XS HWBot Team

[G.Foyle]

I dunno what kind of nail polish you have in the US but it works fine for one time, until you defreeze the board. It will start coming off after thermal cycle.

[l0ud_sil3nc3]

got a dewer, f1 and a thermos coming ver ysoon so i am starting to prep a few trial boards to learn some LN2

i just ordered some

MG Silicone Conformal Coating 422-55ML
MG Chemicals Super Corona Dope 4226-55ML

but i read you can use nail polish, so whats the dope on what is best...

I went out and grabbed some of the Frost King pipe insulation per Kinpin writeup on his forum and I think I am gonna try that route first.

The other option is kneaded eraser the out of everything

not sure on whats the easiest to remove, or what insulates best, however these are both factors to consider when insulating.

3oh6 had a nice write up on kingpins forums as well, its a little different than vinces but they both do the same thing, prevent condensation.

[systemviper]

yea right now i use the eraser and it works great, but i have a few boards that i don't mind insulating for good...
and i want to get full coverage

glad i didn't use nail polish...


thanks for hte info guyzzzz


saw this on crazy PC and ordered some...

MG Chemicals Super Corona Dope 4226-55ML
Quantity in Basket:none
Code: super-corona-dope-934901
Our Price: $9.99
Stock: In Stock






Quantity:

MG Chemicals Super Corona Dope 4226-55ML - a conformal coating that is brushed on PCB boards. Insulates electrical components from moisture, and oxidation. Excellent arc and corona resisting properties.

Features

MG Super Corona Dope

Comes in 55ml (2fl oz) bottle with brushes

Useful for high voltage applications

Insulates components circuits and assemblies

Excellent arc and corona resisting properties

Protects motor windings and coils from arcing and discharge

Protects against moisture and oxidation

Only one coat is required (apply additional coats if extra protection is desired).

ROHS compliant

Specifications

Physical Properties
Type Modified Alkyd
Color Clear
Thinner Xylol
Non-Volatiles 35 +/- 1%
Specific Gravity 0.916 - 0.936
Viscosity (Brookfield, 25C) 105 cps
Viscosity (Demmler #1, 25C) 15-30 seconds
Electrical Properties
Dielectrics Strength, Heat Cured 4100 Volts / mil
Dielectrics Strength, Air Dried 3000 Volts / mil
Cure Properties
Tack Free Time 10 min
Heat Cure (110C / 230F) 30 min
ASTM Dry Time (25C) 20 min
Recommended Curing Cycle @ 125-150 C 3 - 6 hours
Maximum Temperature 180 C
Physical Properties
Moisture Resistance Excellent
ASTM Oil Resistance Fair
Adhesion to Steel Excellent
Adhesion to Copper Excellent
Other Properties
Flash Point (uncured) 27C (81F)
VOC 5.01 lbs/gallon
Application Methods Dip and Bake
Storage Condition Stored at 25C in a dry condition


* For maximum dieletric strength, this product must be heat cured for 3 hours @ 135C (275F).

[namegt]

amd 6-core low clock
Just for fun ...
6core 401.3MHz

http://valid.canardpc.com/show_oc.php?id=1355543







http://www.youtube.com/user/PWNAMEGT.../1/Gy5ojsJQ2NI

[Don_Dan]

You're doing it wrong, you want your frequency to go up, not down!




Edit: We broke 15200 points today, Marco André made the submission that pushed us over the border!

[Marco André]

My humble submission? Glad it helped somehow

[trans am]

Uncle steve, Im proud of you buddy!

[Don_Dan]

Thanks to chew*'s submissions we passed 15300 points today!

That means only 2000 points are needed to catch JMAX-Hardware, should be doable in the next few weeks....

[Linuxfan]

Thanks to chew*'s submissions we passed 15300 points today!

That means only 2000 points are needed to catch JMAX-Hardware, should be doable in the next few weeks....

Lets see it happen guys bench all you can!

[trans am]

I want to bench and Ive had a full dewar sitting since last Friday and its been raining all week. I hooked up the board and ran some ln2 last night for a half hour and the ram was already wet.

[Linuxfan]

I want to bench and Ive had a full dewar sitting since last Friday and its been raining all week. I hooked up the board and ran some ln2 last night for a half hour and the ram was already wet.

Gotta keep er insulated man.

[chew*]

Once again I can only extend the offer........

I have to cpu power to help you guys out in 3d, bring your cards and some loot for ln2 and we can get you some decent results if your local to me.

[Linuxfan]

Once again I can only extend the offer........

I have to cpu power to help you guys out in 3d, bring your cards and some loot for ln2 and we can get you some decent results if your local to me.

When I make the trip "Up North" then I'll try to stop by your place for sure. I'll bring some HW too.
Hope some of you guys can take advantage of this opportunity to learn and pound out some great runs!

[Marco André]

Hello,

I was focusing my attention on 03 because I read somewhere that's very tricky and require a good configuration... I asked google for help on the tweaking side and found this thread: http://www.xtremesystems.org/forums/...=130155&page=3

After reading Hipro5 comment (when he figured out the trick) I started running the benchmark.

980X @ 4292MHz
with 1x 5870 got 103k
with 1x 4890 got 91k
with 1x 5670 got 5,5k

after like 25 runs with each card, i found that my submissions had been blocked

[Leeghoofd]

Let's have a look at your scores :

980X @ 4292MHz
with 1x 5870 got 103k --> your entry in between LN2 cooled cards that run over 1400 core. Your validated score is 90K. This is a huge jump for just a reboot. Don't you think it is a bugged run ?
with 1x 4890 got 91k--> you even beat Kingpins LN2 card on stock cooler :p
with 1x 5670 got 5,5k --> score is still validated, yet it's another bugged run for sure. Ryba runs 1500 core and you 850 ?

I really don't think the reboot trick is that good... sorry before complaining about the moderations get the facts straight sir.

[Marco André]

I just didn't knew any bug or see any during the benchmark but yes it's a big jump, in fact. Well i'll keep going with low scores, no worries Thanks for the help!

[Leeghoofd]

Beating an LN2 cooled card of the same type, that has about +30% higher clocks is impossible in 03 as it's so heavily GPU dependant... logic rules

Though in 05/06 your GT could be able to make up if they ran their scores on core2 or I7... till the moemnt they have a GT too :p

[Don_Dan]

after like 25 runs with each card, i found that my submissions had been blocked

Actually I reported your score with the 4890, that's why it was blocked ( I didn't check the other scores ). Running it after a reboot is a nice tweak, but it will never give you such a huge gain in FPS, in fact your GT4 score looked bugged to me, GT3 was extremely high for the clocks as well.

[mk-ultra]

marco 3dm03 always produce bugged scores when you're near instability, check sub-scores to see if they are in line with the others around your final score to avoid this

btw

good job from the "new" guys and the olders too, i'll put some boints soner than later.

[Linuxfan]

marco 3dm03 always produce bugged scores when you're near instability, check sub-scores to see if they are in line with the others around your final score to avoid this

btw

good job from the "new" guys and the olders too, i'll put some boints soner than later.

Good to hear from you.

[slaveondope]

Someone let me borrow a 980x so I can get top rank in dual 5850's

Lost my 5th place in Vantage due to everyone running 980's above me and Ive got the better gpu score

[Marco André]

Thanks mk, i'll keep uploading all the results I can, less 03 runs with high scores

[G.Foyle]

Lost my 5th place in Vantage due to everyone running 980's above me and Ive got the better gpu score

Story of my life.

[Don_Dan]

We passed 15400 points today...

[Linuxfan]

We passed 15400 points today...

Goin' up is the only way to go here at XtremeSystems!