Anyone notice the very poor thermal properties of the new 2.8 Ghz Intel chip? 75 Watt heat disapation. Ouch! Loud and hot.
Poor thermal properties of 2.8 Intel chip?
- Thread starter KKiller
- Start date
loud and hot?
Yet it uses the same heatsink/fan combination that hte other p4's use, so it is not loud. Nor is it that hot, since overclocking several hundred mhz is possible with the retail heatsink.
And at teh same token, what about the xp2600+ heat output?
Mike
Yet it uses the same heatsink/fan combination that hte other p4's use, so it is not loud. Nor is it that hot, since overclocking several hundred mhz is possible with the retail heatsink.
And at teh same token, what about the xp2600+ heat output?
Mike
Yeh...how about that 3.3ghz overclock with air cooling alone!!! According to THG nothing in amd's near future will hang with it, even the 3400+ (2.66ghz oc'd from 2.133ghz). Plus that chip needed quite the exotic cooling.
IN most instances the p4 overclocks have rarely been limited by heat...Usually the vcore will stop you first before the heat and throttling will limit you.
The higher jump can possibly be attributed to the new default vcore of 1.525v.
IN most instances the p4 overclocks have rarely been limited by heat...Usually the vcore will stop you first before the heat and throttling will limit you.
The higher jump can possibly be attributed to the new default vcore of 1.525v.
i believe that 2600+ has a lower wattage rating than the 2.8 Ghz. I THINK its 68 watts.. not sure though.
Intels stock fan is heat sensing.. meaning it will speed up (read louder!!)... to cool off the chip.
wonder how long it will take for the Intel fan boys to start coming in here with their 6 inch lead pipes? 
I guess I don't really understand the point of this thread...
High performance cpu's are going to put out a lot of wattage. That's just basic microelectronics 101. However, the title implies that the P4-2.8ghz is worse than other chips that perform that high... But it's really not.
But not only that, it dissapates the heat extremely well. Imho, it's very impressive that a 2.8ghz can run so (relatively) cool. The temps of a P4-2.8ghz are less than that of other .13µ cpu's that run at 1000mhz less.
So again... I don't really get the point, unless KKiller was just trying to rile up the "Intel fan boys". And with his last post, I believe that is the case.
High performance cpu's are going to put out a lot of wattage. That's just basic microelectronics 101. However, the title implies that the P4-2.8ghz is worse than other chips that perform that high... But it's really not.
But not only that, it dissapates the heat extremely well. Imho, it's very impressive that a 2.8ghz can run so (relatively) cool. The temps of a P4-2.8ghz are less than that of other .13µ cpu's that run at 1000mhz less.
So again... I don't really get the point, unless KKiller was just trying to rile up the "Intel fan boys". And with his last post, I believe that is the case.
No it's not,the fan itself does not regulate itself,some boards do that but the stock intel fan doesnt have a sensor to regulate speed.
I was actually wondering about this myself. They got a 2.8 to 3.3 thats great but only 500MHz. I see people around here with 2.0's at 2.8 and 1.6's at 2.7
I thought it was ~68W 'typical' thermal dissipation, not 75W. Where did you get that from?
And riptor, heard of core headroom?
An AGOIA 1.4GHz 1600+ can hit 1.8GHz pretty easy. An AGOIA 1.73GHz 2100+ won't hit 2.2GHz though... see what I'm getting at?
The lower speed variants of a core always overclock better percentage wise than the faster chips.
And riptor, heard of core headroom?
An AGOIA 1.4GHz 1600+ can hit 1.8GHz pretty easy. An AGOIA 1.73GHz 2100+ won't hit 2.2GHz though... see what I'm getting at?
The lower speed variants of a core always overclock better percentage wise than the faster chips.
KKiller, the 2.8GHz Pentium 4 has a thermal design power of 68W, not 75W. Page 63 of the datasheet here. And, as previously mentioned, it uses the same heatsink/fan combination as the other Pentium 4's, it's not any noiser.
68W does not seem too excessive to me in light of other CPU's on the market today. Looking at workstation and server CPU's - which is not completely unfair based on the tmpC performance of Pentium 4 CPU's - many of the high-end CPU's are in the 100W+ range. Besides, to put it in perspective, look at the power dissipation numbers for a typical large monitor - like a 19" or 21" monitor. 170W or more is not unusual. I've heard of 21" monitors in the 200W+ range.
FWIW, chips are getting hotter and this trend will continue up until we reach 100-120W - at which point either they will develop a better method of cooling the chips, or they will have pretty hit the cost limit for H/S + fan in the home/office computer market. If you think they are hot now, they will only be getter hotter as we move down into the tends of nanometers.
68W does not seem too excessive to me in light of other CPU's on the market today. Looking at workstation and server CPU's - which is not completely unfair based on the tmpC performance of Pentium 4 CPU's - many of the high-end CPU's are in the 100W+ range. Besides, to put it in perspective, look at the power dissipation numbers for a typical large monitor - like a 19" or 21" monitor. 170W or more is not unusual. I've heard of 21" monitors in the 200W+ range.
FWIW, chips are getting hotter and this trend will continue up until we reach 100-120W - at which point either they will develop a better method of cooling the chips, or they will have pretty hit the cost limit for H/S + fan in the home/office computer market. If you think they are hot now, they will only be getter hotter as we move down into the tends of nanometers.
KKiller, Intel Itanum's max. power dissipation was over 130W at 800Mhz, 75W@2.8Ghz seems rather low compared to that
Yes. It appears that 68 W is "average" power dissipation (sp?). However, I think most ATOTs run their CPUs harder than "average", with distributed computing programs, photoshop, etc...
I read that 75 watts was max power usage... I can't find the source... perhaps some one here will find it.
I read that 75 watts was max power usage... I can't find the source... perhaps some one here will find it.
Actually, it's not really the "average" power - it's the design target power. As in "design to this power dissipation level and things will work fine". It's not "average" or "typical" power - that is much lower.
There has been a lot of debate over Intel's power dissipation numbers online - although I don't hear a lot of engineers getting upset over the numbers. The question is: is max power a useful number to design a processor for? Maximum power is the theoretical maximum power that the design can ever hit - is this the target that you want to use as a design target for a cooling system?
Maximum power isn't anything you are ever going to see - in fact when I worked on the Pentium we spent a good deal of time trying to write a program that actually hit something close to our maximum power number and never got very close at all. Maximum power is calculated by assuming that essentially every possible operation that could theoretically occur concurrently, does in fact occur concurrently. But in reality this doesn't happen. In my experience, maximum power is an unreasonably high target that leads to high system design costs and overspecification of system components. Intel has historically quoted "max power" for designs but changed this practice with the Pentium 4 for various reasons - not the least of which is design cost.
What an engineer really wants, in my opinion, is "what is the maximum power that a user will ever encounter". This is, in my opinion, what "max power" really should be. But this is a hard number to calculate because it you never know if a future program will somehow find a way to use more power than your worst-case program now. It's a lot easier to just take the simulation model of the chip, make the worst case bunch of events happen simultaneously in the model by forcing various parameters and then quote this number as the "maximum power". Trying to figure out the realistic highest power output that a user is likely to see involves the risk that you will somehow underestimate this number and then some application will show up that will cause all of the systems to overheat and crash because you underspec'd the worst case power. So there is a risk aspect to this calculation. But I have yet to meet an engineer who is uncomfortable with the idea of "design target" power instead of "peak theoretical maximum power". It just doesn't make sense to waste money overspec'ing a system to enable operation within a power envelope that will never actually be reached in practice. IMO.
But back onto your original point, KKiller: Thermal Design Power, TDP - the 68W number that I, and several others, have quoted for the 2.8GHz Pentium 4 is defined by Intel as: "A power dissipation target based on worst case applications. Thermal solutions should be designed to dissipate the thermal design power." (Thermal design application note, page 2). So the worst case application set is taken into account with this number, and adequate margin exists to ensure that future applications should not exceed this number as well. Photoshop and distributed computing are already considered in the number's calculation.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 22K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
News NVIDIA and Intel to Develop AI Infrastructure and Personal Computing Products- Started by poke01
- Replies: 407
-
Facebook
Twitter