Wall Street Journal figures out "emergency edition" P4...

[Wingznut]

Originally posted by: zephyrprime

isn't it quite possible they'll lose money compared to what they could profit if they sold them as full fledged 2MB Xeons?

That's possible. Economist would call that the oportunity costs. Say I found a gold nugget on the street and sold it to you for $1.00. I'd be making money but I'd totally losing out in opportunity costs since I could have sold the nugget for much more.

Potential Profit is how I'd define it. 😉

But not really... It's not like anyone out there will be choosing a P4-EE instead of a Xeon, as they are very different markets. I guess you could call it losing potential profit if they were sacrificing some Xeon production (and not meeting demand) for the P4-EE.

But, contrary to what many people here might think, Intel isn't dumb. 🙂

 

[Davegod]

"I guess you could call it losing potential profit if they were sacrificing some Xeon production (and not meeting demand) for the P4-EE"

Yeah but "potential profit" is actually accounted for, reducing it ends up same as a literal cost 😉 Not just directl;y liek that, but the fixed/sunk costs are accounted for over long term expected revenues/margins, and sudden hits to these result in the amortisation having to be written off (i.e. entered as a literal $cost into the accounts) as soon as this reduction becomes known.

Further, Xeon customers are liable to get it into their heads that if they can sell what is basically an Xeon for 1/3 of the price, then WTF are we paying 3x as much? High end high margin product gets massively devalued, the cash cow gets foot and mouth disease.

Basically anything which results in a reduction of profits, past current and future, can easily be called "losing money" 🙂

 

[Wingznut]

Originally posted by: Davegod
Basically anything which results in a reduction of profits, past current and future, can easily be called "losing money" 🙂

Ok, well how much does it cost to have the public perceive that the competition is way ahead?

It's just not that black and white.

 

[dexvx]

Originally posted by: ElFenix
i wonder what intel's variable cost per cpu is? probably $5

One of my neighbors in Oregon is an Intel Finance person for Prescott. The Northwoods he helped managed beforehand cost about $40-50 per CPU in physical costs.

 

[wetcat007]

Originally posted by: Wingznut
I'm not sure why they think Intel would be losing money on these.

Obviously they are a larger die.... Let's say that they do retail for $750, and let's say that the die is exactly twice the size (hypothetically, of course) of a Northwood. Are they suggesting that Intel loses money on a Northwood that retails for $375? (Not to mention all the cpu's that retail for less than that.

The die is friggin huge... they produce 137 watts of heat... just plain impractical..

 

[Wingznut]

Originally posted by: wetcat007

Originally posted by: Wingznut
I'm not sure why they think Intel would be losing money on these.

Obviously they are a larger die.... Let's say that they do retail for $750, and let's say that the die is exactly twice the size (hypothetically, of course) of a Northwood. Are they suggesting that Intel loses money on a Northwood that retails for $375? (Not to mention all the cpu's that retail for less than that.

The die is friggin huge... they produce 137 watts of heat... just plain impractical..

The die is smaller than Hammer, 137w is an exagerration, and what's any of that got to do with the the topic at hand anyway???

If you have nothing constructive to add... Go away.

 

[Markfw]

The die is smaller than Hammer, 137w is an exagerration, and what's any of that got to do with the the topic at hand anyway???

I think talking about the chip and if it's practicality qualifies for this thread. And I read 117 watts for the P4EE. WingZ, Do you have any REAL official watts for that CPU ? The only thing I read for a review site was that it was "figging hot, a real space heater" or something to that effect.

 

[Wingznut]

Originally posted by: Markfw900
WingZ, Do you have any REAL official watts for that CPU ?

Obviously, I cannot talk about specifics when it comes to future products, so I cannot answer you. Sorry.

 

[Markfw]

Obviously, I cannot talk about specifics when it comes to future products, so I cannot answer you. Sorry.

Ok, so it is official, it is a paper launch, and maybe not even a real product (until you can actually buy one). Again, this is pertinent to the thread, as no matter what reasons they had for doing this, it is paper and it still didn't even beat the Athlon FX by many peoples reviews. THIS is what wall street probably figured out.

Edit: And you disclaimed the 137 watts as an exageration, but have no comment on 117 watts. I think that speaks volumes.

BTW, I have much respect for you on these forums WingZ, but this last "stunt" by your employer just stinks, and I don't agree with trying to defend it.

 

[NFS4]

Originally posted by: Wingznut

Originally posted by: wetcat007

Originally posted by: Wingznut
I'm not sure why they think Intel would be losing money on these.

Obviously they are a larger die.... Let's say that they do retail for $750, and let's say that the die is exactly twice the size (hypothetically, of course) of a Northwood. Are they suggesting that Intel loses money on a Northwood that retails for $375? (Not to mention all the cpu's that retail for less than that.

The die is friggin huge... they produce 137 watts of heat... just plain impractical..

The die is smaller than Hammer, 137w is an exagerration, and what's any of that got to do with the the topic at hand anyway???

If you have nothing constructive to add... Go away.

Athlon 64 3200+ / Athlon 64 FX-51: 193mm^2 die size
Pentium 4 Extreme Edition: 237 mm^2 die size

😕😕😕

 

[Wingznut]

Originally posted by: Markfw900

Obviously, I cannot talk about specifics when it comes to future products, so I cannot answer you. Sorry.

Ok, so it is official, it is a paper launch, and maybe not even a real product (until you can actually buy one). Again, this is pertinent to the thread, as no matter what reasons they had for doing this, it is paper and it still didn't even beat the Athlon FX by many peoples reviews. THIS is what wall street probably figured out.

You know what... I owe you and wetcat an apology.

For some reason, I thought wetcat was talking about Prescott. Hence, my comment about the die being smaller than Hammer is most likely untrue. (I say "most likely" because I haven't looked to see what the die size of Hammer is.) So, feel free to ignore everything in that post.

And even still, I cannot comment about the P4-EE because (as far as I know), Intel hasn't officially launched it. There are no press releases about it, that I can recall.

Oh yeah, nothing I say makes ANYTHING "official", even if I wasn't in the middle of a brain cramp. 😛

 

[Wingznut]

Originally posted by: NFS4

Originally posted by: Wingznut

Originally posted by: wetcat007

Originally posted by: Wingznut
I'm not sure why they think Intel would be losing money on these.

Obviously they are a larger die.... Let's say that they do retail for $750, and let's say that the die is exactly twice the size (hypothetically, of course) of a Northwood. Are they suggesting that Intel loses money on a Northwood that retails for $375? (Not to mention all the cpu's that retail for less than that.

The die is friggin huge... they produce 137 watts of heat... just plain impractical..

The die is smaller than Hammer, 137w is an exagerration, and what's any of that got to do with the the topic at hand anyway???

If you have nothing constructive to add... Go away.

Athlon 64 3200+ / Athlon 64 FX-51: 193mm^2 die size
Pentium 4 Extreme Edition: 237 mm^2 die size

😕😕😕

Yeah, yeah, yeah... Boy, do I look like a jackass. I have no idea why, but I honestly thought he was talking about PSC... Which is why I said that his post had nothing to do with the topic at hand.

Sorry guys. :frown:

 

[Markfw]

Apology accepted. Notice my edit comment on the bottom of my last post. I really did think that was beyond you. And I only said "official" since it appeared to be a confirmation when not denied, and at least by this forums rules, they even say official for the inquirer, and you are WAY better than them.

 

[Sohcan]

Originally posted by: Markfw900

The die is smaller than Hammer, 137w is an exagerration, and what's any of that got to do with the the topic at hand anyway???

I think talking about the chip and if it's practicality qualifies for this thread. And I read 117 watts for the P4EE. WingZ, Do you have any REAL official watts for that CPU ? The only thing I read for a review site was that it was "figging hot, a real space heater" or something to that effect.

The 2.8 GHz Xeon MP w/2 MB L3 has a max power of 83 Watts and a thermal design power (what cooling solutions need to meet) of 72 watts. Large, higher levels of cache are designed to consume little power, and usually contribute no more than 5-10% of the total power.

The 2.8 GHz / 533 MT/s bus P4 has a TDP of 68.4 Watts, so the L3 seems to add 3.6 W. The 3.2 GHz P4 has a TDP of 82 W, so around 86 W is likely for the P4 EE.

Page 71

 

[NFS4]

Originally posted by: Sohcan

Originally posted by: Markfw900

The die is smaller than Hammer, 137w is an exagerration, and what's any of that got to do with the the topic at hand anyway???

I think talking about the chip and if it's practicality qualifies for this thread. And I read 117 watts for the P4EE. WingZ, Do you have any REAL official watts for that CPU ? The only thing I read for a review site was that it was "figging hot, a real space heater" or something to that effect.

The 2.8 GHz Xeon MP w/2 MB L3 has a max power of 83 Watts and a thermal design power (what cooling solutions need to meet) of 72 watts. Large, higher levels of cache are designed to consume little power, and usually contribute no more than 5-10% of the total power.

Page 71

Well, you'll just have to guesstimate for a 3.2GHz version😉

 

[Markfw]

Originally posted by: Sohcan

Originally posted by: Markfw900

The die is smaller than Hammer, 137w is an exagerration, and what's any of that got to do with the the topic at hand anyway???

I think talking about the chip and if it's practicality qualifies for this thread. And I read 117 watts for the P4EE. WingZ, Do you have any REAL official watts for that CPU ? The only thing I read for a review site was that it was "figging hot, a real space heater" or something to that effect.

The 2.8 GHz Xeon MP w/2 MB L3 has a max power of 83 Watts and a thermal design power (what cooling solutions need to meet) of 72 watts. Large, higher levels of cache are designed to consume little power, and usually contribute no more than 5-10% of the total power.

Page 71

Yes, but we are talking 3.2 ghz, NOT 2.8. There may be a great deal of difference. Do you have ANY even remotely qualified argument to dispute the 117 watts I read about. If you want me to find that, I might spend another 6 hours reading to find it, and I may do that. I KNOW I can find that "space heater comment" if you didn't read that yourself in the 15 reviews posted. And I don't dispute that 117 watts since it is well rumored the Prescot is at 103 watts, and that is rumored not as fast as the P4EE. ( I know, rumor,rumor..... "Just give me the facts mam")

 

[Sohcan]

Originally posted by: Markfw900

Originally posted by: Sohcan

Originally posted by: Markfw900

The die is smaller than Hammer, 137w is an exagerration, and what's any of that got to do with the the topic at hand anyway???

I think talking about the chip and if it's practicality qualifies for this thread. And I read 117 watts for the P4EE. WingZ, Do you have any REAL official watts for that CPU ? The only thing I read for a review site was that it was "figging hot, a real space heater" or something to that effect.

The 2.8 GHz Xeon MP w/2 MB L3 has a max power of 83 Watts and a thermal design power (what cooling solutions need to meet) of 72 watts. Large, higher levels of cache are designed to consume little power, and usually contribute no more than 5-10% of the total power.

Page 71

Yes, but we are talking 3.2 ghz, NOT 2.8. There may be a great deal of difference. Do you have ANY even remotely qualified argument to dispute the 117 watts I read about.

Umm...I'm a microprocessor engineer? 😛

Reread my previous post after the edit. Dynamic power is proportional to clock rate, while static power (a major contribution to power consumption these days) is proportional to voltage * leakage. Had the P4 EE had a 400 MHz bus like the Xeon MP, power would have most certainly scaled less than the proportional increase in clock rate...the faster bus adds about 5 watts. Between 85-90 W is likely for the TDP, 117 W is out of the question.

There's no reason to be belligerent, I'm just trying to help out.

 

[MadRat]

The L3 cache doesn't necessarily have to be on the exact same die as the rest of the cpu. A second core solely of L3 cache would make it less of a gamble to produce these "EE" chips. Intel has been working on packaging for the past year to do multiple components inside of a single processor package, called BBUL. The trace length from core to cache should easily be short enough to make it possible.

 

[Goose77]

Originally posted by: Sohcan

Originally posted by: Markfw900

Originally posted by: Sohcan

Originally posted by: Markfw900

The die is smaller than Hammer, 137w is an exagerration, and what's any of that got to do with the the topic at hand anyway???

I think talking about the chip and if it's practicality qualifies for this thread. And I read 117 watts for the P4EE. WingZ, Do you have any REAL official watts for that CPU ? The only thing I read for a review site was that it was "figging hot, a real space heater" or something to that effect.

The 2.8 GHz Xeon MP w/2 MB L3 has a max power of 83 Watts and a thermal design power (what cooling solutions need to meet) of 72 watts. Large, higher levels of cache are designed to consume little power, and usually contribute no more than 5-10% of the total power.

Page 71

Yes, but we are talking 3.2 ghz, NOT 2.8. There may be a great deal of difference. Do you have ANY even remotely qualified argument to dispute the 117 watts I read about.

Umm...I'm a microprocessor engineer? 😛

Reread my previous post after the edit. Dynamic power is proportional to clock rate, while static power (a major contribution to power consumption these days) is proportional to voltage * leakage. Had the P4 EE had a 400 MHz bus like the Xeon MP, power would have most certainly scaled less than the proportional increase in clock rate...the faster bus adds about 5 watts. Between 85-90 W is likely for the TDP, 117 W is out of the question.

There's no reason to be belligerent, I'm just trying to help out.

Your right about alot of that, but there still is a lot of variables to contend with. Unless ur an engineer for intel workin on these processor you really cant gaurentee any of the numbers you put up. 117 W is not out of the question. Is there a 3.2 Xeon proc to guesstimate with. My guess, and i mean guess, is no. There might be a reason for no Xeon at higher speeds because of some sort of problem (i.e. prob with heat, might be too hot for a server). Remember, architecture varies from core to core.

 

[jbond04]

Originally posted by: Goose77

Unless ur an engineer for intel workin on these processor you really cant gaurentee any of the numbers you put up. 117 W is not out of the question.

That's BS...I work for HP, and I know a lot about products and technologies I don't work with on a day to day basis. It's kind of humorous, actually, what you can learn from bathroom and hallway talk. Just because Sohcan may not work on the Xeon MP or P4 EE doesn't mean he hasn't heard or read about something about it. There are tons of things I encounter each day that are labeled 'HP Confidential', which means that people like you don't hear about it. I have access to a database that lists reports on just about any technology that HP has ever developed; things that the public has never heard or dreamed of (and probably never will). I would trust the word of Intel's employees when they say that the consumption is less than the rumored 117W...just because they can't quote you a specific number (it may be 'Intel Confidential'), doesn't mean that they're full of it.

 

[Goose77]

Originally posted by: jbond04

Originally posted by: Goose77

Unless ur an engineer for intel workin on these processor you really cant gaurentee any of the numbers you put up. 117 W is not out of the question.

That's BS...I work for HP, and I know a lot about products and technologies I don't work with on a day to day basis. It's kind of humorous, actually, what you can learn from bathroom and hallway talk. Just because Sohcan may not work on the Xeon MP or P4 EE doesn't mean he hasn't heard or read about something about it. There are tons of things I encounter each day that are labeled 'HP Confidential', which means that people like you don't hear about it. I have access to a database that lists reports on just about any technology that HP has ever developed; things that the public has never heard or dreamed of (and probably never will). I would trust the word of Intel's employees when they say that the consumption is less than the rumored 117W...just because they can't quote you a specific number (it may be 'Intel Confidential'), doesn't mean that they're full of it.

Ok, unless he said that he worked for intel in a previous post, he still cant know the exact Wattage! Even if he did hear something on the grapevine, that doesnt mean its true! What he hear could have been from an engineering sample. Until intel says "this is the watts" its all Skepticism. Look, im not trying to argue with any of you, but i dont like someone telling me what the facts are when there is no solid evidence to back up what he says. I can take what he says as his guess, but not as fact. Until intel says what it is, its all an estimate, and even then they could be skewing the fact. Take for example the NV fx fiasco!! man did they skew fact!! My point is that all these numbers are nice guesses!

 

[Sohcan]

Originally posted by: Goose77

Umm...I'm a microprocessor engineer? 😛

Reread my previous post after the edit. Dynamic power is proportional to clock rate, while static power (a major contribution to power consumption these days) is proportional to voltage * leakage. Had the P4 EE had a 400 MHz bus like the Xeon MP, power would have most certainly scaled less than the proportional increase in clock rate...the faster bus adds about 5 watts. Between 85-90 W is likely for the TDP, 117 W is out of the question.

There's no reason to be belligerent, I'm just trying to help out.

Your right about alot of that, but there still is a lot of variables to contend with. Unless ur an engineer for intel workin on these processor you really cant gaurentee any of the numbers you put up. 117 W is not out of the question. Is there a 3.2 Xeon proc to guesstimate with. My guess, and i mean guess, is no. There might be a reason for no Xeon at higher speeds because of some sort of problem (i.e. prob with heat, might be too hot for a server). Remember, architecture varies from core to core.

Yes, I do work for Intel; and no, I don't need to have worked on this particular processor to know that 117 watts TDP is out of the question.

You don't need me telling you this, there are more than enough data points from the publically available datasheets for the P4 and the Xeon MP to estimate power consumption.

Until intel says "this is the watts" its all Skepticism. Look, im not trying to argue with any of you, but i dont like someone telling me what the facts are when there is no solid evidence to back up what he says. I can take what he says as his guess, but not as fact.

I'm not about to start posting information I might have from confidential meetings and documents. I linked the Xeon MP datasheet (here's the P4 datasheet), there's more than enough public info you need.

Until intel says what it is, its all an estimate, and even then they could be skewing the fact. Take for example the NV fx fiasco!! man did they skew fact!! My point is that all these numbers are nice guesses!

Intel can't lie about power consumption, vendors rely on the datasheets to supply enough power to the processor and provide adequate cooling. If the numbers were intentionally lowered, systems would fail.

 

[orion7144]

Originally posted by: dexvx

Originally posted by: ElFenix
i wonder what intel's variable cost per cpu is? probably $5

One of my neighbors in Oregon is an Intel Finance person for Prescott. The Northwoods he helped managed beforehand cost about $40-50 per CPU in physical costs.

Lets just say that is a high average and those cost include everything including pay, benefits, tool cost (including depreciation), design, and so on. The reason that Intel's die costs are so low is their die yield. That is why AMD has so much trouble making money. If they could get their die yield to an acceptable level they would be making some good money.

 

[T9D]

..

 

[Markfw]

This is such a sad smear campagne, Seriously this is all the AMD supporters could come up with?

Smear campaign ? Intel comes out at the last minute with a CPU out of the lab which will probably never be produced that can heat your house, and still can't beat the best AMD is SHIPPING TODAY, and you call this a smear campange ? Go back to your Intel fanboi's and try to make a logical argument. This argument takes two flushes to make it to the kitchen......