Most of us are so focus on the upgrade and downgrade by the analyst. It is true that analyst can move the market, but most are just followers and bean counters. I think I can qualified as a networking expert and I have talked to many bean counters before and I know how ignorance they could be.
I believe 95% of bad news for 3COM is over. 3COM does not have sales problem, it had the channels stuffing problems before. If you look at the upcoming products cycle, 3COM has one of the strongest in its history. 3COM will be able to maintain its shares in NIC and Modem ($99 internal 56k softmodem), and I believe it will cause lots of problems in the enterprise switching market for CSCO and Bay. Let me make a comparison of 3COM 3900 (will ship by the end of March) with CSCO and Bay products.
3900 can deliver 9.8Mpps at around $5k(24 FE + Gig), CSCO Cat5 is one generation behind, 10% the performance at 2 times the cost (without the Gig) of 3900. support only FE and CSCO say it will have expensive Gig module to plug in slow backplane. The newly announce 2916 is a total screw-up from CSCO. CSCO markets 2916 as FE + Gig Ethernet product, but it has a CELL backplane, frame from one port to the other could go through frame to cell, then cell to frame conversion, it is a lot of overhead. It is at best 1/3 of 3900's performance at roughly the same price and the Gig module is not available today, Gig is only promised by CSCO. Let me predict here - CSCO will have to cut price in a month on the 2916 to get it moving. No way customers is going to pay less functions, 1/3 the performance at the same price. Bay's 350 and Accelar are also have less than 1/2 the performance of 3900 at about the same price.
Layer 2 workgroup switch is price/performance game after all. |
