Dear Readers, I love you all! Thank you all for talking back about my recent post re the Pre overclocking issue. Please allow me to clarify something here.

The passage that most of you took offense on was this:

Given that Palm never did this before, it IMHO is highly likely that the Pre has severe thermal issues, which would be exaggerated by raised CPU frequencies! This means that an extensive computing session could quickly lead to a toasted Pre.

Please allow me to explain a bit more. Every object has a “thermal capacity”; which we can imagine like a bucket for water. Heat is dissipated to the environment (water out of the bucket), and it can also be produced from electric energy (water into the bucket).

Obviously, a healthy device would always have an empty bucket/steady water level as all heat gets dissipated – and the Palm Pre also has that. Your Pre’s work flawlessly at 800MhZ – but this does not mean that it will work so for everyone.

Let’s, for a moment, assume the following: you are on a train (aka bad wireless signal), are charging the aged battery (extra heat dissipation), tethering to a PC which downloads Windows Updates, have BT on and – finally – play a 3D game.

Let’s go on to assume that the amount of heat generated is just a tiny teeny bit more than what can be dissipated. This means that the bucket fills up slowly but surely. It might take hours, but it might just happen.

For knowledgeable users (who see that their box is hot and turn it off), this is no problem whatsoever. But lets now put the Pre into the hands of a dumba$$: he could fry his Pre, theoretically hurt his hands due to the high temperature, and sue.

Palm could theoretically have worded the statement as: this is not for dumb users who have no common sense. However, dumb users are offended easily…which is why Palm chose the wording which they used.

Hope to have this cleared up now!

Related posts:

  1. Intelligent design is lawless design
  2. Hard cases could cause Pre overheating
  3. On the power of small steps
  4. On sensible user interface design
  5. Palm: don’t use overclocking patches

2 Responses to “On thermal budgets in handset design”

  1. Heat isn’t the only problem. You claim to be an embedded developer yourself, so you should should know that there always is timing margins that have to be met. Overclocking MAY work on one device and MAY NOT on another. Or, what makes things worse, MAY work on lower ambient temp and MAY NOT work on higher. This does not have to be related to thermal stress.

    A device manufacturer usually knows very well (who else would?), what margins to meet and set the TOTAL timing and operating parameters in a way that any average user will not experience any problems.

    If a device can be “overclocked”, this does not mean that this is generally a good idea, as you can never ever make sure that it will work for ANY device out there.

    If you do so, you do it at own risk. No more, no less. You are unlikely to toast your device, but you want reliable performance.

    One thing you get for sure from overclocking: Less battery life. As the Pre already sucks on this, you might want to underclock it…

  2. It’s probably an even simpler explanation.

    Heat comes from one thing only: power usage.

    Power usage means battery drain.

    The Palm pre’s battery life is already not so good. Palm is probably concerned that overclocking will drain it even better. The hardware itself can handle the extra heat, but the battery is not strong enough to supply it!

Leave a Reply

(required)

(required)

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Subscribe without commenting

© 2012 TamsPalm - the Palm OS / web OS Blog Suffusion theme by Sayontan Sinha