If you sleep the computer, the main battery continues to be used to
maintain the RAM contents. If you run the battery down until it's
empty, the laptop may shut down immediately. That could be why they
invented hybrid sleep, where not only are the RAM contents kept
in RAM, but also written to the disk. If the computer loses power
due to a drained battery, the computer can use the hiberfile on a
restart, to return to the same session.
Generally, the computer takes care of itself. Unless the user
manual warns otherwise, you can just use it. Hardware has
advanced enough, that the computer knows a lot about what is
going on. The hardware designers have choices, if they want to
use them.
A lithium battery is not supposed to be patterned by the way it is
charged and discharged. There is a claim, that leaving it full charged
to 100%, and then leaving it in the corner for six months, isn't the
best of conditions. It may be better for long term storage, to leave
it half charged. However, if you planned on not using the laptop
for that long, you should remove the main battery. If a lithium battery
is discharged below a certain level, the charger is not allowed to
charge it on the next startup. In normal usage that won't happen, but
in some storage scenarios, it's possible it could drain to the point
that it cannot safely be charged.
Hibernation, may cause all power to be removed from the motherboard.
In which case, you may have more battery capacity left later when
you need it. If you sleep, and the computer really keeps contents in
RAM, that uses perhaps 1 watt of electricity, so will eventually
drain the battery. The end result, is the next time you go to use
the computer, you may have to plug in the charger before the
laptop will respond. If the laptop has hybrid sleep, your session
will be saved. If it has "ordinary sleep", the session contents would
be lost.
Talk to the vendor first. Ask the vendor, what is their policy on batteries.
Do they stock large quantities, so that stale stock is the result ?
The battery may come with a factory date stamp, or it's even possible
there is electronic information in a smart battery, to indicate when
it left the factory. I haven't investigated this, because this is
my first laptop, and my battery is still working. But I have read
complaints from people, who are getting a lowered capacity from
their replacement battery.
A rotating hard drive has moving parts. They can wear. I've had one hard
drive at work, that lasted for around seven years of pretty well 24/7
operation (part of a Unix workstation). But the hard drive needed to be
rotated 90 degrees from its normal orientation, when powering the system
from a cold start. The mechanical parts had too much friction for the
drive to start up properly in its normal orientation. That hard drive
had a ball bearing motor (not an FDB). The motor was noisy, but there
was no indication I was about to lose any data. That's a pretty good
lifespan. The drive was still working, when it was finally retired
from service.
Modern drives have fluid dynamic bearings. Those have zero friction, as
long as there is lubrication in the reservoir. Motors can fail at any
time, if the lubrication is missing. The motor will stall and not turn,
if the lubrication is gone.
Other parts that wear, are the bearing for the head assembly, which is
a good quality bearing (cannot have much play in it, or the drive won't
track properly). There is a flex cable which connects the electrical
end of things, to the controller board. Each seek, will flex the cable.
I've never been able to find any information on how many cycles that
cable can take. Whether it is 10**6 or 10**9 or whatever. When a
rotating hard drive is used in an Internet server application, where it
constantly seeks 24/7, such a drive lasts about a year, before it falls
apart and fails. So that number should give you some idea how long
a good quality drive lasts, under extreme service (never rests).
SSDs on the other hand, can only last for the total_data * write_cycles
amount of writes. Say the flash is only rated for 3000 writes. The SSD
is 100GB. Once you write 300,000 GB of write operations to the SSD disk
(write, delete, write again), it is finished. Flash blocks will fail,
the spare area will be used up, and eventually CRC errors will be reported
to the user, or (more likely), the drive will simply cease to communicate
with the computer.
The less writing you do to the SSD, the longer it lasts. If you did no
writes at all, then the drive still has reliability issues that can
cause some units to drop out. But being a solid state device, the
overall reliability is pretty good. The reliability then, has two
components. A "bathtub curve" type reliability associated generally
with electronic components, with a "wearout" phenomenon added to the
result. If you do lots of writes, such as write continuously to the
drive on purpose (say, some kind of benchmark), then it won't last
very long at all. If the SSD drive was rated for 200MB/sec write rate,
how long would it take to hit 300,000 GB at 0.2GB/sec ? About 17 days ?
So if you abuse my example SSD drive, it could wear out in as
little as 17 days.
Unless you pay "enterprise prices", the drive will be made with MLC flash..
If an MLC drive cost $200, and an SLC drive cost $800, will a company
just "giving you" an SLC drive stay in business for very long ? No.
So if you want SLC, you have to actively shop for it. It's expensive.
Play a game, or run Folding At Home, or some other 100% loading
application. You should understand how the unit operates under
full load. Sitting idling in the desktop, is not an indicator
of how bad it can get.
The Prime95 stress test is multithreaded, and will give you some
idea how hot your CPU can get.
http://www.mersenne.org/freesoft
While Prime95 is running, then you can start this benchmark, and
load up your GPU as well.
http://majorgeeks.com/3Dmark_d99.html
The thing is, if you play a game with a "high power" laptop, the
GPU temperature can go up to 100C, which causes mechanical stress
on the GPU. Some of the parts in the laptop, have thermal limiters,
and will crank down their efforts, if they get too hot. But the
temperatures allowed in there, are still quite hot. Continued
burst cycling (play a game for five minutes, go get coffee),
where the component temperature goes up and down, causes
fatigue. And even if the component doesn't fail, the solder
joints can fail from the stresses.
Years ago, stereo systems used to have the high power transistors
fail on them, after several thousand temperature cycles. Hardware
has come a long way since then, and they're better able to
match the thermal expansion coefficients, or redistribute the
stresses in modern components (with materials like underfill).
If I tried