New 3DMark06 WR with Phenom II X4

[monteboy]

@Team Finland
concrats guy's impressive work !

What's the main purpose of that slip fit?

The F1 is ideal for the Phenom, really. Never had a 7-8h bench session using not even half of my 35L dewar with intel rigs

Copper Mass can only be replaced by more Copper Mass

Hope i can start next week my first Deneb LN2 session with this babys.





7 Kg copper mass !

[SF3D]

Monteboy.. that is the old way of thinking. It is not working just like that with CPU's and huge heatload from one small spot. You need to have grooves or holes in the base and certain mass. You have to find the right balance.

Huge mass doesn't help in the top of the pot and huge mass do only harm in the base, cause heat must travel more to be transfered away. Copper is not ideal to hold cold, but it can transfer heat quickly.

This is the reason why F1 and my own pot are different from these traditional designs.

[jellyrole]

Great job with that score!

[monteboy]

Monteboy.. that is the old way of thinking. It is not working just like that with CPU's and huge heatload from one small spot. You need to have grooves or holes in the base and certain mass. You have to find the right balance.

Huge mass doesn't help in the top of the pot and huge mass do only harm in the base, cause heat must travel more to be transfered away. Copper is not ideal to hold cold, but it can transfer heat quickly.

This is the reason why F1 and my own pot are different from these traditional designs.

Copper is even better to hold cold as aluminium. I agree it is importent to find right balance but for what application (CPU) ?

With a massive copper pot on Intel chips you have two problems :

- first if you run in cold bug. It takes lot of time until the pot warm up.
- second problem you need more LN2 to cool this monster pot down.

On Deneb you don't have this two main problems.

Once you cool this massive copper down your LN2 comsumption would be slightly because no coldbug. Only what you need is to hold up -185°C

And as far as I can see for this application the main think should be mass.

3 stairs and 16mm (diameter) nipple


Phenom II Die Size is 258mm² and the area form nipple is 201.062mm² , i agree a little to small for this CPU.
Perhaps it is time to think about a new Base for my pot

We will see how the old way of thinking perform

[massman]

The F1 currently has enough mass to hold the load of the Deneb well enough, so I don't think more mass is needed. Imho, designers should be working on a container that allow the container to go faster to -185°C, but has the same mass to keep the load

[KELL5]

I ain't no ln2 pot guru, but have you guys tried creating small holes inside the pots on the smooth surface to increase surface area which is in contact with LN2? Much like a cpu waterblock has raised fins inside to increase surface area and heat exchange but in reverse?

[monteboy]

The F1 currently has enough mass to hold the load of the Deneb well enough, so I don't think more mass is needed. Imho, designers should be working on a container that allow the container to go faster to -185°C, but has the same mass to keep the load

Nobody said that F1 did'nt have enough mass. F1 is a great pot.
In some sessions with my friend joe_cool (F1 owner) i had a chance to compare with the Undertaker So i know very well what this pot's could perform

[SF3D]

Too much mass in the base will still be a problem, even if you are using Deneb. If you start for example Wprime 1024 and watch the temps, you can see, that it will be stable first and then it is starting to get warmer and warmer.. If your base is not balanced right, you can't do nothing. You have a full pot of LN2 and the temperature is still raising up. Very light applications will not suffer from this and heavy pot might work well.

Monteboy- Let us know how it is working.

In general, everyone should remember to check how well the sensor + thermometer can read temperatures.. I have three different sensors and only one is reading correctly at -196C. Othe one show -160C and one -175C

[Foamy]

Impressive results Team Finland Now freeze those cards and perhaps with hard pushing 36K...

AMD back on top, i like

[monteboy]

Too much mass in the base will still be a problem, even if you are using Deneb. If you start for example Wprime 1024 and watch the temps, you can see, that it will be stable first and then it is starting to get warmer and warmer.. If your base is not balanced right, you can't do nothing. You have a full pot of LN2 and the temperature is still raising up. Very light applications will not suffer from this and heavy pot might work well.

Monteboy- Let us know how it is working.

In general, everyone should remember to check how well the sensor + thermometer can read temperatures.. I have three different sensors and only one is reading correctly at -196C. Othe one show -160C and one -175C

PT100 is the Answer for you.

I'will present soon a selfmade 5-channel - thermometer based on PT100 resistor/sensors with cheap components. What we need here is not exactly measurement.
It is importand to receive fast response time and therefor PT100 is better than K-Type.

I think your thermometers are designed for K-type

[Foamy]

PT100 is the Answer for you.

I'will present soon a selfmade 5-channel - thermometer based on PT100 resistor/sensors with cheap components. What we need here is not exactly measurement.
It is importand to receive fast response time and therefor PT100 is better than K-Type.

Paired with this

[monteboy]

Paired with this

I'am a poor man
Don't have enough money for that

Prefer, to build it in may way and write also the Monitoring software.

This Thermometer project will be specially developed for LN2-cooling measurement (-200 .... -50°C)

simple cheap components :
operational amplifiers
cheap 8-bit AD-converter
small ceramic PT100 - resistors

[Foamy]

I'am a poor man
Don't have enough money for that

Prefer, to build it in may way and write also the Monitoring software.

Yeah it costs a bit, in may case i could get it for free due my work. But it would be could to have this device controling your LN2 flow with a valve or anything compatible with an LN2 dewar..

anyhow nevermind offtopic

[boblemagnifique]

Impressive freq & score , congratz for your oveclocking SF3D , Sampsa & Macci

[monteboy]

Yeah it costs a bit, in may case i could get it for free due my work. But it would be could to have this device controling your LN2 flow with a valve or anything compatible with an LN2 dewar..

anyhow nevermind offtopic

you have seen my LN2-Control on AOCM2 ?



I use conventional valves the next version should be better
Ln2 valves are expensive !

In combination with this cheap selfmade PT100-thermometer this could be a nice equipment

[Foamy]

you have seen my LN2-Control on AOCM2 ?



I use conventional valves the next version should be better

Wow sweet

[SF3D]

Guys.. the storm is coming tomorrow.. I can feel it in my bones

[Lightman]

Guys.. the storm is coming tomorrow.. I can feel it in my bones

And it was such a lovely day today, who would think that storm is coming?

I hope after the storm it will be like this (LN2)


or if LHe involved, then like this



PS. Winter in Poland, Europe. First picture presents a car

[jugeen]

who cares the car, tell us more about second picture

[G H Z]

poured in the last drop halfway thru Canyon flight and the pot was empty at the end of that test.

Was it full @ the beginning of the test?

[FlanK3r]

yes.. who is a girl on second photo :-)? She looks good as Phenom II X4 945 :-D

[dpa]

shes old, and she has a burberry bag

[posershadow]

If you run out of ln2 during the bench, just start shoveling some of that snow in the pots and hope you make it long enough to finish and save your wr's.

[dinos22]

Andre

I am a bit depressed about overclocking circles as of late and have been taking some R&R this summer.

I hope that some of the smart overclockers here will refrain from side taking and protect what we all hold dear and what all the big companies dont get. I have absolutely nothing against any of these companies being a part of our circles but when it comes to affecting our competition I feel depressed. It is getting out of hand.

We need to engage HWBOT or XS or form a new body that would work towards forming a strict, independent and rigid platform which to follow to protect overclocking as a sport from desparate manufacturers one upping eachother over our backs.

I think we should be more agressive about this then we've ever been. Not only is it not good enough to claim a W.R. on unreleased hardware we should completely stop engineering samples being accepted particularly CPU and GFX manufacturers as they make the biggest impact on scores and any other in future that will significantly affect benching (i.e. HDD manufacturers with PCMARK suite for example). The reason why i think ES hardware has to stop entering our competition specifically is primarity to manufacturer superbinnig we have seen from both Intel and AMD in particular from their latest core i7 and Ph2 platforms.

A superbin chip is not a normal ES........these are chips a cut above the rest where no retail or almost impossible to be 100% confident of finding another chip like that from retail sample of say 20 chips and personally binning them. I think both manufacturers have used their binning power to give Ocers access to chips far better than that which we have never had happen before. i knew i could go and find a better retail E6700 or E6600 than the absolute best ES CPUs that came out only shortly after release. This is NOT the case with i7 for example and AMD from what i can see as well.

We should prevent scores from being accepted by manufacturers or their employees. This is different from people receiving hardware to bench as long as it doesnt become another avenue for manufacturers to feed their hardware....in that case it could get more complicated and people getting extremely good overclocks against other benchmarkers could be one sign of "manufacturer seeding" behaviour which should be stopped

I reckon since this is becoming so professional now as Charles said recently that we should then treat it as such. Other professional sports have super comprehensive rules & penalties over their sport so should we no? We should create new categories of benching in addition to current ones where users are encouraged to get the most out of the system.........lots of examples like 32M SuperPi challenge and many 3DMARK challenges created by legends like Vince and others wanting to see how we all stack up with same set of rules. How about low budget challenges we had as well....that was a massive hit and lots of fun.....This enables a lot more people to participate as well and some new names have surfaced showing off enormous skill that would have otherwise gone unnoticed with wallet battle style comp we are a part of right now.

This would also create a good platform for international competitions as currently they just cannot be created fairly due to not all hardware being created equal and being absolutely impossible to bin it to all be the same.

We should probably look at some disclosure related clauses as well in terms of hardware support and general support being received from manufacturers as well to improve transparency and promote openness.

I know that there will be a lot of people huffing and puffing about my post here but i tell you what DONT post a response, think about it, sleep on it and then think about it some more. i know i have for the last coupe of months. I am sure you will see where i'm coming from and try to deal with this constructively

Dino

[dinos22]

sorry fellas for the thread hijack

monster CPU to say the least

liquid helium will gain how much?? 200MHz
man freeze those cards and show the real results heh

or if LHe involved, then like this



PS. Winter in Poland, Europe. First picture presents a car

holy
look at that snow

life is hard in Sydney i have to say my foot sunbathing on one of the beaches close to that shark attack hahahah



summer rules