Can i do better

[NEOAethyr]

 

 

Finally, finally got some real stability going for me.

It only took me 3 days to doit or somehing like that :.

 

 

Btw, I found that the +3.3v rail is actually like the rest int he bios, .1 off.

 

 

Anyways I made a new bios, 2 in one sorta thing so I can do 266 and 275.

Just got 275 stable, now I need to check 266, etc.

To many things talk about :.

 

 

Here's some guidelines I figuered out while my little adventure:

 

Disable any hd controller thingys if they are'nt in use.

Seems somewhat inportant, probably not as serious as my other discoveries.

If you don't disable them, you get instability.

 

Don't set a cd/dvd rom drive to use LBA drive translation/access mode!!!!!!!!!!!!!

Lol, it is inportant, severely.

CHS and Auto works fine.

The drive will act perfectly fine on lba, heck I've ben doing it for a very long time now.

I recently was told it can be bad, from a buddy.

Well I blew it off until yesterday, can mean th diffrence between memtestable instabe bigtime and s&m stable(but was'nt superpi stable).

Still a huge diffrence.

 

Floppy drives.

If you don't have one conencted, and it's setup to use one, it can mean trouble.

Again like the dvdrom thing, just not as severe.

It's more severe though then letting the hd controller go.

 

 

Measuring rails, I think I got it down, but I'm just not sure yet.

Measure +12v at the hd's.

Measure the +5v at the hd's???, I did it last time.

Measure +3.3v at the atx.

 

 

Atx vs's sata with no load, well there is a few 0.0xv diffrence.

 

ATX: 3.30 = 3.34-3.35 :sata (No load)

ATX: 3.33 = 3.36-3.37 :sata (No load)

 

The 3.33 may be slightly worng, as my connection on the atx was poor at the time.

It's sorta freaky jamming the dmm's pin thing in there ^^.

 

Anyways I setup the +12v to not go under 12, and that's it.

Same goes for 5.

3.33 I setup not to go under 3.33 idling in the bios, like I did the rest.

Got me 275 stable again, in dc mode.

what a pain.

 

The other problem I've had was that I was setting up drives and slews...

Well, what I know is best when in sc mode, is not best in dc mode :.

Somehow they combined, so the old 4-11 2-11 thing is what I'm using now at 275.

 

 

Oh yeah...

The dfi ide udm133 cables.

The ones that came stock with the mobo, the round green ones.

These are major trash.

I am defently sure.

They cause memtest errors, and make drives dissapear ramdomly.

A generic round unshielded cable my mom got for her system at bestbuy is actually working much better then that lame cable of dfi's.

 

 

I now know what can give me poblems.

I'd liekto take the time and say I was wrong about the ocz, and it's cables.

It's a good psu, then again some parts can use filtering, better cooling, and better wire management, etc etc.

Hoever it's a good psu.

The power shield cables, I retract my statement before, these are just fine.

It was just the small things I was having problems with.

If I dont' have my drive's balanced to the point where every single one of them is doing great, I'll have huge problems in stability.

 

Drives that are not liking the env, freak the system out.

This is probably what the deal is with sata, just a guess.

The are probably more picky, this may explain why some can oc ok on the nf2 with sata, and why some can't.

 

 

So many wierd problems I've ran into the last few days :.

I sure get it bad with these sorta problems lol, but it's all good, now I know what can be a problem ^^.

 

 

Oh one more thing, the other day, was'nt stable though, I got 3971 read in everest, and 66.7 latency.

That was 275 dc.

I'm now getting 1701 for memtest at 266 :, wicked I guess.

[Clay]

:nod: Magic Clay :nod: Real glad for you. Now you don't need to save all over again :cat:

yeah.. I'm sure happy about it myself. Don't gotta' tear the house up anymore looking for reciept. OCZ has them anyway. I got a note yesterday telling me to expect replacement.(Soon I hope!!LOL)

 

Nice post "Neo" Another saved for reference. I think I'm lookin' for another "Infinity" or LPB. Yesterday I discovered my DVD's on my DFI hate DFI's round IDE

cables. When I first put it together I jus' used the cables I was usin' before. But I was buildin' a rig for a lady and grabbed those one's outta' my rig. figured I'd use the new ones. At anyrate I put the new cables in and everytime I tried to open a cd or dvd on that computer or burn 1. I'd get an instant system restart!!! Finally figured it out! Put some old cables on it. Now all is well... So I guess I don't need them fancy Ide cables after all.For some reason the barracudda 120 likes them.(Ata100 also) though the barracuda 40 didn't.(System restarted everytime I try to open that drive too)LOL So I guess my lil' cuda' 40gig ata100 ide is probably ok after all also. Sumptin' about some ata-33 to 100 drives not likin' them cables. Maybe I should start a thread.(Naw...) But I thought I might post it.

96f here today on the porch in the shade good breeze too. Got the ac on again. Don't want my "puter" gettin' hot.

humm... about the ide cables I wonder if using them enables something that some drive shouldn't use? Well I'm done with those. Hehe they work now. LOL

[NEOAethyr]

Nah dude, it's the quality of the dfi cables that's bad I think.

 

I have a copper shielded round cable, which works just fine.

I cut it in half though, still working fine.

 

This generic black color non shielded round cable I'm using for now, works fine.

 

 

You may be right, my udma2 and 3 drives(33 and 44, my burners) hate these dfi cables.

Maybe that's why I was getting ramdom lockups after burning with nero.

Never got lockups burning with anything else.

 

I thaught that those cables may be a problem but I was able to get round it.

Well yesterday It was made obvious to me that they were an issue.

I was using only my dvdrw, no other drives, and etc etc, trying diffrent things to see what the heck was going on.

I figuered if I did'nt have any drives to hold me back, I could figuer things out.

That's when I discovered that those cables were defently a no go.

 

Ramdomly dissapearing drives, even when my dvdrw was the only drive on the cable.

I used to have the same issue when my cd burner was on that cable.

Some stability issue with that cable too, prbably has to do with it dissapearing, a drive that dissapears when it's supposed to be there would make the system unstable for sure.

 

I never really had probs except for the lockup thing because I allways ran at lower fsb's.

These days I want 275 and everything else stable.

 

 

I was never inpressed by those cables, but they were shielded and round.

I thaught they were ok, but I guess they are'nt ^^.

 

 

I got the cable right here.

Just took the top clip off.

The side I unclipped is fine, however the other side snappend in half.

The cable was made by starconn, so I know know if I see another starrconn cable, to stay away from it.

 

The actual connector came off way to easy, it just fell after I took off the 2nd clip thing.

 

Maybe the cable is'nt starconn, I would assume it's just the conenctor and dfi actually made the cables.

In any case, I think the connector may be cheap.

Whatever though :.

Thanks dfi for such cheapy ide cables lol , only took my a few years to figuer it out ^^.

 

Edit:

Hehe, yeah they work, they work well in the garbage.

[NEOAethyr]

Hehehe, check this out.

 

Note the write speed is freaking crazy.

Also note I compressed all these bmp's into a single rar, with 5% recovery recored and it's small , I don't know what that is about but it's all cool lol.

 

I'm gonn alay down for a nap while I prime this.

I don't think it'll be stable but who knows.

I need some breakin anyways.

 

My 266x8 was'nt s&m stable yet, bummer.

263x9.5, I was superpi'ing it, and it bsod'ed after a few loops.

May be heat releated I dn.

I know I'm not quite up to par when it comes to stability though.

One thing I notcied, it was super fast general latency wise.

Matter of fact, everest got allmost 3900megs a sec at that speed which is not normal, because of the multiplier and the timings I setup, I setup uper slack timings so I can just test it.

 

 

Anyways take a look at my attachment.

This sucker is superfast as well.

Nf3's, eat my dust lol, something like that.

Now only if I could gt the latency down more and such so I can compete with the nf4's.

I need to beable to up my cpu, still I'm not clocked very high so It's pretty good.

 

 

What do you guys think of this?

Especially the write speed, the small rar is pretty crazy too, that's never happened to me before, no matter how simple and simuler the shots were.

[Clay]

near 4k memory writes? 4154 reads? Slack timings! LOL T2? I dunno But that's pretty good! Ya' broke 4k like it wasn't there. @263 dc? Pretty good! The 5k barrier will fall!!!LOL 3.9k memory writes!! my oppie only does 3.3k when clocked up there pretty good.(Onna' fresh boot too.) LOL NF2 Lives!!!......

 

Edit: Geeze "Neo" do ya' think ya' can make it go faster?LOL...... (Those write's are impressive!!!)

Edit2: Yup those Ide cables are crap. Both my dvd 's hate em' and 1 hdd hates em' too. But I get total system-restarts. When tryin' to open the drives.(Not freezes) I thought my lil' 40g barracuda was dead then when I decided to try the other cable on the dvd's then I had big trouble's. Took me awhile to figure that out. These cable's are nasty.LOl Better now though.

Edit3#: Workin' on gettin' my raid in. Had problems with fat32 file system.(Don't know why though. Same problem exactly thatI had about 10 days ago bsod in the same exactly place in the xp install. I thought it was a partially corrupt bios. Cause of the now becoming infamious IRQL less than equil error (Exact same). When I reflashed that time all was good. This time not so good. I converted the file system to ntfs I hope it works. If not I guess I'll run memtest all night. Humm.... makes me wonder fer' sure. LOL

[NEOAethyr]

Those benchs were at 275 .

 

I got 266 dc stable too, just a while ago.

The 275 I had stable before I posted the last time.

 

I'm pretty happy the way things are turning out so far, the last weeks or 2 it's been like I finally get it stable, I goto sleep, check afterwards and it does'nt even pass memtest.

Now it just keeps getting better and better.

 

Matter of fact, 278, when I checked maybe a few hours ago, only had 8 errors on 1 pass of memtest test #5.

I've apperently gotten 278 dc stable once, memtest wise, so if I can get that far, I should be in top form.

 

 

Anyways this 266 , I have'nt fooled with the older set of timings.

Well I was, an I was close, but I laid off that for now.

I just want it stable, plus I got something better anyways...

 

Old timings: 3-3-2-3 9-13 cas3 cmd2 0-0-2-2-0-6-3, vdd 1.8v vmem 3.0v(or 2.9v, depending on the bios)

New like timings: 11-3-2-3 9-14 cas3 cmd2 0-0-2-2-0-0-3 vdd 1.7v vmem 2.8v

 

The tRFC I have'nt checked for yet to see if I can lower at this speed, 266.

 

The whole point was to check to see if I could doit with lower voltages, I figuered I could since I doit at 275, but I wanted to test it to be sure.

Plus 275 is a bit more slack.

Works good too .

 

That new timing setup is actually faster then the old one.

Matter of fact, it's faster even if i have tRC at 11.

It's not tRAS, it's that one alpha timing that makes that much diffrence.

 

If I lower tRAS to 8, I get 1783 or somehting liek that in memtest, and it passed, I never saw my rig rated so high in memtest and pass .

Did'nt boot windows all the way though :.

 

 

For the heck of it, I tried some 2 timings.

I found out with the new alpha setup that in sc mode with my tcc5, I could allmost use another 2 for the main timing.

Been some time now.

 

Ends up it was tRP.

IE, just what I tested:

11-3-2-2 cas3 cmd2 11-14 0-0-2-2-0-0-3 vdd 1.9v vmem 3.0v

Yep needed that high of voltage , when I up the vdd to that my system is rated slower in memtest, I know that's really nothing to go by but still, probably not good.

Passes memtest easy, games for a while too.

Does'nt pass s&m.

Was fun to see it in dc though, not bad for 250 hehe.

 

I swore it was'nt tRP though with my tcc5, I thaught it was tRCD-R, maybe it was..., can't do that in dc though.

 

 

Found another problem with my previous setup.

Again the drives and slews.

Instead of setting up the 4-11 for empty slots and 2-11 for filled ones.

I found that I can now setup my adata to it's optimal setting, 10-1.

I can fill he empty slot with 1-1(I modded the bios, where the missing slot is 1-1, helps out alot it seems).

Actually that mod I think is wht is allowing me to setup my drives and slews anyway I want them...

 

The samsung, the 1-1 which I thaught that was best for it must be wrong.

I do know that adata is right, 10-1, I've used that in sc for some time now.

 

 

Apperently thse are good for my samsung so far(in order of best to worst):

2-10

4-11

1-10

 

Maybe the 1-10 as better then 4-11, I'm pretty sure it was 4-11 that was better though.

2-10 wins straight up.

 

I'm starting to wonder.

If 2-10 should be setup genericly for single sidded sticks, and 4-11 for a generic setup of dual sidded.

 

 

 

There was somehting I forgot to mention before, I experimented agin with the memory dividors.

Got them workin way better .

I even got the all mighty 1:2 dividor working ^^.

That's the one where memory is 2x faster then the bus.

 

The trick is in the 2 timings, sync memory bypass and super bypass.

When you disble them both, dividors work much better.

However when you use 1:1, you should enable both of them, because at certain fsb's, it freaks out about them being disabled @ 1:1.

 

I learned a few things from testing these.

SPD works better then a dividor unless it's 1:1, maybe 1:2 works better I dn(this one is supposed to work good, I did'nt test it again).

Anyways spd boots alot easier then 2:1.

 

Say you setup by spd or 2:1 at 100mhz fsb, so you get 400mhz ram(ddr).

I went to memtest and it was'nt stable ^^.

That's sorta freaky.

Might have to do with the bios I was using a the time.

 

 

I found that the dividorws work off the romsip tables.

This is expected.

However I was expecting it would cross ref a diff table for that memory speed, it does not.

Say you use a dividor at 100mhz fsb to brind it up past 100mhz, say 200mhz for example.

The table used for that is gonna be 100mhz, bad news.

 

Here's the kicker, I tested the hellfire romsip and mine.

 

The multiplier sectin of the romsip, 2nd half, with the hellfire romsip no boot at 275 even if the memory is slowed down.

So it's not to aggressive on the memory, it's the bus.

 

Mine is a fixed like ver, though a tad slower apperently.

So it works at high fsb's.

However the romsip I had in place needed a super fast memory setup to even run or boot(I have an idea of what this may be now, just a guess, more later).

So it freaked out as well with a dividor, becuse the memory was to slow.

 

I'd have to make a special bios, or change out on half of the tables just to use dividors more correctly :.

 

 

That thing about needeing high speed memory, that one offset I've meantioned a million times.

I'm betting, since I know it's directly related to memory and not the bus, that it's straight up the memory's latency timer.

Like old school days, where you had options like 60 or 70ns memory.

If this is truly that timing, then hellfire's bios is setup for 66ns, and mine from 11ns to 00ns!

 

I can't do the 00 with dc yet, I can't get my timings low enough to run it stable.

11 and 33 I'm using right now on this other bios I made yesterday.

33 works good from 250 and up, allowing lower timings and lower voltages for the same sort of setup.

11 only works from at least 272 with my type of ram, and it's gotta be tight, even then I dont' know if it works, 275 works though.

00 only works at 275 and above with my setup, and it needs the tighest timings I can do in sc, with my adata(and my tcc5 can't keep up with those timings).

 

 

I can't remmeber the other thigns I was gonna say lol .

 

I have 2 more thigns to talk abotu though real quicky.

For one, I'm doing 266x9 right now, @ 1.65v vcore, s&m stable, priming now .

 

For 2, a buddy of mine appenrently found out how to add new options to a bios, on his via boards.

You never know with sorta stuff though, then again the stuff he's been saying from time to time ends up having some truth to it.

 

Like the hd cable thing where you plug the middle of the cable into the mobo for more speed(shorter cables), then the lba thing on cd/dvd drives ends up being true...

This new thing, breakin in the fpu using gtasa seems to do well, very well.

Only messed with it once a day or 2 ago, some decent results(nothing that allowed anything new yet).

 

 

I won't know until this next month though, if I can even get a ride up there.

I know it's not modbin because I allready teached him that some time ago.

So who knows :.

Would be cool though if I can add new stuff to our bios, like tREF, that would probably be the 1st thing.

Then again I don't want to get anyone's hopes up on that one, it may be nothing.

 

That cable thing...

You have to use the short end on a cd/dvd drive, and the long end on a hd.

If you only got a hd hooked up you can use the short end.

Lowers the access time alot, however the bios thinks you are on udma33 cables and you have to force it back up using the nv drivers.

Even at udma33, it's faster then 100 using eh standered cable setup.

 

With a single hd hooked up to a channel that way, I get 0ms access time in sandra.

Then again sandra is a bit wacked, but it is faster, you can tell, it's one of those type of things lol.

 

 

Later dudes

[NEOAethyr]

Just got 276 and 277 s&m stable in one shot.

I got 2 errors @ 278, this is dc btw, same timings as usual.

 

11-4-2-3 cas3 cmd2, 0-0-2-3-3-0-3

1.475v vcore, 1.7v vdd, 2.8v vmem

8x multi, 1:1 ram

 

 

Ruffly another 40mins(actually less) I could most likely get this stable and perhaps another mhz in fsb, like the jump I did from 275 stable to 277 stable.

For now I'm just gonna prime it while I surf.

 

I am quite happy the way things are turning out.

 

 

The rails, god d@mn this sucks...

In the morning, I woudl guess the ocz psu is running in the mid 30's to upper 30's, in c.

My fluctuation of the +12v rails goes from under 0.01-> 0.01, to 0.03-0.04 .

This makes me alot less stable.

Not onyl that it doe'nt decrease, it increases at that time, making things worse probably.

 

When I try to compensate, it never works out, I can get memtest to behave, but not s&m.

However the more I work on breaking it in, I believe it'll start behaving, plus I did a few things to help...

 

The +5v rail goes from 5.00-5.01 max, to 5.02 to maybe(have'nt checked this until recently so I can't remember) 5.03 :.

This may be the main cause, not sure.

 

The +3.33v rail does about the same too.

It goes nuts in the morning, making things very hard.

In the afternoon, it fluctuates less then 0.01v, something I can't even measure.

 

 

Most of the peeps on this forum would tell me that it's not the psu, that's well well within good specs, well bs...

I actually need it not to fluctuate, I can only have a max of 0.02 going on the 12v, and a max of 0.01 on the +5 and +3.33.

 

 

Anyways before the day broke, I took it apon myself to hook up my enermax psu as a secondary psu, to lighten the load off the ocz.

Right now I got all my fans on that enermax.

After a while I'll add my lights to that enermax psu as well, when I get new sticky pads thingys.

 

This makes it so when it is fluctuating, it fluctuates much slower, still the same amount though.

 

I also redid all my cables arrangements so that each drive has it's own cable, I did not re-wire anything, just that the hd's have the powershield ide cables, which are single ended.

The dvdrw has my modded ide cable ground, but it's on it's own, at the 1st conenctor.

This helps alot, I noticed a speed increase by eye in all drives.

The cdrw I can't put in this rig, my case is not big enough, if I did it would go under my dvdrw, and my dvdrw would get hot, I can't have that.

 

 

My +12v rail, I've been adjusting this during the day to try getting it stable liek it is in the afternoon and night.

With no luck like Iv'e said above.

Anyways I found that I can use ruffly around 12.07-12.08 wihtout probs.

I get a pseed increase as well as extra stability.

I'm right around the optimal level with this rail, for now it's staying.

I would'nt beable to do much better, I'd liekly make it worse.

 

The 3.33v rail, I tried adjusing this again today.

This is the 2nd time after I got it stable in dc with 266 and 275.

This time however Ig ot it more stable, but at a price, my wram boot issues are more frequent.

This is because of those newer bios ver's, ie the 728.

The 505 I would'nt have issues with, I can't help this for now until I make a new bios from scratch based apon the 505.

So I'm dealing with it, plus my warm boot issues are going away with breakin.

I have not reset my cmos since this afternoon so It's hard to say if I would have trouble booting say 266 for example initially.

The optimal level is right around 3.33-3.34 or 3.35.

On the low end of that.

Higher or lower is very bad news on my rig, very bad.

Higher gets better boots, however the better ver is not measireable by the dmm, it's in the blind spot, and it's less stable.

This is by sata, measuring this by the atx has only got me trouble, probably because I'm not used to it.

 

Funny how if I adjust by certain conenctors in certain cercumstances I can get it stable, then sometimes a diffrent one.

:

Tjhe way I got my cables now though, it shoud not be hard to get decent ref values for later use, but liek I said, my rails are going to stay this way for at least a week.

I'm gonna try for a 24/7 stable thing, prime is 24/7 stable if I leave the rails along, but s&m is not.

I hope to change that.

Not even stable at 266 when the crap hits the fan :.

 

 

The +5v rail is setup I think it was 5.01 and 5.02 optimally, if it's off by heat, it's 5.03 or something.

 

 

Further inporvements woudl involve a complete re-wiring of the ocz 600w powerstream with much better gauge cable.

Though that atx I would have to skip on, the rest would help alot, more so then filtering.

Inproved filtering would help alot, however I would have to compensate.

Filtering adds resistance, I checked this today with a few experiments.

So in order to get around that, I would have to take off the epoxy of the 2 vr's inside the psu, for the sperate +12v rails, add them externally(no ready for this yet anyways), then adjust them accordenly.

This would allow me to get the drives down pat to the optimal speed, and adjust to optimal stability.

Allowing me to add more filtering as well, without unbalncing the rails in favor for one device or the other.

However I got this +12v rail setup right now pretty high compared to the other day, so who knows if I truly NEED it yet.

Breakin at this level should do the trick.

 

 

When you adjust the rail levels, it takes a little while for your rig to break into that.

So if you keep adjusting looking for a quick fix, it's not likely gonna happen I'm afraid.

This is one of the reasons I've had so much trouble figuering stuff out.

It's something that would likely piss anyone off, I know it pisses me off lol(not trying to be offensive).

 

 

How am I doing this breakin now these days?

I found that I can totally skip memtest, even if memtest is failing, I can get it stable in a matter of hours max ^^.

Prime95 with the max sized fft works allright, but it's the slowest method.

40096k-4096k inplace mode.

I'm using my little custom read only config.

Takes around 47mins to 3 hours and 17mins to break in a fsb, even if I'm skipping straight to 266x9 or 275x8, etc.

Those are just example of times I stoped it and tried s&m.

 

Superpi however does the trick with a singole pass of 32m.

Which takes ruffly 38mins for me at 275 and above(I got a 36 recenlty from some speed, but I don't know what).

After this is done, every single time the last 3 days it has broke in the fsb.

This last time, I got 1 error in test#4 in s&m, did the suerpi thing, bam stable(276 dc), and 277 dc was automaticly stable as well .

 

I've also ran prime95 while watching tv on my rig, using graphedit, as well as playing some gtasa marking locations for teleportation, etc.

And surfing the net.

 

I got 2 errors at 278 in s&m test#4.

Been running prime since just before I got on the net to type this.

Still going strong.

 

If I'm failing test#1, 2 or 3 even, superpi or prime does it too.

If prime was'nt enough at the time, the for the short peroid I ran it, I either run it more, or use superpi.

It does the trick, though last night it got me 3 errors @ 276 test#4 s&m.

Tonight no errors (Remember rails were re-adjusted so I needed to get used to that) .

 

I forgot wat else I was gonna say :.

Ohwell I guess.

This speed is quite nice, at this mhz it woud put my at par with 280 single chan at the same timings.

That's what I calc'ed long long ago, howeve I'm using much much diffrent timings now and much lower voltages too , and I'm actually gettting it really stable, not just gaming stable - I thaught it was stable sort of thing.

 

I can't wait to get 280mhz fsb 1:1 dc stable and beyond, I can only imagine what 283 and 286 will look like hehehe.

 

 

Oh I rmember what I was gonan say.

Prime95 with that setup of max fft's, is like running somewhere past 12+hrs of 102k-4096k, but in 3hrs.

Blend would probably take days to gets what I can do in an hour ^^.

 

 

 

Edit:

Oh yeah, the +v rail I setup at 5.02-5.03 for optimal, sorry, fluctuation must be .04 or somehting, I forgot.

I can check right this sec while priming...

 

Well I don't normally check on load because I can't use tose for ref, but I'm getting 5.01-5.02 on the +5v.

While priming at 278 dc x 8.

Now dvd or dvdrw in my drive, that was off a totally unused cable from the ocz psu.

An ide cable.

 

 

I would liek to increase my +12v rail on my enermax too sometime, it's around 11.73-11.75v :, it only powers the stuff that does'nt need stability, ie my fans.

Since it's used for stuff like that I'd like to crank it past 12.40v or something, but I would have to take a serious look at the inside of it and take measurments and etc.

 

 

Edit:

278x8 dc is stable , prime all is all it took

279..., gotta try that in a few.

[Clay]

280 will fall!!!! Neo isn't 278fsb prime stable 1-1 (No divider) a "DFI-street" record? Next time you might consider lettin' us force a toaster on ya'.LOL

 

@Kitfit Thanks the Raid0. Werk's like a dream.(Funny I forgot how I did it but it werks great!)

LOL

[kitfit1]

Don't forget to post us up a screeny when(not if)you hit 280 Neo. I will be flashing to Suzuna again tommorow with some 4800 plats on board. Will post up the results(fingers+toes crossed) :nod:

 

@Clay, glad the raid setup went ok. How did you split it up partition wise?

[NEOAethyr]

I'm glad you guys belive in me still and all, I was leaving the scene, I'm pretty much gonna stay around in the background but that's about it, posting once in a great while.

 

I can get 278 going well, and even 275x9, both of which can pass superpi and s&m.

However when the heat is on I'm screwed.

I can't even get 266 stable with the slackest of timings when it's hot in here.

 

I'm gonan redo my gel on my cpu for the heck of it.

Even though it's only been like a month, on ceramic, it seems that I have the worst darn luck with gel going out on me all the time.

It's allready hitting 38c and my south bridge at 36c!!

 

Last night, 277x8 @1.475v unstable s&m memory(because of heat, passed before easy at this votlage).

However right afterwards, at 1.5v, passed easy.

For the heck of it, 1.525v failed big time, errors all over in s&m, I hit I think it was 38 errors before I stoped it.

 

It's no wonder why adding my torado helped alot, it probably was't the extra cooling by moving the panaflo to my ram and north ridge, but the actual cpu it's self.

I do know that extra cooling on the ram and north bridge helps a bit though.

It uysed to keep my stable @ 3v vmem and such at 266 when it was hot.

 

 

I'm gonna try as5 again.

I'm starting tho think my 1st results were from somethign else,, liek breakin, the less vcore needed thing.

I'm betting as5 is better for better surfaces, and ceramic for those uneven ones I guess.

[NEOAethyr]

Redid my gel and sorta redid the spacing int he pins of my swiftech heatsink.

Anyways my warm boot isues are gone now because if that.

 

I assume the reason why 506 doe'snt have the warm boot issues is because it has the idle tweak built in.

I'm using my 728 mod though because it's more stable, maybe it's the bpl/nvdamc I dn.

But it's more stable even with the same romsip.

So I will need to figuer thart oen out before a new official release of my bios.

 

I'm still not s&m memory stable on 280, but I have not tried breaking that in either.

Matter of fact I'm actually supposed to be trying for 279 still...

 

But anyways, it passes the fpu test, short length is all I checked, at 1.5v, 8x.

I'm priming it now, 1024-4096k fft's.

 

I wikll relese this very same bios that I'm using to do all of this in a few hours, at best maybe 10 or so, I promise, but remember it's a beta.

The slots have been switched around too..., so it may take a bit to get used to.

When i tried the 506 again I was having trouble remembering what slot was what lol, because it was stock.

 

Anyways in a while I will, I ust want to get through tonight and see what I can do.

My bedroom has been re-arranged so it does'nt get as cold, so it's a wee bit harder to clock up there like I was.

 

 

 

Edit:

New beta uploaded btw.

Should be ok for testing anyways.

Give me a heads up on sata, if it even works, I'd liek to know because I never got feedback on that one :.

The reason being is because it's using the raid 5 rom.

 

Btw, warm boot issues gone with gel redo.

And tasl switching probs are gone with that too, ie it took me maybe 10 secs with 512m of memory to switch out of gtasa to the desktop.

But it was instant with 256m.

Now it's instant with 512m, I thaught it was a win2k prob :, I guess not lol.

The cpu is becoming an interesting obsticle ^^.

 

Oh, and the darn ceramic gel left some crap on my cpu I can't get off...

Like as5 does.

I must of used to much, err...

Now I will have to give the cpu another bath to get that film crap off it, will do it later someday...

[Clay]

@kitfit: On the raid0 I went 15-38-42-42.(I think..) The last 2 are "ntfs" Sandra sez C: 99mbs though "pcitstop" sez 120mbs. I haven't installed "atta" or hd tach yet. To see how it compares with some others. But I will today. I made the first part a bit bigger because when partitions start gettin' more than 1/2 full they slow down a bunch. More so than if they fairly large to begin with.(More than 40gigs is a no) Larger than 7gigs the seek time goes up.(<98 thinkin') Plus I have some apps that insist on being on c: (Coupla' games "Ubisoft" no option to install elsewhere(But I haven't put em' in inna' long time)) But all in all I'm really happy w/that.

 

@"Neo" What do you use to get clean yer' processor.(A bath?) Because I have an XP3000 that I got as5 all over.(It works OK, but that can't be good) Unfortunatly the body of this one is that organic crap Looks like baked crap!!. Which I think the as5 may have soaked into. But next time it's out, I thought I might give that a try. Maybe not on the count that the body is that organic stuff and water? Might also soak in.make it soft? I dunno I still might if I can do it quick.