Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

2.6 GHz worse battery life than 2.3 GHz?

M

macbook123

macrumors 68000
Original poster
Feb 11, 2006
1,869
85
I'm surprised that I haven't seen this being discussed on here (apologies if I missed it!): http://www.engadget.com/2012/06/13/apple-macbook-pro-with-retina-display-review/

The 20% improvement in battery life from 7:49 to 9:22 is pretty remarkable as it means that the 2.6 GHz processor is much less efficient than the 2.3 GHz one. 20% less efficient assuming both processors were made to complete the same tasks, and nearly 10% less efficient if the 10% faster processor was made to do 10% more tasks in a given time interval.
 
Last edited:
gentlefury said:
Gigahertz are a measurement of energy consumption. It's pretty obvious that a higher energy consuming chip would use more battery power.
Click to expand...

No, watts are a measurement of energy consumption, a hertz is a unit of frequency which describes the number of cycles per second of a periodic phenomenon.
 
Serously? Everyone here is so damn negative that I failed by stating a fact???

GHz has NOTHING to do with speed.....how many pulsing cycles of electricity the CPU can produce. This is how we now have low energy consumption cpu clocks that are MUCH more powerful than they used to be. They can utilize those cycles better and perform more flops per cycle. Less GHz does NOT equal slower speeds....

In computing, most central processing units (CPU) are labeled in terms of their clock rate expressed in megahertz or gigahertz (106 or 109 hertz, respectively). This number refers to the frequency of the CPU's master clock signal ("clock rate"). This signal is simply an electrical voltage which changes from low to high and back again at regular intervals. This signal is a square wave.
Click to expand...

But of course, the genius masses here are ALWAYS right.

How awesome is it to be 15 and SOOOO smart?
 
gentlefury said:
Serously? Everyone here is so damn negative that I failed by stating a fact???

GHz has NOTHING to do with speed.....how many pulsing cycles of electricity the CPU can produce. This is how we now have low energy consumption cpu clocks that are MUCH more powerful than they used to be. They can utilize those cycles better and perform more flops per cycle. Less GHz does NOT equal slower speeds....



But of course, the genius masses here are ALWAYS right.

How awesome is it to be 15 and SOOOO smart?
Click to expand...

Voltage isn't a measure of power consumption either, it's the potential energy between two points.

You're completely and utterly wrong about the CPU's GHz measurement having nothing to do with speed, it's everything to do with speed; it's a measurement (along with the instructions per clock) of how many instructions per second a processor can perform.

We're only being negative because you're passing off erroneous misinformation as though it were gospel.

Better to be 15 and well informed than old and refusing to learn or accept that one is wrong.
 
theBostonian said:
Voltage isn't a measure of power consumption either, it's the potential energy between two points.

You're completely and utterly wrong about the CPU's GHz measurement having nothing to do with speed, it's everything to do with speed; it's a measurement (along with the instructions per clock) of how many instructions per second a processor can perform.

We're only being negative because you're passing off erroneous misinformation as though it were gospel.

Better to be 15 and well informed than old and refusing to learn or accept that one is wrong.
Click to expand...

wow,I guess to not upset the trolls on this board you just have to agree with everyone.....ok, here goes.

Cool story bro. You're awesome! Everything you say is true!

Anyway, know-it-alls not withstanding. In response to the OP. More cycles per second requires more power consumption, therefore draining the battery faster. Low energy cpus that run at 1.8ghz are used in smaller units to get more battery life out of a smaller battery pack.
 
Isn't this always the case? I thought it was common knowledge that higher clocked processors almost always delivery worse battery performance. They generate less heat, fans don't need to spin as fast, etc. etc. I'd never buy a maxed out notebook CPU.
 
Dangerous Theory said:
I think he was referring to himself being 15 and self proclaiming his (non-existent) intelligence.
Click to expand...

I wasn't nearly as self centered as most of the users of this forum when I was 15. Tho when I was 15, if you wanted to be a dick you had to do it to someones face...since there weren't anonymous internet forums to do it in.
 
I'm voting every post in this thread 'UP' in a protest to them removing the vote 'DOWN' button.

Now how's that for intelligence? :D
 
wonder

wonder if we could come back to the initial topic? Are there other reports/test that confirm this information, that indeed the "slower" 2.3 processor will give a couple HOURS more battery life?
 
gentlefury said:
I wasn't nearly as self centered as most of the users of this forum when I was 15. Tho when I was 15, if you wanted to be a dick you had to do it to someones face...since there weren't anonymous internet forums to do it in.
Click to expand...

What did you mean by "How awesome is it to be 15 and SOOOO smart?" then?
 
Yeah, who the hell cares about some snot nosed kid and his ego. Let's get back on the topic here!

a 2 hour difference between the 2.3gh and 2.6gh makes no sense.

The only different component is the CPU and it doesn't account for the difference.

Did En gone and funked? Maybe not using the discrete GPU for the 2.3 test?

But getting 7 hours with the discrete on doesn't sound right either.
 
Meever said:
Yeah, who the hell cares about some snot nosed kid and his ego. Let's get back on the topic here!

a 2 hour difference between the 2.3gh and 2.6gh makes no sense.

The only different component is the CPU and it doesn't account for the difference.

Did En gone and funked? Maybe not using the discrete GPU for the 2.3 test?

But getting 7 hours with the discrete on doesn't sound right either.
Click to expand...

I agree, 2 hours does seem a bit extreme for such a small cycle difference.
 
gentlefury said:
Serously? Everyone here is so damn negative that I failed by stating a fact???

GHz has NOTHING to do with speed.....how many pulsing cycles of electricity the CPU can produce. This is how we now have low energy consumption cpu clocks that are MUCH more powerful than they used to be. They can utilize those cycles better and perform more flops per cycle. Less GHz does NOT equal slower speeds....



But of course, the genius masses here are ALWAYS right.

How awesome is it to be 15 and SOOOO smart?
Click to expand...


Way too many people just stating stuff with no expertise, knowledge, or cited sources on these forums.

The power consumed in chips is proportional to the operating frequency (i.e. the chip's clock) and the square of the operating voltage. (Source.)


The engineering tradeoff here is that, for higher frequencies, you want more voltage since you need to reduce the rise- and fall- times of the signal, but more voltage means much more power consumption. Alternatively, you could develop new/better semiconductor material device science techniques or improve manufacturing to reduce these times through in a material properties sense (or, better, reduce the required operating voltage since it's proportional to the square here). This is why ultra-low voltage (ULV) CPUs are so sought after for ultrabooks.


Anyway, clearly the CPU speed can't account for the roughly 20% power difference here. The GPU, screen, and other electronics make up a substantial proportion of the power consumed during operation so it would be like going from 2.3 GHz -> 2.6 GHz made the CPU consume 300% more power or something. Sounds ridiculous. More data is needed.


Edit: With regard to the other points raised, CPU clock (i.e. GHz) has everything to do with "speed". The other critical factor here, as theBostonian pointed out, is instructions per cycle. These two parameters work together to produce almost all of the raw "speed" a CPU has. Not part of this discussion are pipeline tricks or the buses which feed the CPU its data (i.e. RAM speed and others).
 
gentlefury said:
Serously? Everyone here is so damn negative that I failed by stating a fact???

GHz has NOTHING to do with speed.....how many pulsing cycles of electricity the CPU can produce. This is how we now have low energy consumption cpu clocks that are MUCH more powerful than they used to be. They can utilize those cycles better and perform more flops per cycle. Less GHz does NOT equal slower speeds....



But of course, the genius masses here are ALWAYS right.

How awesome is it to be 15 and SOOOO smart?
Click to expand...

OMG. This is funny. Thanks for some levity on this thread.....so funny.
 
monksealpup said:
Way too many people just stating stuff with no expertise, knowledge, or cited sources on these forums.

The power consumed in chips is proportional to the operating frequency (i.e. the chip's clock) and the square of the operating voltage. (Source.)


The engineering tradeoff here is that, for higher frequencies, you want more voltage since you need to reduce the rise- and fall- times of the signal, but more voltage means much more power consumption. Alternatively, you could develop new/better semiconductor material device science techniques or improve manufacturing to reduce these times through in a material properties sense (or, better, reduce the required operating voltage since it's proportional to the square here). This is why ultra-low voltage (ULV) CPUs are so sought after for ultrabooks.


Anyway, clearly the CPU speed can't account for the roughly 20% power difference here. The GPU, screen, and other electronics make up a substantial proportion of the power consumed during operation so it would be like going from 2.3 GHz -> 2.6 GHz made the CPU consume 300% more power or something. Sounds ridiculous. More data is needed.


Edit: With regard to the other points raised, CPU clock (i.e. GHz) has everything to do with "speed". The other critical factor here, as theBostonian pointed out, is instructions per cycle. These two parameters work together to produce almost all of the raw "speed" a CPU has. Not part of this discussion are pipeline tricks or the buses which feed the CPU its data (i.e. RAM speed and others).
Click to expand...

Ummm, I pointed out its flops per cycle actually. If GHz were the only factor in speed than 1.8GHz would still perform like crap! Its how many flops per second that matters! Sure the higher the ghz the more potential it has to processor more instructions...but power is not the determination of speed....only the minimum determination. A 1.8 GHz proc running 3 gigaflops per second is faster than a 2.6Ghz running 2.6 gigaflops per second. (these numbers have nothing to do with actual statistics...since everyone is so damn literal here, I have to preface that)
 
This intellectual debate is most annoying. If you know something to be true why try to convince someone else? You are not getting paid for this so let them stay uninformed and ignorant so that you may take advantage of them in the future.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back