To: Jock Hutchinson who wrote (13854 ) 7/24/1998 12:55:00 AM From: Grand Poobah Respond to of 25814
I just listened to the conference call. Thanks, Jock, for the number, and also very much for your excellent, prompt, and thorough notes and insights. My overall impression was that it was very impressive. This is only the second conference call I've listened to (the first was not LSI) and this one compared very favorably to the other both in terms of the depth and the optimism of the discussion. My very first impression was that Wilf has an interesting accent. It was a long call (~1 hour), but I considered it a good investment of my time as an LSI investor. Although I believed strongly in LSI going in, I am now much more comfortable in my position as a long-term investor in LSI. My belief that they are in the right (growing) markets was strengthened. Now, a few notes on things I picked up on that Jock did not mention or which I found especially interesting. (Jock made an excellent summary, but sometimes it is helpful to have an additional perspective.) Wilf's comments on the gross margin improvement were that it was the result of lower manufacturing costs due to the weaker yen, higher factory utilization, and a richer product mix. It was mentioned that they had 2x the design wins this quarter compared to last quarter, and this was driven by networking. The strength of networking was mentioned repeatedly. The communications segment had revenues >$120m. LSI, as mentioned by Jock, had 25% market share in DVD. They said this makes them the dominant supplier worldwide. I thought this was quite significant. They expect DVD to double for next year. LSI also has 80% of design wins in Fibre Channel. Some very significant comments were made regarding Fibre Channel. This looks to be a dynamite market for LSI and one that will especially benefit from synergy with the Symbios server business. LSI currently has a digital Fibre Channel controller and Gigablaze analog transceiver. The transceiver is operating at 2.5 GHz, and LSI is the only one doing this in CMOS, and this gives them a cost advantage over GaAs competitors. Although it was not said, it would seem to also give them a big leg up in making a SoC out this. It takes some expertise to make a good mixed-signal chip out of one digital and one analog, but you dang sure can't integrate a GaAs chip with a CMOS chip. Various notes on Symbios: Symbios has two main product lines--SCSI host adapter chips and boards (~$400m) and storage systems (~$200m). They also have some non-storage markets in wireless and set-top decoders. My understanding was that this acquisition accelerates LSI's position in the storage market by 3-5 years, not for 3-5 years as Jock interpreted it. I could be wrong. Symbios brings SCSI IP to LSI, LSI brings SoC expertise, semiconductor processing technology, and global presence in Japan, Taiwan, and Europe to Symbios. Regarding Symbios' manufacturing capabilities, they have an 8" fab with linewidths from 0.8u to just starting up 0.35u. Their SCSI products won't shrink (presumably will stay at 0.8, 0.6, 0.5u) because they need to maintain a 5 volt power supply. This 5V requirement is presumably due to the design of the SCSI bus protocol, which will not change, and requires 5V at the pads. You cannot shrink a product below around 0.5u without lowering the power supply voltage below 5V, or going through gyrations in I/O pad design which are temporary measures at best. I liked Terry Ragsdale's question about the semi industry. I think it about sums up my concerns. And I quote, "Like, are we kinda done with the OEM inventory reductions?" Like, enough already with this slowdown, dude. <g> Sony revenue was down 16% to 11% sequentially, partially due to the yen and partially to long-term pricing agreements on Playstation where prices go down after certain volume points have been reached. However, margins on Playstation chips remain good due to chip shrinks. G.P.