>>>WSTLs DSLAM was supposed to be out by today. does anyone think there is a correlation? maybe BEL is delaying because of FCC and WSTL DSLAM delay.
Trey: Makes sense to me and maybe the chip situation is the reason for the DSLAM delay.
>>>P.S. Ti doesn't care about anybody let alone WSTL. The Mot chips works the best at lower power anyhow, so why go for TI.
The above quote is from one of Pipers posts. I've been wondering about this recently, whether WSTL would use C6X or CG for DMT. Maybe Piper or Jim W from the Amati board or someone more knowledgeable on the chip situation can comment. I remember reading an interview with someone from Amati that mentioned that while they were disappointed with the CG delays they still expected CG to be the most efficient, power and cost wise, solution. Granted this was long before the TXN/AMTX merger. There was a post here, from Jim Miller I think, a month or two ago that talked about the benefits of using an ASIC as opposed to a DSP. The benefits of using C6X as I see it are the ability to subscribe multiple modems from a single DSP (4 now I think, 16 later) and a migration path from G.Lite to full blown DMT via software download. Does this outweigh the power advantage of the ASIC?
Westell and TI will initially work together to define and integrate TI's technologies into Westell's SuperVisionO xDSL Access Multiplexer (DSLAM) and customer access equipment. The DSPs will incorporate discrete-multi tone (DMT) software technology. According to industry analysts, DSP-based solutions are expected to provide the industry's best cost/density/performance implementation for standards compliant DSL system solutions. The Westell DMT products are planned for customer availability in the first half of 1998.
This was from the Strategic Alliance PR. Westell's PR that Lucent will distribute the DSLAM makes no mention of the C6X, either does the recent PR, DMT solution to be Available for Lucent 5ESSr-2000 Switches. In fact I don't think they've mentioned the C6X since the original strategic alliance pr. Could be that Westell is testing CG now and that's the reason for the delay. If they use C6X at both ends they've got power and cost issues, (I assume), but the modem density and upgrade benefits of the DSP. If they use C6X at the ATU-C and CG at the ATU-R they have the ability to oversubscribe the modems at the CO but lose the migration path from G.Lite. If they use CG at ATU-R&C they lose the migration path and modem density benefits of the DSP but have a less expensive more efficient product (I think). I read an article at Techweb that mentioned that C6X had so much computing power it could support both CAP and DMT line codes. You could have C6X at the ATU-C, Globespan or CG at the ATU-R. It's all to complicated. I'm not surprised its delayed. Dave |
