Ballistix PC8500

[zsamz_]

hey turtle look what da balistix help me do
i think i maxed out on my cpu fsb my damn board doesnt have too many options damn 1.60 in bios next step is 1.80 wich is 1.78 real too much for me on air

[Zeus]

At those volts 1350 5-5-5-15 bench stable is pretty good RAM AFAIK. I haven't tested more, but if someone bought them off me they could trash them all they like. I'm not bothered of what speed or timings they reach, I only run 1000-1066 5-5-5-15 at 2.1V given the choice. Thats not my experience or findings. Corsair RAM on the same everything is quicker in 1M/32M than the Crucial I tested. I've already tested this in the CDT thread, it's all there and someone else posted with similar test findings before me.

Oh, if you're only running that low we should swap out our memory's.

Are you planning to sell them?

I did not say Pefromance Level doesn't make any difference in performance, it does indeed but it ain't a memorysetting, it's a chipset latency or something.

Doing 7x514 4:5 PL 6 i have to raise Vmch, for PL 7 i can leave it at auto.
That's a sign it is chipset related, not memory.

[turtletrax]

hey turtle look what da balistix help me do
i think i maxed out on my cpu fsb my damn board doesnt have too many options damn 1.60 in bios next step is 1.80 wich is 1.78 real too much for me on air

Sweet!!! I wish quads weren't so hard to FSB clock Still learning this DFI board, but it is coming along... slowly

[KTE]

Oh, if you're only running that low we should swap out our memory's.

Are you planning to sell them?

Yep.

I did not say Pefromance Level doesn't make any difference in performance, it does indeed but it ain't a memorysetting, it's a chipset latency or something.

Doing 7x514 4:5 PL 6 i have to raise Vmch, for PL 7 i can leave it at auto.
That's a sign it is chipset related, not memory.

I understand what you initially said meaning performance is no different from RAM to RAM. But I found different results and I did this inside a matter of 1 day with the same config and no other variables, so to make sure. I'm aware of the NB effect as well, that's why the P35 DS4 would run slower than a P5K in Pi at the same settings.

[nullface]

I know this has been asked a million times, but what voltage is safe for every day use, 2.2v? Or can i give them some more?

[zsamz_]

I know this has been asked a million times, but what voltage is safe for every day use, 2.2v? Or can i give them some more?

\
the lower the better lol but hey its up to you
but when i get my mushkins back from rma they gonna see 2.3v
i think 2.2-2.3 is ok but what do i know lol i lost 6 stix to 2.26v
i think its my damn board

[maxwedge]

May be you want to check your board with a multimeter?

[zsamz_]

May be you want to check your board with a multimeter?

well i did that long ago loli know my exact volts

[TinTin]

More results of my Ballistix PC8500 1G X 2



2.05V @ DDR1200 CL5-5-5-12, memtest 140% passed - Original Heatsink



2.55V @ DDR1300 CL5-5-5-9, ortho 30mins blend test passed - Original Heatsink



After Changing Thermatake RAM Heatsink,
2.55V @ DDR1200 CL4-4-4-5-4-25, memtest 100% passed



2.55V @ DDR1398, CL5-5-5-15, direct boot to window, superpi1M passed


2.55V @ DDR1401, CL5-5-5-15, direct boot to window, superpi1M passed


Love these rams

[Zeus]

So, you changed the heatspreaders and it gave you almost 100MHz gain?

What heatspreaders do you have now, do you have a picture? How easy was it to remove?

Very nice DDR1400 man!!

[TinTin]

So, you changed the heatspreaders and it gave you almost 100MHz gain?

What heatspreaders do you have now, do you have a picture?

Very nice DDR1400 man!!

I used that heatsink. I just used my fingers to remove the original heatsink, tear it form one end and then the other end, then removed it very very slow, the original heatpad can be kept.

Before I changed the heatsink, DDR1200 CL4-4-4-5 can boot to window but can't pass memtest or even finish superpi32M, giving me error very early!

and for CL5, the highest clock before changed heatsink is about DDR1350, 1400 is no way to enter window.

Recent D9GMH is very hot but very overclockable compared with my old Team DDR667 and DDR1000, so the heatsink may help to clock my crucial.

[KTE]

Those volts scare the daylight out of me, it's pointless for me. After killing some at 2.2V I would never try those again on something I want to keep whether for 2GHz or 1.2GHz. But anyone doing it would obviously not be bothered enough to kill 'em.

There's a bug on Phenom with some platforms and it gives you DDR2-1400/1500/1600/1700//1800 validations and bandwidth.

[TinTin]

I just received a pair of Tracer PC8500 1G X 2 yesterday, some initial tests with My DFI P35-T2R.




DDR800 CL4-4-4-5 @ 1.64V memtest 100% passed



DDR1200 CL5-5-5-9 @ 2.04V memtest 100% passed


DDR1200 CL4-4-4-5 @ 2.51V, Performance Level = 4


DDR1300 CL5-5-5-12 @ 2.39V



From the voltage with DFI, it seems not bad, I will test it more when I get my P5K-Premium back.

[Vatos_locos]

Why everybody hides the Serial number ?

[Don_Dan]

I just received a pair of Tracer PC8500 1G X 2 yesterday, some initial tests with My DFI P35-T2R.

I will receive some Ballistix Tracer soon too, so these results make me very happy, but why do you hide the serial number of your kit?

[Zeus]

He's hiding the serial numbers for the mfg, if they would see the voltages he's throwing at them warrenty is lost.

TinTin, how did you get PL4 to work?

[Don_Dan]

He's hiding the serial numbers for the mfg, if they would see the voltages he's throwing at them warrenty is lost.

Yeah, that makes sense... Thanks!

[jas420221]

I have some of these sticks on the way... PC8500 tracers. Im looking to get an e6750 to the 3.8Ghz (+/- .2) range on air (Tuniq 120) on a X38 maximus formula...

Advice? These are only cooled by the case airflow (Antec 900 so good flow).

Can I 1:1 these and get that goal? or us 2:3 more likely?

Im an AMD overclocker so this is alllllllllllllllll new to me.

[Vatos_locos]

He's hiding the serial numbers for the mfg, if they would see the voltages he's throwing at them warrenty is lost.

TinTin, how did you get PL4 to work?

ok it hasn't crossed my mind thanks

[flesheatinvirus]

I just got my sticks in a few days ago and had a chance to play around a bit this weekend with them. So far they look pretty good.

[TinTin]

He's hiding the serial numbers for the mfg, if they would see the voltages he's throwing at them warrenty is lost.

TinTin, how did you get PL4 to work?

It is my first time to see it works with PL=4, I just always leave it AUTO in BIOS and have to use 2:3 divider, very strange, just Tracer, then it boots to PL=4 and works

[Zeus]

It is my first time to see it works with PL=4, I just always leave it AUTO in BIOS and have to use 2:3 divider, very strange, just Tracer, then it boots to PL=4 and works

Will it still do PL4 if you lower tRFC?

So only the Tracers did this?

[KTE]

That's odd. I can never change PL in Windows-> higher or lower will freeze system, but PL5 was easy at 4-4-4-4 1200 stock volts from bootup on auto on these sticks.

TinTin: if you're sticks are anything like mine then 3-2-2-2-1 will be easy at 400-450 and changing tRFC at or below 38 won't be plus 1250 4-4-4-4 PL5 tWTR 10 tWTP 10. Increase any of these and you get more stability and can lower another one down.

You guys posting massive images: the page is empty and taking forever to load. 7 and some minutes now and still hardly any pics. Smaller would be far better.

[eva2000]

That's odd. I can never change PL in Windows-> higher or lower will freeze system, but PL5 was easy at 4-4-4-4 1200 stock volts from bootup on auto on these sticks.

TinTin: if you're sticks are anything like mine then 3-2-2-2-1 will be easy at 400-450 and changing tRFC at or below 38 won't be plus 1250 4-4-4-4 PL5 tWTR 10 tWTP 10. Increase any of these and you get more stability and can lower another one down.

You guys posting massive images: the page is empty and taking forever to load. 7 and some minutes now and still hardly any pics. Smaller would be far better.

Yeah guys for smaller screenies read http://i4memory.com/showpost.php?p=61904&postcount=3 easily save up to 75% in image file size while keeping quality and physical size of image intact

[jas420221]

Maximus formula has EPP profiles and such right? So I will be able to run these @ 1066 2.2v 5-5-5-xx???