Slashdot Mirror
Sony Pledges More Accurate Laptop Battery Figures
Slatterz writes "Ever wondered why you never get the 10 hours of battery life advertised with your new ultraportable? Battery life ratings have been a joke for years, so it's interesting to hear that one big vendor is picking up its game. PC Authority says Sony is abandoning the usual (and wildly misleading) JEITA method for coming up with those 10+ hour battery numbers (they're still using JEITA, but not the usual way). Interestingly, the story has links showing the old and new steps Sony takes to come up with those battery predictions. It's good to see the industry coming clean on this issue."
Properly, we should be told the capacity of the battery and the consumption of the machine at highest and lowest levels.
For example, my Lenovo X61 gets between 4 and 8 hours on its large battery. The difference comes from how I tune the machine.
At least for laptops using Intel chipsets and Linux, powertop makes it very easy to measure battery life, and (more importantly) tune it. I get my 8 hours by by switching off the wifi, usb ports, killing programs that do too many interrupts, turning down the brightness, etc. Powertop shows exactly how many watts the machine is using. The battery has about 70 watt/hours so when I get it down to 9 watts, that gives me about 8 hours.
My blog
Everyone should read that, it's good for most computers running Linux.
Hard disc manufacturers are in the right though - mega means million, giga means billion, tera means trillion. It's the world of computers with their binary-derived values that are wrong.
This has already been discussed in great detail, and the decision was that a binary gigabyte (2^30 bytes instead of the decimal 10^9) should be called a gibibyte (GiB).
2^10 bytes (1024) is a Kibibyte (KiB)
2^20 bytes is a Mibibyte (MiB)
2^40 bytes is a Tibibyte (TiB)
There are even a few people who took notice of the decision and switched usage.
It's clearly not the whole computer industry, though, is it? Otherwise we wouldn't be having this discussion in the first place. Some parts of the computer industry call a gigabyte 1,000,000,000 bytes, other parts call a gigabyte 1,073,741,824 bytes. One of these standards is consistent with the usage of 'giga' in all other scientific and technical fields, while the other is unique to computer science. To my mind, calling 1,024 bytes a 'kilobyte' was just about acceptable, since the difference wasn't so great and 'kilo' was a convenient shorthand. But calling 1,073,741,824 bytes a 'gigabyte' is really pushing it, and now we're starting to build terabyte drives and it's getting ridiculous. If you want to use substantially different multipliers from the standard, don't use SI prefixes for them. Make up your own unit names.
Real Daleks don't climb stairs - they level the building.
Well they don't use powers of 1024 but powers of 2. ;)
Ummm. They do use powers of 1024 because they want to match the SI prefixes reasonably closely.
1024 is used because it is a power of two making the SI approximations powers of two.
1 B = 1024^0
1 KiB = 1024^1
1 MiB = 1024^2
1 GiB = 1024^3 and on.
Not really. The difference goes up by 2.4% every iteration. It becomes relatively greater, not just absolutely greater. Terabyte disks are now available: a tebibyte is 10% more than a terabyte, and that's almost 100GB - quite a substantial difference.
But who's 'losing' bytes? I bought four gigabytes of RAM when I built my new PC, and it seems that the manufacturer has thrown in just under 295 extra megabytes for free! I'm quite delighted. Everyone else in this thread is just wilfully taking the 'half empty' viewpoint :-)
Real Daleks don't climb stairs - they level the building.
Ya know, when I ask Linux "How much can this 1TB drive hold?" and it reports back "900 GB".
And when you copy files to it, you find that it only holds 760GB, because your files are small enough that the file system overhead eats more than 10% of the available space.
Gibibyte is dead because the difference between 2^30 octets and 10^9 octets is small. The computer industry uses Gigabyte for 2^30 octets because it works in powers of two, so the storage PART of the industry should do the same thing.