Slashdot Mirror
Rob Enderle Announces Death of Bluetooth
prostoalex writes "Rob Enderle is typing away (perhaps even on his very own Ferrari laptop) at Intel Developer Forum, noting that Intel gave up on IEEE Ultrawideband and decided to switch to Wireless USB derivative. This, in Mr. Enderle's opinion, signifies the end of life for Bluetooth standard, although Enderle calls Bluetooth 'dead' in the title of the article and 'all but dead' in the actual text."
Its pretty clear most of us don't care much about what Rob Enderle has to say. Apple has integrated bluetooth and I love it. Its in many devices and its cheaply priced.
Plus I would never be caught dead with a Ferrari laptop.
Yeah- don't tell Intel either.
/. story is nothing but a troll designed to whip you geeks into an anti-Intel frenzy? Imagine that...
Wait? You mean Intel is adding bluetooth support to next gen Centrino, and this whole
The death of Bluetooth view is being advocated from a perspective that says Intel is in the driver's seat - a very PC centric view. Take a look from another perspective - device centred - and the picture looks much much different.
Cheap devices use ASICs and ASSPs to implement Bluetooth. IBM, NEC, Toshiba, LSI, and somewhere down the line Mitsubish are the major ASIC and ASSP players. Now, with that in mind re-read my post, especially this part:
IBM and NEC both just dropped support for Bluetooth in their ASIC core selection (which is key to cellphone, other cheap device, and mobo mfg'ers), LSI and Mitsubishi stopped development altogether after wasting some cash trying to figure out what the spec actually was and how to plug the holes in it safely.
Sorry to be the bearer of bad news, but you bought a phone with an ASIC inside that includes a core that is no longer supported. That means there will be nore updated models of your device, no big deal, but it also means no new Bluetooth support in that line either. Which is what we're discussing. As much as I hate to agree with that ferarri-licking laptop monkey, he's right.
everything in moderation
1. Intel chipsets had USB builtin since the days of USB 1.0, thus for about half of the sold motherboards (those with Intel chipset), USB was just there without additional cost. Firewire required extra hardware until non-Intel chipset designers started to put it into their products.
2. Firewire uses an own controller design to handle the protocol, making the chip design to support Firewire more complex. USB does a lot of work in the software USB driver, thus making for a more simple chip design or an easier integration of USB into the I/O part of the chipset (mostly in the southbridge).
This basicly covers also the pros and contras for USB and Firewire.
USB is cheaply to implement in hardware, and you can add functionality later in the driver. So USB-support for non-commodity platforms is more complex, because you have to write more complex drivers. USB transfer rate is coupled with system load, a loaded system can't keep the full transfer rate, and USB transfers in reverse generate considerable system load at higher rates.
Firewire is more complex to do in hardware, but once it is implemented, the drivers are quite straight forward and generic and thus easily implementable on different platforms. Protocol extensions will break backward compatibility though or require at least a software compatibility layer to run also on older hardware. But firewire transfers are not coupled with the system throughput and can run with high rates on highly loaded systems or slow CPUs.