Jock,
Thanks for sharing the report from your call to IR. Some good stuff in there.
Looks like I may have been too negative about embedded FPGA. I still stick by everything I said, but I purposely overlooked the field-programmable part, which has not been the driving force for the PGA's that I'm familiar with. But Cisco seems to be very interested in them, and that could in fact be a big deal. I'm not that familiar with routers and such--does anyone know what Cisco would be doing that would utilize field-programmability on a wide scale? I would think that things like routing tables would go in software instead of hardware and things like protocols would be stable enough that you wouldn't want to be changing the hardware all the time for them.
After many false starts, it appears that the ATM is about to take off. (As previously noted, it was not expected that the capabilities of Ethernet would be expanded as long as they have, which is what forestalled the growth of ATM)
I had a recent conversation with a brother of friend who is an engineer in the telecom business. I asked him about ATM replacing IP for WAN. He said he didn't think it would ever happen, because bandwidth is going to rapidly get so cheap that the network protocol won't matter. Since IP is already so widespread, there will be no reason to replace it with ATM. OTH, that reasoning may not apply to LAN's where Ethernet is currently king.
Regards,
G.P. |
