Originally Posted by benanderson89
You need to firmly establish the context of your sentences. Also sarcasm.
Strange how you should say the clock directly relates to wattage.
Current Gen Ivy bridge Mobile Processors as a prime example:
i7 2630QM - 2.0GHz 45W
i7 2860QM - 2.5GHz 45W
i7 2640M - 2.8GHz 35W
If its directly related to wattage then why the hell has the 2.8 got a 35W TDP? Please explain your logic.
The reason the QM range will need a higher wattage is because its on chip features require a higher power draw - screw the clock speed. Its clearly shown to be irrelevant here in processors of the same generation and family.
Are the 2630 the 2640 and the 2860 not different models? Take any one of those chips and run it at different clock speeds and get back to me.
If that doesn't convince you then explain why Intel uses variable clock rates in its new chips to manage power usage.
It was to establish that clock and wattage mean nothing when two CPUs with different specs but identical clock and wattage can perform vastly different. By your logic they should be neck and neck. Also, why couldn't you pit an x86 and a Sparc together to see which performed better? I think that would be a rather interesting comparison.
Your inability to read for content is amazing, this discussion was never about different CPUs with different specs. It is about what happens when you up the clock rate on any given chip. Point in fact they run hotter.
As to the x86 and Sparc comparison the two chips aren't even designed for the same types of workloads. Well that and it just pulls the discussion off track which seems to be your goal here.
Yes, there are more important factors. But saying someone goes for a light machine because they are a weakling just shows the idiocy on your part. I wanted a really lightweight machine when I attended University because it was on my back the majority of the day in a bag full to the brim with stupid-thick text books. Because I also needed something powerful for programming.
I can see where you might not like that, however I stand by my remarks. It is silly to focus on weight at the expense of all other features. Frankly I suspect that this was a big factor in the Netbook failure.
The HP ProBook my dad has is a BRAND NEW SANDY BRIDGE MACHINE. Its not an old 'prime' computer from 6 years ago, its brand spanking new! You say thinness sacrifices performance, up to a certain point (sub notebooks), yes. Though the MacBook Pro clearly out performs the ProBook and yet is strangely thinner AND lighter.
So its low wattage but it isn't. What? Compared to other CPUs yes, but its still a Low Wattage, Low Voltage CPU.
ATOM as it currently stands is pretty much a worthless offering. It is not low wattage in the sense of the common ARM chips. Further at slightly higher wattages you can get much better performance out of AMDs Zacate/Ontario chips.
I own two iPads (1,2), an iPhone 4 and an iPod touch. So yes I have clearly seen an iOS device to base it upon. But my point still remains - by your logic of wattage = performance, why do the ARM CPUs seem to outperform or be neck and neck with the Atom when the atom has a higher clock and higher wattage (2~10 vs 0.5~2)? You have been harping on about wattage and clock rate equating to better performance?
You have not seen an ATOM based iOS device, unless of course you work for Apple. Thus you have no idea how an ATOM based iOS device would function. In any event you missed that point just like you missed the point on wattage.
With the AMD and Intel comparison I was making, I was highlighting your statement of "higher clock = better performance". The AMD V140 and Sandy i7 are both current Gen CPUs. The AMD runs at a higher clock rate, yet performs far worse. You either need to explain yourself better or rethink the statements you are making.
I'm really wondering how in the hell you will make it through school, you repeatedly twist simple statemeevenings false representations even after I've repeatedly explained where your problems is. there is no value in comparing clock rates between different manufactures hardware, you can't even do it with different chips from the same manufacture. It isn't my statements that are messed up, it is you that have taken the wrong position and the try to pull in totally unrelated information into the argument.
Maybe you are trolling or maybe you just don't grasp the technology, but you have yet to address anything I've said.
No, the wattage would not increase above 45W. I have a 2.0GHz i7, the 2.5GHz i7 in the same family/gen is still 45W TDP. When my i7 turbo boosts to 2.9GHz, its still 45W TDP. I think you might be confusing wattage with voltage. In which case yes you would be correct. Voltage governs the clock rate, though I shouldn't need to tell you how over-clocking works since you apparently know everything :roll:
Again you are referencing material that isn't related to the argument. Ask your self why the wattage stays at 45 watts when turbo boosting. I will help you along here, it is because the chip either turns off or throttle some of the other cores on the chip. When that core turbo boosts to 2.9GHz it is expending a lot more power than at its base speed. It really seems like this concept has gone completely over your head. Increasing clock rate increases power in CMOS circuits, it always has and always will.
Look at this another way, if you take that 2.0 GHz chip of yours and over clock it's base frequency will it run hotter?
Also, to answer your question of "why do they charge more for higher clock rates" - they do to make money, but is always poor value. Over-clocking has shown that all it takes is a minute increase in voltage to up the clock rate to the level of the top-of-the-line processor of the same family/gen and gain the same performance.
OK what do over clockers concentrate on to get those chips to operate reliably over clocked? Usually the first thing an Over clocker installs is a better than stock heat sink. Why? Because the chip runs hotter. Why does it run hotter - well two things impact this. One is the bump in voltage and the other is the increase in clock rate.
All comparisons I've made are new and recent processors. I think you may be confusing generation with family.
I honestly think you are confusing wattage with voltage. A higher clock rate increases the voltage.
The definition of voltage is "Voltage is equal to the work which would have to be done".
No I'm not confusing anything here. If you increase the voltage to a processor core you increase disapated power. By the way the clock rate does not increase voltage on most processors, people do that from their BIOS settings. Wattage does go up with clock rate though, it is a rather fundamental point too.
I was showing that a processor with a 299MHz clock could outperform a 700MHz processor (that by your logic is impossible)
My god are you dense or what? Seriously where have I ever said such a thing? Seriously where do you come up with this stuff?
from the same generation despite its lower wattage and vastly slower clock speed, yet the Emotion Engine came out on top because of its on chip features. Not its clock rate and wattage
. Funny how that works.
I'm beginning to think you have a reading comprehension problem because even a troll would not make such convoluted arguments. This discussion has never been about how specialized processor perform in comparison to more run of the mill processors. It is about what happens when you up the clock rate on a given processor. When you do that you increase performance and increase the power disapated. There is no magic here, it is the way CMOS circuits work.
What may be causing you some confusion is that modern processors actively manage power. They do that by turning off cores and adjusting clock rates. This only goes to prove my point though, one can adjust the power disapated by a processor by adjusting its clock rate dynamically. Don't buy this from me then read some of Intels documentation.
I really don't know how to make this any clearer. I suggest you read what I say instead of pulling unrelated information into the discussion. In the end what don't you like about the idea that power disapated varies with clock rate?