Vladimir:
Good questions. I tried to post this last night, but SI seemed to be out to a late lunch in the post receiving department. Now a bit dated, bu FWIW, here it is.
I think the board has touched on an important technical issue- the ASIC (Application Specific Integrated Circuit) vs. the DSP chip.
My understanding is that ASICs do the job faster, and without loading the CPU with tasks that slow it down, but are really skill intensive to design, and require volume use to be cheap enough. New ASIC designs can include the whole ball of wax. Previously, there were practical limits to ASIC complexity. Their ability to accept downloaded software upgrades is dependent on design, and may be less than a pure DSP.
DSP chips offer non- specific and very versatile processing power, easy upgradability, but are innately slower. Since they are less task specific and less optimized, they contain a lot of stuff that may not be necessary for the job at hand, and for other reasons.
TXN has repeatedly made it clear that it now wishes to be thought of as a "DSP" company, and seems therefore committed to the less specific dsp chip approach.
Throughout the entire three year period in which the xDSL technology has been developing (and has been repeatedly stalled), it seems to me that the time factor leads us away from the generic DSPc solution, and toward ASICs. The issues are speed and power consumption, cheapness and simplicity of manufacture. As standards solidify, more of the guts are "ASIC" able, and the need for radical software alterations is reduced. Have we reached that point in the standards game? I think not, but convergence comes.
I gave up on AMTX when time reduced the value of their proprietary technology to the point where they were not viable as a stand-alone, and because time showed their management was dropping too many balls. As a TXN aquisition, AMTX might be useful, if their (TXN's) objective is to retain some legal control of the idea. Bears directly on your question.
I am, and have been, long WSTL, but--Ah, where have I seen this face before?
The never-ending delays have changed the landscape here, fellow sufferers. I feel for WSTL, their employees and their management. As has been said, the last months must have been tough on morale. Still, the key questions are these:
1) How exactly is the scene changing over time? Will the new guys steal the eggs out of the basket?
2) Will (can) WSTL cope?
WSTL was rightly criticised for putting too many eggs in the CAP basket. I bought WSTL when they made it clear to me that they were taking a broader approach to the problem. Pair po-pooed the whole xDSL thing for a long time, while carefully developing a damn good, low power drain ASIC chip. They were, I then thought, late to the party, but--maybe not.
Perhaps we were just REAL early.
I fear an omelette is in the making here, and it is our eggs.
The party better start soon.
Thoughts?
Jim |
