Tom, how about sharing your wisdom on this company with us? If this is just a big cap vs small cap thing, I hope LSI can find two more Symbios. If VLSI knows what's ahead of them, they should just merge into LSI. LSI is already $2B, another $1B or so and it will be big cap too.
Personally, I think LSI will become big cap after the next round and they can show they are in the right areas. I look at them as STM without the flash. But for the stock to move, it would take 2Qs of profits and sales increases. If 4th Q is the first one, then the move upwards may be 6 months away.
Just hope that I have more cash now.
Another thought: IDTI just wrote off $120 worth of 0.35um fab because they could not fill it. Here's LSI running at 90% capacity last Q with 0.35 or bigger. Tell you something about their products, doesn't it. Specifically IDTI is trying to make 0.35um gelly beans that any design house can supply through a foundry and LSI's 0.35um is making design-in stuff that have minimum competition.
Somewhere down the road, there will be a shortage of sub-0.25um capacity. A lot of guys who built 0.35um, like IDTI/Vanguard/LG/Kuotech/Seimens etc. etc., did not recover their investment to go sub 0.25um. LSI is just going at it slowly but surely. One can make an argument that they should farm out their 0.35um stuff because there is excess 0.35um foundry capacity and then convert their 0.35um fab to sub-0.25um. However, my guess is their 0.35um fab may be too old to convert and in the long run it will be much more efficient to build Gresham. After the next upturn, they will be competing with TSMC/UMC/IBM and the likes and fab efficient will really count.
I know that someone here said that they are going to have competition from 6 - 7 person design houses with $1 - 2M CAD budget. Some chips yes, a whole lot of other chips no. I know somebody who is trying to design 0.25um 32 bit microcontroller. Supposed to tape out beginning of April. He's still at it. And he works for a very big company with a lot more than $1M - $2M CAD budget. Turns out a lot of the simulation tools are having trouble simulating surface and perpherial effect of 0.25um vs bulk effects of larger ums. Personally I think they should have gone with SrongArm or something like that. Trying to do a 32 bit microtroller ground up is also like re-inventing the wheel. Can be done but will probably lose out on time to market.
I can tell that the PC thing was going well from DRAM pricing since early July. But Oracle saying they are going to have a good Q, perhaps the infracture building slowdown has rebounded. Well, time will tell. If LSI's news leaks are as efficient as the last two months, the movement of the stock the next two weeks will tell us what this Q is going to be like.
Understand your sentiments regarding this stock though. But no need to throw daggers at someone who is still thinking like you when you bought your stocks at the higher price.
patrick
|
