Itanium's MASSIVE Heat Sink

[Lithan]

Am I crazy or did AMD not get some crazy patent on using tec's to cool processors recently and everyone laughed at them?

Heatsink is a beast, but nothing really new. I remember the first opteron heatsinks (skived fin, someodd hundred fin pure copper) sinks had people gawking too. The Tec is a shock though, especially if it only dissipates 100w. Intels current cpus are speced up to 115+w and they get away with a crappy aluminum HS/F with no tec. Why would a supposedly cooler chip want a tec and a beefy copper cooler? Even at 150-200w a tec is certainly not necessary. If the Inq is right here... this could be very bad news for fans of the itanium.

 

[Wingznut]

Originally posted by: Lithan
If the Inq is right here... this could be very bad news for fans of the itanium.

Did you happen to read ANY of what pm wrote??? Or did you just happen to pick and choose what you wanted to?

 

[BW86]

thats a big heatsink :Q

 

[Lithan]

Originally posted by: Wingznut

Originally posted by: Lithan
If the Inq is right here... this could be very bad news for fans of the itanium.

Did you happen to read ANY of what pm wrote??? Or did you just happen to pick and choose what you wanted to?

Inq could be wrong, hence the "if". But I'm not going to take the word of some guy on a forums just because he says he works for intel. Frankly I'd be worried how good the inq's cooling engineers were if he is wrong/lying. But I think that the idea that intel has to throttle it's chips (as the article he references suggests) simply to achieve the speced thermal properties is much, much, much more of a red moon.

 

[Wingznut]

Originally posted by: Lithan

Originally posted by: Wingznut

Originally posted by: Lithan
If the Inq is right here... this could be very bad news for fans of the itanium.

Did you happen to read ANY of what pm wrote??? Or did you just happen to pick and choose what you wanted to?

Inq could be wrong, hence the "if". But I'm not going to take the word of some guy on a forums just because he says he works for intel. Frankly I'd be worried how good the inq's cooling engineers were if he is wrong/lying. But I think that the idea that intel has to throttle it's chips (as the article he references suggests) simply to achieve the speced thermal properties is much, much, much more of a red moon.

So, instead of using information from the actual source, you choose to believe whatever you want to believe... Using theinquirer as your primary source of information.

Hope that works out well for you.

 

[pm]

Originally posted by: Lithan
Inq could be wrong, hence the "if". But I'm not going to take the word of some guy on a forums just because he says he works for intel. Frankly I'd be worried how good the inq's cooling engineers were if he is wrong/lying. But I think that the idea that intel has to throttle it's chips (as the article he references suggests) simply to achieve the speced thermal properties is much, much, much more of a red moon.

My email address at Intel is in my profile - feel free to email me.

What's a red moon?

 

[Thermalrock]

i think a red moon is the opposite of a blue planet. i could be wrong on this one tho.

 

[IntelUser2000]

"Inq could be wrong, hence the "if". But I'm not going to take the word of some guy on a forums just because he says he works for intel. Frankly I'd be worried how good the inq's cooling engineers were if he is wrong/lying. But I think that the idea that intel has to throttle it's chips (as the article he references suggests) simply to achieve the speced thermal properties is much, much, much more of a red moon."

Doesn't that contradict your statement since the article saying Montecito will run cooler than current Itanium and won't require a big heatsink came from the Inquirer. Other than the fact that you are a AMD fan and a troll I can't see where you got that from.

 

[IntelUser2000]

"yea its total system power. and the 3500+ winchester is freaking awesome 114W whole sytem under load. im so gonna buy that cpu."

Why does anyone care about desktop system's power consumption if in A64's case they are already low? Especially when you pair it with 430W power supplies. Power supplies don't have states between idle and full power so you are consuming 430W whether you have rest of the system consuming 350W or 100W.

 

[IntelUser2000]

"But I think that the idea that intel has to throttle it's chips (as the article he references suggests) simply to achieve the speced thermal properties is much, much, much more of a red moon."

Who cares if the processor throttles if it meets thermal properties and maintains performance, in this case you'll get better performance out of it.
The technology used in Montecito is much more advanced than SpeedStep or DBS, and in this case if your processor consumes less than 100W it will clock the processor up temporarily to increase performance. Whatever it is you are an idiot who argues with ITS feelings rather than facts.

 

[Thermalrock]

Originally posted by: IntelUser2000
"yea its total system power. and the 3500+ winchester is freaking awesome 114W whole sytem under load. im so gonna buy that cpu."

Why does anyone care about desktop system's power consumption if in A64's case they are already low? Especially when you pair it with 430W power supplies. Power supplies don't have states between idle and full power so you are consuming 430W whether you have rest of the system consuming 350W or 100W.

thats the most stupid thing ive read in a while. hahahahaha.

 

[Thermalrock]

do you actually enjoy making bs up and embarrassing yourself or do you actually believe that nonsense?

 

[Lithan]

Originally posted by: IntelUser2000
"But I think that the idea that intel has to throttle it's chips (as the article he references suggests) simply to achieve the speced thermal properties is much, much, much more of a red moon."

Who cares if the processor throttles if it meets thermal properties and maintains performance, in this case you'll get better performance out of it.
The technology used in Montecito is much more advanced than SpeedStep or DBS, and in this case if your processor consumes less than 100W it will clock the processor up temporarily to increase performance. Whatever it is you are an idiot who argues with ITS feelings rather than facts.

How do you figure a processor throttles down without impacting performance? You do know what the term "throttling down" means?

Doesn't that contradict your statement since the article saying Montecito will run cooler than current Itanium and won't require a big heatsink came from the Inquirer. Other than the fact that you are a AMD fan and a troll I can't see where you got that from.

What you quoted explained my opinion on the situation in either of the proposed cases. As well as an explaination for my conclusion drawn earlier. But don't let that stop you from deciding that people who you don't understand are "trolls".

pm

My email address at Intel is in my profile - feel free to email me.

What's a red moon?

How many people work at intel? How many are in a position to know the specific details behind this new processor? I'd like to see some white pages before I accept any claims as sacrosanct.

"Red moon" has a double meaning. First as a bad omen, second as a "sign of great heat".

thermalrock
thats the most stupid thing ive read in a while. hahahahaha.

Most stupid in a day or two for me. If you really look around anand you can find some really stupid stuff fairly easily. Someone gets their 2.4c running 3ghz and starts thinking they are qualified to be an electrical engineer. The reason I don't consider it the most stupid is that it isn't exampling a basic failure of reason, simply ignorance. Someone not familiar with electronics (my mother for example) could easily make that assumption, not out of stupidity, but only ignorance.

 

[gwag]

guess no one has seen a G5 heatsink

 

[pm]

How many people work at intel? How many are in a position to know the specific details behind this new processor? I'd like to see some white pages before I accept any claims as sacrosanct.

As I mentioned earlier, I will be presenting details on Montecito at this spring's International Solid State Circuits Conference. That's the earliest any papers on Montecito will be available.

 

[Lithan]

So are you saying that the earlier Inquirer article is correct? The processor will consume a maximum of 100w, and maintain this as well as temperatures by throttling? Is it safe to assume that the processor will be rated and sold according to it's performance at full speed (without having reached the 100w threshold)? If I'm reading the article correctly, this isn't your typical p4 thermal throttling. This is throttling inside of normal operating conditions when the processor is performing power-hungry tasks, or in essence, the processor (or all aspects of it) will only operate at its (their) full potential given that certain variables (set in processor design, not as an aftereffect of) are met.

 

[pm]

Originally posted by: Lithan
So are you saying that the earlier Inquirer article is correct? The processor will consume a maximum of 100w, and maintain this as well as temperatures by throttling? Is it safe to assume that the processor will be rated and sold according to it's performance at full speed (without having reached the 100w threshold)? If I'm reading the article correctly, this isn't your typical p4 thermal throttling. This is throttling inside of normal operating conditions when the processor is performing power-hungry tasks, or in essence, the processor (or all aspects of it) will only operate at its (their) full potential given that certain variables (set in processor design, not as an aftereffect of) are met.

I only meant to say two things:
1. that the original link points to a heatsink that is completely unfamiliar to me and appears to be much bigger than anything that we are using in our test set ups.
2. that the article at the top that mentions that Montecito dissipates 150W-200W and that this is incorrect and that Montecito is substantially lower power that this.

Foxton technology is an interesting method for maximizing performance while maintaining a given power limit. It appears at first glance to be throttling - I admit that when it was first explained to me, I thought this and spent an embarrassingly long while missing the point on it. It's unique to Montecito and more interesting than it appears at first glance. The full implementation details will be disclosed early next year and it's not my place to discuss it further right now - as much as I'd like to.

 

[IntelUser2000]

"Most stupid in a day or two for me. If you really look around anand you can find some really stupid stuff fairly easily. Someone gets their 2.4c running 3ghz and starts thinking they are qualified to be an electrical engineer. The reason I don't consider it the most stupid is that it isn't exampling a basic failure of reason, simply ignorance. Someone not familiar with electronics (my mother for example) could easily make that assumption, not out of stupidity, but only ignorance."

Hmm, sorry to piss you off with my ignorance but that doesn't have to do with my electronics knowledge. I know electronics, but just basics. From hearing from people I know and stuff, I came to a conclusion that Power supplies have two states: Idle and Max power. I guess its not true. Maybe you can tell me?

"How do you figure a processor throttles down without impacting performance? You do know what the term "throttling down" means?"

I am guessing it means "clocking the processor down." Okay, I may not know the proper words for explaining Foxton Technology and its related technology, but I do know one of the major factors that makes Montecito have TDP of 100W despite all the enhancements is Foxton Technology.

The massive heatsink that was shown on Inquirer was thoroughly dissected and analyzed at www.realworldtech.com. The guys at the site said 150-200W is the TEC's cooling AND Montecito's power dissipation requirements TOGETHER, since TEC needs to cool the processor and itself.

Anyway don't trust the Inquirer, take it with a truckload of salt. I saw couple of articles that gave misleading information, one example being how Montecito will use Blackford chipset. Blackford chipset is for Xeons, and platform merger of Xeons and Itanium is supposed to happen only in 2007 with Tukwila, and Blackford is supposed to come next year!!!


"The 2005 ISSCC advance program1 offers quite a few interesting details about Intels upcoming dual-core Itanium processor, Montecito. Here's a breakdown of the papers and the information which has been disclosed:"

Go to link here: http://www.realworldtech.com/f...ID=42339&roomID=11

"With both cores operating at full speed, the chip consumes 100W."

Anyway, 100W is what Intel is trying to achieve. There is no TDP in this case, its just 100W max at any given time.