Chipmakers Admit Your Power May Vary

Dylan Knight Rogers writes to mention a News.com story discussing the realities of chip power consumption. From the article: "Assessing only pure performance is passe. The debate these days is about performance-per-watt, which seems like it should be a simple miles-per-gallon type of calculation. However, miles are miles, and gallons are gallons. There's no one simple way to measure processor performance, and measuring the amount of power output by today's chips is proving just as difficult."

Comments

Comment removed

[account_deleted]

Comment removed based on user account deletion

How is this news?

[AuMatar]

Performance being difficult to measure is well known- you can't go by clock speed, or even clock speed*instructions per clcok since these will differ based on instruction mix. For power, a simple inverter will use different amounts of power depending on if its on or off- exact power for a chip is impossible to guess. This is all old news.

Re:How is this news?

[dsanfte]

Uh, actually, watts sucked per hour/minute/etc has been very easy to measure for many decades now. There is no reason why chip wattage drawn should be difficult to gauge in the slightest.

Re:How is this news?

[kfg]

Performance being difficult to measure is well known. . .

it should be a simple miles-per-gallon type of calculation. . .

It is. That's the problem. Mr. Krazit seems to be utterly clueless. I defy him to predict the milage I get the next time I go out for a drive.

Hasn't he ever noticed, like most of the rest of us have, that the milage he gets is not actually the same as the EPA test "prediction"?

That's because the EPA test only gives valid results for. . .the EPA test, which is actually an average of multiple tests.

There's no one simple way to measure. . . gas milage. Sheesh.

KFG

benchmarks

[Anonymous Coward]

This is what benchmarks are for. Compare the performance of two systems with other variables held as constant as possible. This has been going on for years, has it not? If I want a computer to play games, I see what different CPU configurations yield in, say, HL2 with the same ram and video card.

Is this perfectly scientific? No.
Is it practical? Hell yes!

Re:benchmarks

[KermodeBear]

This is exactly how I feel. Theoretical speed is nice and all, but just where does theory and practice meet? Usually never. It's the practical application that matters, which is why, in my opinion, benchmarks of common operations are important. Things such as frames per second in a video game, or how long it takes to encode a DVD, or how long it takes to open up a large PDF... Those are things that matter and, perhaps more importantly, things that the average joe can wrap his mind around. How many Joe Users know (or care) about stuff like clock speed and cache size?

Some miles are up hill and some are down hill...

[EmbeddedJanitor]

Especially with caching and pipelining, MIPS per W gets very difficult to measure. If you can live in the cache you don't need to go fetch from the outside world. If you stall the pipeline, you lose performance. Some operations (eg. DIV) clock a lot of transistors, some (NOP) don't. It was a lot easier to measure MIPS/W when devices were synchronous. Now they're a group of asynchonous entities (core CPU, cache,...).

BTW, EPA mpg are measured without using real mile on real roads.

Well...

[radish]

Miles per gallon are hardly constant either. Uphill? Downhill? 10mph or 100? Highway or city? Same difference.

It's not just the CPU

[Rosco P. Coltrane]

For most users (i.e. not power-users doing heavy calculations for some scientific purpose, or high-quality video editing, or raytracing), most processors provide way more power than needed, and have done so for years. Or at least, they *would* provide all that power if the software running on top of it wasn't bloated and unnecessarily complex, unoptimized and badly written. And no, I'm not just talking about Windows, I'm including Linux, MacOS and all the others in the bag.

The best proof that modern software makes modern hardware suck is that, back in the mid-eighties, I used an Atari ST to do desktop publishing, and it wasn't all that different from what I can do now with a simple PC that would look like a supercomputer back then.

Re:It's not just the CPU

[cgenman]

The brother post may have used a car analogy, but he's basically right. There is a similar situation in Game Development.

Right now, you have machines that will do amazingly powerful things, especially with the Next-gen coming out. So what do you do with that power?

Quite simply, one of the things you can do is optimize less.

For example, early FPS games were written largely in assembly in an attempt to eke out every bit of power from the system. It worked, but it was really expensive financially and broke at the slightest provocation. But as computer power increased, you started to be able to write parts of your game in C, and use assembly for the rendering and other intense processes. Then you could write parts of your game in script, and your engine in C. Each step up the chain gets easier to maintain and extend, but at the expense of more clock cycles.

Another example of the above is simple data abstraction. On the NES a simple boss, like Frankenstein, would need his own code base and optimizations to get him to work. Now that boss might be pulled in through a "Boss" class of an "NPC" class of a "Character" class of a "Things" class. His behaviors might be tagged by a designer from a separate Behavioral AI library.

This might seem like it is being lazy, but remember that largely the limiting factor on huge projects is the complexity of the code. If you had to code the behavior of all of the monsters in your game one at a time without abstraction, you'll need to pare back the number tremendously. If you had to write the display code in assembly, you can forget about having an artist script a wispy smoke polygon shader effect. It would be much faster to hit the raw iron of your graphics card, but if you don't go through Direct X or Open GL you can forget about ever finishing.

So you optimize less, you create broader, slower code, and you get a lot more done. You get more game for your buck.

Is the code the most efficient ever? No. But that's not the point.

Re:Some miles are up hill and some are down hill..

[tomstdenis]

You need a stats class, badly.

No offense, but nothing is perfect. This is why we have a thing called "standard deviation".

Me hitting the letter "e" will probably not take the same amount of energy to process twice. But I bet over 1000 e's the standard deviation could be found and would indicate that 66% of the time it's "x J +/- y" and so on...

So you sample something like "building the linux kernel to a ram drive" 100 times, find the deviation and use that. The tighter AND lower the better. The wider and higher the worse.

Tom

Re:We have a nation of SUV's

[tomstdenis]

Because they're stupid. Then they bitch at the bank for the 1.50$ "service fee".

Let's see... processor running full steam instead of low power mode when idling probably amounts for a waste time of more than 90% (unless you work/live at the box).

Opteron at full == 95W, at low == 35W, diff 60W. price per KWh is about 7 to 10 cents. Let's say 8.5 to be close to middle. 60W * 24 * 31 * 0.085 = $3.80 per month. Probably double that once you factor in power supply inefficiencies and cooling costs. So you spend ~8$ per month because idling the box is "too hard". But $1.50 at the bank is robbery. ... hmmm ... Now tact on a bunch of other needless things like gas guzzling cars, retarded cable, etc... Each on their own is trivial to expense, but together you're spending hundreds of month stupidly... /rant...

Besides, multiply your wasted 44.7KWh per month by the millions of other people. That results in higher demand which raises the price. The price goes into other things like the cost of producing things that require electricity. So because people like you are apathetic and think "me being wasteful is ok" you end up paying more at the gas pumps and the stores for everything you touch.

Tom

I would like to know...

[Garabito]

Why, for a given chip, power consumption raises with clock speed? I know there's corelation, but I'd like to know the physical relation between the two variables.

Re:The Solution

[Dorceon]

Optimized for what pipeline length? The wrong one for the competition's processor?