Yonah to suck up more power.

[Googer]

Yonah, a new notebook chip coming from Intel early next year, will run slightly faster than expected, but may also consume more power than its contemporaries.

 

[Hacp]

I think thats when both cores are on and running. I beleive that Yonah will shut down one of its cores when its not in use, so the actual power consumption should be similar to A regular Pentium M model when used unplugged. Extra power consumption will most likely happen when your plugged in and doing more intensive tasks.

 

[Googer]

Originally posted by: Hacp
I think thats when both cores are on and running. I beleive that Yonah will shut down one of its cores when its not in use, so the actual power consumption should be similar to A regular Pentium M model when used unplugged. Extra power consumption will most likely happen when your plugged in and doing more intensive tasks.

But still it may not be as good on power conservation as its predecessor.

 

[Hacp]

Originally posted by: Googer

Originally posted by: Hacp
I think thats when both cores are on and running. I beleive that Yonah will shut down one of its cores when its not in use, so the actual power consumption should be similar to A regular Pentium M model when used unplugged. Extra power consumption will most likely happen when your plugged in and doing more intensive tasks.

But still it may not be as good on power conservation as its predecessor.

Why wouldn't it when it has a die shrink? The only thing that is added is the extra core, and that is likely not likely going to be activated alot when your unplugged. I forsee very similar, maybe even a little better battery life with the new yonah processors.

 

[Googer]

Originally posted by: Hacp

Originally posted by: Googer

Originally posted by: Hacp
I think thats when both cores are on and running. I beleive that Yonah will shut down one of its cores when its not in use, so the actual power consumption should be similar to A regular Pentium M model when used unplugged. Extra power consumption will most likely happen when your plugged in and doing more intensive tasks.

But still it may not be as good on power conservation as its predecessor.

Why wouldn't it when it has a die shrink? The only thing that is added is the extra core, and that is likely not likely going to be activated alot when your unplugged. I forsee very similar, maybe even a little better battery life with the new yonah processors.

A die shrink does not always mean lower power consumption.

130nm Pentium Northwood V.S. 90nm Prescott. Need I say more?

 

[alimoalem]

what's wrong with high power consumption? is the only negative shorter battery life?

 

[Googer]

Originally posted by: alimoalem
what's wrong with high power consumption? is the only negative shorter battery life?

Higher energy bills, larger bulkier notebooks (no thin and light), harder to keep cool, enviromental polution, and almost no run time- sort of defeats the purpose of a laptop. I left one out, I forgot.

 

[Sunner]

Originally posted by: Googer

Originally posted by: Hacp

Originally posted by: Googer

Originally posted by: Hacp
I think thats when both cores are on and running. I beleive that Yonah will shut down one of its cores when its not in use, so the actual power consumption should be similar to A regular Pentium M model when used unplugged. Extra power consumption will most likely happen when your plugged in and doing more intensive tasks.

But still it may not be as good on power conservation as its predecessor.

Why wouldn't it when it has a die shrink? The only thing that is added is the extra core, and that is likely not likely going to be activated alot when your unplugged. I forsee very similar, maybe even a little better battery life with the new yonah processors.

A die shrink does not always mean lower power consumption.

130nm Pentium Northwood V.S. 90nm Prescott. Need I say more?

There were many siginificant differences between Northwood and Prescoot, far more than just a die shrink.

 

[First]

Yonah will not consume that much more power drawing straight from the battery, and the notebooks won't be any bigger.

 

[hurtstotalktoyou]

Whoa, guys. This is all just speculation. We don't know how much power Yonah will actually consume in real-world applications. Until then, I don't see the point in whining.

 

[Googer]

Originally posted by: Sunner

Originally posted by: Googer

Originally posted by: Hacp

Originally posted by: Googer

Originally posted by: Hacp
I think thats when both cores are on and running. I beleive that Yonah will shut down one of its cores when its not in use, so the actual power consumption should be similar to A regular Pentium M model when used unplugged. Extra power consumption will most likely happen when your plugged in and doing more intensive tasks.

But still it may not be as good on power conservation as its predecessor.

Why wouldn't it when it has a die shrink? The only thing that is added is the extra core, and that is likely not likely going to be activated alot when your unplugged. I forsee very similar, maybe even a little better battery life with the new yonah processors.

A die shrink does not always mean lower power consumption.

130nm Pentium Northwood V.S. 90nm Prescott. Need I say more?

There were many siginificant differences between Northwood and Prescoot, far more than just a die shrink.

He was sayig that a die shirnk was going to be the solve all cure for power consumption. In reality a die shink has often lowered power consumption from about 0-15% depending on the achitecuture and if the chip was at all changed or altered in the process as in the prescott case then it could increase. But going from 25 to 49wattts even with a die shink is not a good thing.

 

[Jeff7181]

I think this was posted in the CPu forums too. So I'll repeat here what I said there...

Surprise! Two cores consume more power than one!

 

[Sunner]

Originally posted by: Googer

Originally posted by: Sunner
There were many siginificant differences between Northwood and Prescoot, far more than just a die shrink.

He was sayig that a die shirnk was going to be the solve all cure for power consumption. In reality a die shink has often lowered power consumption from about 0-15% depending on the achitecuture and if the chip was at all changed or altered in the process as in the prescott case then it could increase. But going from 25 to 49wattts even with a die shink is not a good thing.

Not saying a die shrink will always bring x% power reduction, just saying that Northwood->Prescott isn't a good example since there are vast architectural differences between the two.

 

[Googer]

Originally posted by: Sunner

Originally posted by: Googer

Originally posted by: Sunner
There were many siginificant differences between Northwood and Prescoot, far more than just a die shrink.

He was sayig that a die shirnk was going to be the solve all cure for power consumption. In reality a die shink has often lowered power consumption from about 0-15% depending on the achitecuture and if the chip was at all changed or altered in the process as in the prescott case then it could increase. But going from 25 to 49wattts even with a die shink is not a good thing.

Not saying a die shrink will always bring x% power reduction, just saying that Northwood->Prescott isn't a good example since there are vast architectural differences between the two.

I would not call the differances between them vast, that is a bit of a streach. Vast differances would be comparing an AMD 64 to a Pentium 4 or comparing a Pentium 3.8 GHz to a 1.8GHz Itanium there is a big differance between those. The differance between Prescott and Northwood is not too great and their architecture is so close to being identical.

 

[LTC8K6]

I think all the extra pipeline stages and the greater cache and SSE3 are significant differences.

One of the reasons Prescott was disappointing was because all of the changes resulted in no real performance increase over Northwood and even a decrease many times.

 

[DaFinn]

Originally posted by: Googer

Originally posted by: Sunner

Originally posted by: Googer

Originally posted by: Sunner
There were many siginificant differences between Northwood and Prescoot, far more than just a die shrink.

He was sayig that a die shirnk was going to be the solve all cure for power consumption. In reality a die shink has often lowered power consumption from about 0-15% depending on the achitecuture and if the chip was at all changed or altered in the process as in the prescott case then it could increase. But going from 25 to 49wattts even with a die shink is not a good thing.

Not saying a die shrink will always bring x% power reduction, just saying that Northwood->Prescott isn't a good example since there are vast architectural differences between the two.

I would not call the differances between them vast, that is a bit of a streach. Vast differances would be comparing an AMD 64 to a Pentium 4 or comparing a Pentium 3.8 GHz to a 1.8GHz Itanium there is a big differance between those. The differance between Prescott and Northwood is not too great and their architecture is so close to being identical.

I thinks biggest difference in power consumption was due to the 1Mb cache vs 512kb in Northwood. Also new pipelines, sse3 etc....

 

[Sunner]

Originally posted by: LTC8K6
I think all the extra pipeline stages and the greater cache and SSE3 are significant differences.

One of the reasons Prescott was disappointing was because all of the changes resulted in no real performance increase over Northwood and even a decrease many times.

Yep, extending the pipeline significally is definitely a big change, and they made some pretty big changes to the ALUs as well, doubled the trace cache, and a bunch of "other stuff".

DaFinn, the L2 doesn't get very hot by the way.

 

[FishTankX]

Did the Pentium-M's power consuimption go up, even though it doubled it's cache and increased it's clockspeed, from .13 to .09? No.

 

[Pariah]

The Pentium-M can disable parts of its L2 cache to reduce power consumption so it's not a valid comparison.

From the article:

"Yonah chips, though, will carry higher maximum-power-consumption ratings than current Pentium Ms. Most likely, that's because most Yonahs will sport two processing cores, rather than the single core found in today's notebook chips."

OK, so 2 cores will use more power than one. There's some breaking news. Move along folks, nothing to see here.

 

[Accord99]

Originally posted by: FishTankX
Did the Pentium-M's power consuimption go up, even though it doubled it's cache and increased it's clockspeed, from .13 to .09? No.

Yeah, it's specified TDP went down from 24.5 for the Banias family which maxxed out at 1.7GHz to 21W for the 400MHz FSB Dothans, which maxxed out at 2.1GHz.

Also, 2CPU had a look at a dual dualcore Sossaman system:

http://www.2cpu.com/review.php?id=109&page=5

94W power usage at the wall. Unknown clock speeds but it looks to be using 8 DDR2 DIMMs and 94W load usage compares favorably to many single-core systems at idle. Since Sossaman is the server variant of Yonah, I'd expect power usage for typical notebook CPUs to compare very well with Dothan.