Oliver, in six months, Sun's outlook will still depend on its earning
outcome. To be practical, I do not think Java will significant effect Sun's earning in six months. I think Sun will generally do fine. According to Sun's style, if they made more money, they would put more money on R&D (it's good thing for Java) and made EPS only 2~3 cents higher than estimate. Last quarter the result caught me by surprise, made my prediction well off the target, Sun had enough room to make EPS meet the estimate. Listening the messages from Sun, they would spend less on R&D if they have known the Sept. would be weaker than they thought (July, Aug. are always soft), I don't expect Sun will let it happen again. One uncertain factor is how the sentiment will change about Sun and Java from the majority investors, those people know nothing about Java, UNIX and NT, they tend to absorb whatever MSFT says, if these people understand a little bit Java's power and future, I do not expect Sun's PE to stay on the current level.
A lot of people try to understand how Sun will make money for Java,
the current situation, Sun mainly get Java money from
1)Java license fees. There are about 116 Java licensees around the world, expecting these numbers will increase, from the revealed MSFT's contract with Sun, MSFT pays more than 10 million (14 or 18 millions ?) to Sun during 5 years, about 3 millions a year. I don't expect other licensees pay as much as MSFT's, but these are pretty good numbers.
2)A few java products from Sun. Java WorkShop, Java Web Server, JavaOS, Java studio etc..
Java WorkShop: retail price is $99, HP bundles JWS with its HP-UX,
so Sun can eat some lunches here.
Java Web Server: retail price is $295 ($95 without SSL), a pretty cool product, sales will increase with time.
JavaOS: With the JavaOS products gradually appearing on the market (do
you see the news Japanese companies and LG will ship NC will JavaOS 1.1 soon), Sun will get its big meals here.
Java Studio: retail price is $76. released last month, target on both java developers and no developers, you do not need to write a single java code to produce a java products, you will probably see it on your local retail stores next month, it has potential to be a big hit.
3)Java service.
Don't forget Java also made more people know about Sun(become a brand name), the thin client model based on java will boost the sale Sun's servers.
Next year, Visa will ship Java smart cards and pay Sun loyalties per JVM copy, So will the Java smart phone for North Telecom, TXN will join these ranks with its DSP chips being implemented with JVM, there
is no reason why we would not see the loyalties for Lucent's inferno.
I stop doing the math counting here and let you do the rest, otherwise it would be a long list.
My personal opinion the JavaChips will be one of the greatest potential area for Sun, imagining if there are enough Java applications and if Sun could build JavaChips around 600~1000 MHz (64 bits) level in 3~4 years, what reactions can you expect from Intel?(VBG).
Currently Sun is in partnership with LG Semicon, Sun designs(probably CO-design in some areas) and LG Semicon manufactures, Sun gains exclusive rights to sell these cheaper JavaChips and LG Semicon gains priority to license future generations of Java(TM) technology from Sun. This market is too huge to predict, expecting a lot of companies
will join the race (even Intel, the japaness companies also licensed PicoJava from Sun for their JavaChips, we will see them next year), it would be turly good news if we could see JavaChip from LG Semicon next month.
Be patient.
Cheers
Gordon Shen |
