Blitz Extreme in our hands was able to reach 550 of FSB.

For these reasons I am thinking about some trouble with bios CPU support. I think you have to try with E6420 or E6850.

Yeah i think i just got a bad Blitz Extreme as every cpu i tried goes higher FSB on other boards heh.

did you try different slots?
this sounds exactly like what im getting on the orange slots of the p5k3, i didnt notice a corelation to the strap tho... very weird...

interesting results! thanks goerge!

for Blitz Extreme tried both blue and white slots.. but on P5K3 DX only started with black slots never used orange.

[PanteraGSTK]

Yeah i think i just got a bad Blitz Extreme as every cpu i tried goes higher FSB on other boards heh.

for Blitz Extreme tried both blue and white slots.. but on P5K3 DX only started with black slots never used orange.

I think the low fsb of the blitz extreme is why people are doing the cap mod. This is from Victor. http://67.90.82.13/forums/showthread.php?t=158732

This one is from George (hipro) http://67.90.82.13/forums/showthread.php?t=159973

Hope this helps.

[guess2098]

hey man. good review as always
a lots of info here hehee


Denny

[saaya]

Yeah i think i just got a bad Blitz Extreme as every cpu i tried goes higher FSB on other boards heh.

for Blitz Extreme tried both blue and white slots.. but on P5K3 DX only started with black slots never used orange.

was there a difference between the blue and white slots on yor extreme?

[eva2000]

I think the low fsb of the blitz extreme is why people are doing the cap mod. This is from Victor. http://67.90.82.13/forums/showthread.php?t=158732

This one is from George (hipro) http://67.90.82.13/forums/showthread.php?t=159973

Hope this helps.

Thanks for the info.. unfortunately no volt mod/soldering skills LOL

hey man. good review as always
a lots of info here hehee


Denny

Glad you like Denny.. doing more testing this week on P5K3 Deluxe as no X38 in sight for me at the moment heh

was there a difference between the blue and white slots on yor extreme?

definitely was a difference

P5K series = black dimm optimised
Blitz series = white dimm optimised

[saaya]

interesting... thanks for the infos george

btw, your images load incredibly fast, very nice host

[eva2000]

Thanks i think my web host did a router/network upgrade recently and helps to optimise images with 75% reduction in file size http://i4memory.com/showpost.php?p=61904&postcount=3 and i did some tweaks on the web server /apache itself

[3oh6]

Thanks for the info.. unfortunately no volt mod/soldering skills LOL

it is about time we did something about that George. i have moved my "Soldering for Eva" article up a few notches on the priority list. i really just need a video camera to make it complete...taking photos and soldering at the same time is still something i have to work on as still frames just don't explain what i am showing that well.

i now solder for a living and my soldering skills have gone well past 'iron-ninja' in the last few months so hopefully i can at least get you doing the standard volt mods

now if only i could bring my camera to work to show some extreme macro shots through the Mantis...nothing like soldering at 10X under a stereo microscope

[saaya]

cool, a video guide on vmodding would be great!
we have people with 20+ years of soldering experience here at cellshock, but id still like to mod the boards myself hehe

[eva2000]

Thanks a volt mod guide would be nice.. but i pride myself in having never killed any hardware myself from oc'ing (only DOA stuff) in the past 7 yrs since i started.. probably cause i never did volt mods LOL

But looks like bad luck comes in pairs first my BenQ FP91G monitor and just now my P5K3 Deluxe/E6600 system's connected Belkin Surgemaster power board went up in smoke - literally big pop sound and smoke came out and the entire house blacked out tripping several safety switches

Smell of burnt plastic from the Belkin Surgemaster power board but looks like it did it's job and protected my connected P5K3 DX/E6600 system from the power surge as I'm typing from this P5K3 DX/E6600 rig right this moment.

Have yet to boot up my usual main pc yet as i'm diagnosing one pc at a time hoping nothing is damaged.

Wish me luck!



Just before the smoked Belkin Surgemaster power board happened, I did pull off my fastest <3600Mhz 32M time to date (faster by ~2 seconds)!

Crucial Ballistix PC3-12800 @900Mhz 7-6-5-18 1T at 2.0v vdimm bios set.





mixed emotions right now LOL

[3oh6]

But looks like bad luck comes in pairs first my BenQ FP91G monitor and just now my P5K3 Deluxe/E6600 system's connected Belkin Surgemaster power board went up in smoke - literally big pop sound and smoke came out and the entire house blacked out tripping several safety switches

wow, that is crazy but be careful...bad things come in threes is how i understand the superstition, you've got one more to go

as for the whole dead hardware...perhaps your right but i don't think it is something to worry about. i haven't killed any hardware where it wasn't a direct result of my own stoopidity. volt mods haven't killed anything for me yet...*knocking on wood*

hehe, its funny you were working on 3600MHz time with the P5K3/Ballistix, so was i last night. its all about the 514*7 though

click for full size including Memset...


i might have to try your run at 450*8 though with that higher mem speed...although at 514FSB the bandwidth seems to be better despite a PL of 7. my personal best >3600MHz BTW too last night.

[eva2000]

Seems like it was a power surge and the surge protector and maybe mains safety switch saved the day.. both bench and main pcs seem fine to me! I consider myself dam lucky this time!

Thank goodness for Belkin 8-Way Surge protector board which has lifetime warranty http://catalog.belkin.com/IWCatProdu...duct_Id=173139 - 2700 Joules of protection - so that surge must of been a whopper to smoke the surge board!

But hope that's the last of the bad luck hehe

Wow 12m 53s... congrats on your PB too!

Looking over the bandwidth and memset timings and memory frequency - probably a bit more to do with tweaking/maxmem etc than straight bandwidth/FSB/Mem frequency there but very nice!

I see over-charge = 0.9v - what does that setting do ? One thing i miss on Blitz series is being able to read CPU PLL voltages! Does P5K3 Deluxe CPU PLL on AUTO, automatically raise CPU PLL volts ??

[3oh6]

Wow 12m 53s... congrats on your PB too!

Looking over the bandwidth and memset timings and memory frequency - probably a bit more to do with tweaking/maxmem etc than straight bandwidth/FSB/Mem frequency there but very nice!

I see over-charge = 0.9v - what does that setting do ? One thing i miss on Blitz series is being able to read CPU PLL voltages! Does P5K3 Deluxe CPU PLL on AUTO, automatically raise CPU PLL volts ??

thanks...yeah i think your right, a whole combination of things are resulting in that bandwidth.

when running at 514*7 PL7 i thought i needed the over-charge cranked up for the best times. i was assuming something to do with a NB voltage but i am just doing some runs and seeing if i can lower it. we'll see what happens to the times.

as for PLL, i haven't tested to see if it auto raises but based on the history of Asus boards and AUTO settings, i would imagine it does. you gotta pull out a DMM, its all i trust these days for actual voltage readings.

[eva2000]

Do you know the read pts for DMM for P5K3 Deluxe ??

[saaya]

glad everything is alright eva!
have you tried low fsb/memspeed with tight timings and tweaked pl compared to high fsb/memspeed and tweaked pl yet? I did that some time ago and low fsb low memspeed was very close to high fsb/memspeed. pl is really a weird thing...
i really wonder what exactly it does

[3oh6]

Do you know the read pts for DMM for P5K3 Deluxe ??

i have been using the read points from the P5K-Dlx thread:

http://www.xtremesystems.org/forums/...d.php?t=145142

they seem to be scaling correctly with the voltage being set in the BIOS so i have just gone off them. for vDIMM, i have been using the "test pad" in-between the two sets of DIMM slots. the white half gives the same reading as the capacitors in the area.

same with the "test pad" near the vMCH read point in the thread. the white half gives the same reading as the capacitor shown in the P5K-Dlx thread for vMCH.

[eva2000]

Thanks for the link

saaya, haven't got that in depth so far as Blitz Extreme low FSB did limit me to <485FSB... so just starting out with P5K3 Deluxe and 530FSB headroom.

So many ways to tweak on Asus P35 DDR3 you have static read and write delay, transaction booster levels, 1T/2T and 266FSB/333FSB strap so there's so many ways to reach your desired speeds

[eva2000]

System:

• Intel Core 2 Duo E6600 L709A958
• Corsair Nautilus 500 H20
• Asus P5K3 Deluxe 0604 official bios
• 128MB Gainward FX5200 PCI
• 2x1GB Crucial Ballistix PC3-12800 DDR3-1600 8-8-8-20 at 2.8v rated dual channel kit
• 74GB WD Raptor 8MB
• LiteOn 165H DVD Burner
• Sony FDD
• 1KW PCP&#37;C SLI PSU
• WinXP Pro SP2 Luna Royale Blue themed with nLite + RyanVM updates

7x514FSB 3:5 856Mhz 7-6-5-15 1T at 2.00v bios set
I finally broke the sub 13min Super Pi 32M barrier at <3600Mhz and when I mean broke, I mean I slashed 8.25 seconds of my previous personal time of 13min 02.094s!

I needed 2.0v vdimm as 1.95v vdimm resulted in memtest86+ v1.70 errors (4 errors at test #5 looping).

How's 12m 53.844s for 32M with WinXP Pro SP2 fully updated with no maxmem tweaks at all!





Now I wonder if I can go sub 12m 50s for 32M at <3600Mhz ?

[3oh6]

very nice, congrats! i must have something screwy with my 32M OS installation because it is tweaked to the nines (or so i thought) and everyone is kicking my butt with this setup at 514*7 / 3:5 / 7-6-5. i might have to try with a bit higher vDIMM, perhaps it will speed it up a little.

your using that n'lightened XP install you posted at over a i4memory?

also, your vDIMM is .15v higher than mine for the same frequency timings. you said you got Memtest 86+ errors, i am assuming then that 32M wouldn't run with lower vDIMM? i see your using the black slots already...have you tried swapping the order of the sticks and running at lower voltage? if so, then i guess i finally got lucky and have a better kit of something than someone, at 7-6-5 anyway. usually i end up with the dud DIMMs

i just ran a couple hours of Prime Blend Pr9 at the same settings i pinched my 12:53 out at. Prime blend isn't the best memory test, HCI is next, but its still usually a little bit better a test than a single 32M.

[eva2000]

Yup nLite custom WinXP Pro SP2 cd i posted in OS forum at i4memory.com

I don't even boot into windows let alone run super pi 32M, if i get memtest86+ v1.70 errors, so haven't tried lower vdimm for 32M hehe

Haven't tried swapping dimm slots yet.. but will try other stuff later

For now set RAS to CAS Write delay in memset from 6 to 3 and squeezed out 0.906 seconds faster 32M time

[3oh6]

For now set RAS to CAS Write delay in memset from 6 to 3 and squeezed out 0.906 seconds faster 32M time

wow, nice find! i would have never thought a whole second could have been had from that change. i always thought it was a superficial one with DDR2. thanks for the heads up...i might have missed it in the screenshot if you didn't say something

[eva2000]

Yeah same for DDR2 that tweak in memset ranges from 0.2 to 0.9 seconds improvement for me.

[NoDsl]

....

Typical Crucial conservative memory rating scheme
....

I really like this thing crucial does. No fancy packages, just performance.

BTW, nice review