HP's ambitious plan to slash energy consumption in its PCs by 25 per cent by 2010 has been greeted with scepticism in some quarters. At the Consumer Electronics Show (CES) in Las Vegas the computer firm proclaimed its intent to cut the carbon footprint of its desktop and notebook PCs by adopting more efficient power supplies …
The manufacturers making their laptops component upgradeable, that way you don't have to splash for a full laptop every time you want a processor upgrade or if you damage the screen. What if the mobo was upgradeable/swappable as most laptops are "too slow" way before their screens go out of acceptable quality. This way you don't throw all of the manufacturing energy away each time something needs to change.
Fit a better power supply unit with a more efficient transformer. Not sure it will gain you 25%, but if you use other higher quality parts you may get there.
lower powered tech already right in front of everyone
Look at the OLPC $100 laptop for how to achieve lower energy consumption.. or even the intel classmate.
Of course running a decent (i.e. non-MS) OS that doesn't waste mountains of power and processing helps alot.
Laptop power supplies
Do laptops still have transformers? I thought pretty much all consumer electronics today has switch mode power supplies?
You'd probably save 25% power consumption...
...by ditching certain well-known Anti-Virus softwares which throttle the whole system...
Seriously though, with all modern processors and especially mobile processors slowing down their clock (reducing power) according to computational demand, more efficient and less bloated software really should translate to electricity savings.
On the scale of other domestic and corporate power-wasteage, modern laptops really barely register (otherwise the batteries would last even less time). 30-60watts for a laptop, compared to a 24/7 1kW base-load per employee at my workplace...
Point of reference?
What are they using a point of reference? Their PCs of 2010 will consume 25 percent less energy than WHAT? 25% less than a PC using an Intel Itanium? Xeon? AMD Athlon 64 X2 6000+ (125W)? Athlon 64 3800+ (35W)? Are they talking business PCs with lower-end integrated video or gaming PCs with high-end, high-power video cards?
They can already make that claim now. Build two systems -- one based on the Athlon 64 X2 6000+ at 125W, and the second based on the Athlon 64 3800+ at 35W. The second system will use roughly 49% less power than the first system (assuming the other components use a combined power of 60W). They could even use a VIA processor such as the C7 to reduce the power even more, though most people probably wouldn't like or accept the performance hit.
It's all in your point of reference. Stuff like this is nothing but marketing talk, signifying nothing.
Guess what? Switched mode PSU chip manufacturers (Maxim, Linear, et al) are *always* coming up with smaller, faster, cheaper, MORE EFFICIENT chips and PSU designs. Cos more efficient means less heat, so more reliable, and usually smaller footprint, less components, which means cheaper. So I really doubt there will be 2.5% saving, let alone 25%, in a 'more efficient transformer'.
What's so difficult
What is so ambitious about a 25% cut in power consumption over a . Inneficient power supplies, more efficient power supplies, better power management, more efficient CPUs, DRAM and graphics.
I managed to substnatially reduce average power consumption on my home PC moving from a single core Athlon to quad core Intel through careful choice of components, better power management etc.
Of course some of the potential gain is bound to be swallowed up in higher specs but really 25% is not exactly demanding.
HP can do this easy
Since they are broken about 25% of the time, they won't be turned on. Green machines!
Whatever happened to "Reversible Computing"?
Which promised major energy savings, not to mention avoiding the potential meltdown of increasingly powerful processors...
It wasn't for real world use, sorry if you were taken in by the hype and misrepresentation, but you weren't alone.
When you accelerate (or decelerate) an electron (which you need to do when changing a logic 0 to a logic 1 or vice versa, which goes on a lot in a computer), it emits electromagnetic waves which means it loses energy into its surroundings.
There are a tiny number of circumstances where that electromagnetic energy can be mostly recovered (e.g. in a tuned circuit containing a coil and a capacitor) but in the general case the energy has GONE FOREVER from its original home out into the big wide universe, and therefore the process of computing cannot be reversible in any general meaningful sense. Put another way, entropy has to be increased, as it inevitably and inexorably does.
I just saved 30%
The fan in the power supply of my home desktop system started making some ugly noises just before Christmas. So I looked around for a new one with "efficient" and "quiet" as my two primary criteria. I bought a Seasonic S12 II SS-330. Since it has a honking great 120mm fan, I disconnected a case fan while I was installing the new PSU.
Result: at the wall A/C power consumption is down by 30%
I'd ask Mr Handy's opinion.....
.... but then IBM stuffed up their PC bizz, admitted defeat and flogged it to the Chinese!
Penguin suit, please!
I know, I know, I know!!
I know exactly how they're going to do it. They are going to lead the way in firing lots and lots and lots of people. And probably investing in hampster wheel tech too, but the firing seems more likely to work.
Then if everyone fires loads of people the world will be a greener place - and we haven't even had to think about virtual millarkies or any of that other complicated gobbledygook. By the way did you know there's a correct spelling of gobbledygook and Firefox knows what it is? Irrelevant to most I'm sure, but it impressed me.
Shiny objects and all that I suppose.
(sorry, getting back on the meds immediately).
OLED screens and Solid State Discs then?
Should be a whopping powe saving for laptops - also cooling can be cut so there's more.
300W+ desktops can be replaced with 5W dumb terminals in office environments as the vast majority of work is dealt with on servers anyway.
All money in the bank - and we don't need any new nuclear power stations either.
You can reverse the computation as you as you don't look at the result
So-called "reversible computing" promises zero-power computations by having the calculations done as a wave and recovering all of the energy. This requires careful attention to design details. All N-input logic elements have N outputs, and all of the input energy is directed through to the outputs. Every output must be used, so that the energy can be "reflected" back and recovered.
One of the problems here is that sampling the 'result' output removes that energy from the system, making the computation irreversible and therefore consume energy. Doh!
@ Dunstan Vavasour: "transformers"
"Do laptops still have transformers? I thought pretty much all consumer electronics today has switch mode power supplies?"
This may be just a terminology problem. Switch-mode power supplies still contain transformers - just much smaller ones because you can get a higher energy throughput if you increase the frequency you send them from the 50 or 60Hz that's supplied by the mains. But I think the commentator meant power supplies generally.
Switch-mode *regulation* is the big efficiency gain - instead of getting the right voltage under differing conditions by a varying resistance, giving off a lot of heat (linear regulation), you only supply what's required by switching the supply on and off very quickly. Switch-mode power supplies do this inherently, but you can also add a switch-mode regulator to the output of a conventional transformer if you want. There's no point in doing this nowadays because the saving in copper and iron (with their associated space and weight) more than outweighs the cost of the high-voltage components needed to make a switch mode power supply. So everyone uses switch mode nowadays.
There is often quite a large scope for improvement in the efficiency of switch-mode power supplies by using more sophisticated designs, using more and/or more expensive components. As has been pointed out, the disadvantage of a inefficient supply, apart from its wastefulness, is its heat generation. The heat has to be got rid of. In a desktop computer it's cheaper to make a horribly inefficient supply and bolt on a fan to get rid of the heat than it is to make an efficient supply. Putting a fan in a laptop power supply makes it larger and less reliable, so these designs are more efficient to avoid needing a fan. Couple this with the efficiency of the laptop itself, both for battery life and heat dissipation reasons, and a laptop does a lot better than a desktop PC. By putting more laptop technology into desktop PCs (as is already happening with CPUs) you can get quite a gain for not much pain.
@ Matt Bryant
".... but then IBM stuffed up their PC bizz, admitted defeat and flogged it to the Chinese!"
Course they did. I mean, I wouldn't want to question your obviously superior intellect, but isn't there a small possibility that IBM might have decided that PCs and laptops are commodity items, and therefore didn't fit in with IBM core strategy? As opposed to 'stuffing it up', as you so eloquently put it?
easy fix for how soon we forget the old P3 class which ran between 20 to 30 watts !
Only sell low end machines.
I can see this easily happening. All they have to do is only sell low end machines.This is all that 90% of the users need anyways so no big deal.
85% by 2010...how about NOW?
Wow, HP is looking to increase power efficiency to 85% by 2010. Today’s power supplies - at least the ones Dell is using - are already running about 80% to 85%. By my estimations, this translates to about 25% less power use. So companies talking about 85% power efficiency by 2010 haven't been paying much attention to the efficiency gains made by the rest of the industry, and more importantly are pretty much behind the curve.
Am I missing something here?