Talk:Environmental Impact

Does it make sense to add discussion of food and/or recharge facilities to Environmental Impact? If the laptop battery is recharged using muscle power, it will also impact the food a child needs. A simple conversion assuming a 90% efficient charging cycle and 15% efficient human food to muscle power conversion results in 105-140 calories of food required for each laptop recharge. Extra calories used due to thinking more aren't included! --MWarren 20:15, 5 December 2007 (EST)

	NiMH	LiFePO4
Wh	16.5	22.0
cal	14187.4	18916.6
kcal (aka food Calories)	14.2	18.9
charge cycle efficiency 90%	15.8	21.0
human food to muscle work conversion efficiency 15%	105.1	140.1

One of the reasons why the solar option is so appealing. --Walter 21:11, 5 December 2007 (EST)

Also, even in the 3rd world, a majority of children have some play room between the calories they consume and the physical work they are expected to do. If they don't, a laptop is really not what they need anyway, and probably not what anybody would give them, to boot. Homunq 22:01, 5 December 2007 (EST)

Hopefully the children do have some extra play room; the issue of food and/or recharge facilities seemed like another good topic for this page. Here is how Dr. Negroponte responded to the laptop vs. food question in an OLPC Analyst Meeting:

"...Why give laptops to kids who are dying of hunger, malaria, who don't have clean water?"

The answer to that question is very simple, just substitute the word "education" for "laptop" and you'll never say it again. Because, clearly, if a child is dying of hunger that child needs food right then and there. Education is not on .. it's a separate thing.

But when you solve and you address and you work on all of those problems nobody I know would say: "By the way, let's hold off on education." The reason you don't do that is because education happens to be a solution to all of those same problems." --Nicholas Negroponte

--Mwarren 13:28, 6 December 2007 (EST)

Power Consumption

This information appears to misleading to me but it could easily be cleared up:

"Assuming 2000 hour use per computer per year at the low end of power consumption (idle-mode), and an average 3 year lifetime, 87 billion kilo-watt hours are used to power the world’s personal computers per year. If every personal computer was replaced by an OLPC XO laptop, just 1.5 billion kilo-watt hours would be needed: thus 85 Billion kilo-watt hours could be saved."

I tried to sort out the calculation and came up with a power consumption of 3.3 watts. A previous calculation using a 16.5 watt-hour battery capacity divided by the 3 hour operating duration came up with 5.5 watts of power required... Something seems off but it could possibly be a blending of operating versus idling power requirements. The statement above could be cleared up by listing the OLPC standard "operating" power consumption.

After surfing the site some more I've seen a power consumption of 2-4 watts listed a few times. It seems essential that this info be above the bar graph if this is indeed the standard power consumption.

I looked for this page in search of the power requirements of this wonderful little computer. Initially I found a statement that was more word problem than solution to my query. That's funny but it shouldn't be. The word problem blurb is a misplaced analogy because this computer isn't going to replace my desktop and hypothetical pollution abatement is irrelevant. What's really important is how much power will be required in the remote area this computer will be deployed in. This information is essential for those of us interested in solving the power problem in these remote areas. Take 10 kids and 10 computers and measure the average power consumption for 10 days and post that info. A simple computer program could be written to perform this task. Why not? Lee 23:06, 18 January 2008 (EST)

XOs compared to servers?

At the OLPC office it was suggested that a server farm of linked XOs would use less than regular servers. Impressive if true - but can anyone offer more insight on this?

This isn't to suggest that linking XOs is the best way to set up servers, just that there's a lot in XO design to be applied to servers. --Chriswaterguy 17:19, 2 November 2008 (UTC)