The Wall Street Journal published a biographical article (free till May 10th) on George Gilder, one of the chief exhalers that inflated the optical bubble of 1999.
I mentioned Gilder last week in my first post on Wintegra (WNTG). He has consistently spoken positively of a network processor company called EZ-Chip and the holding company LanOptics Ltd. (LNOP). His support for the company has been unwavering for years as he continues to predict an NPU revenue ramp for EZ-Chip 6 months into the future. I hope this isn’t because EZ-Chip is a leading sponsor of his annual Telecosm shindig. How about some introspection into why things aren’t working out as he projected?
Posts from this site are occasionally linked to from his walled garden subscription site (I don’t subscribe), particularly anything that mentions Ciena (CIEN).
I’ve been cautious of Gilder because I think he took technology like WDM and purposely dumbed it down to enrapture investors who lacked the technical discipline to really understand what they were buying. Occasionally I get comments that what I write is “too technical” or “Not targeted at a wide audience”. I take them as compliments as that’s exactly what my goal is – to operate at a higher technical level than most. While my audience may be smaller I think it’s a better source of feedback, which is why I write this blog.
Barry Ritholz at the Big Picture summarizes the WSJ article and adds his opinions:
The key tell that his newsletter was a house of cards was that even Gilder himself would note that “He doesn’t do price.” When you consider how utterly absurd that is for an investment newsletter, you can understand how totally out of hand things can get.
This is a simple but important observation. Buying technology equities for technology’s sake without attaching a valuation is a formula for investment disaster (as well proven 6 years ago). His posts on EZ-Chip (that I’ve seen) lack any financial justification, like how big he thinks their TAM is (not what some market research wonk thinks) or what he thinks the company is worth.
I think his service and brand would be significantly enhanced by adding financial rigor, perhaps by adding a partner to his company.
I particularly liked this comment of Gilder’s:
“The typical Gilder subscriber lost all his money and that made it very hard for me to market the newsletter.”
Well, yeah, I would think so.
The guy has always been fun and he is often spot on. Which is somewhat irritating. I think he really nailed the optical revolution that is happening still today. “Everything that can go optical will.” Who really liked PON 2 years ago? Who really liked companies like Infinera’s chances 2 years ago?
But he really hates Access and the regulatory/political aspect of that market segment. It really messes with his ideas ;-) He has no idea of over-subscription etc. ( ie filling the pipe )
Anyways, I think its great that we’ve got a colourful pundit. Rigor be damned, he is great fun.
Andrew … on your side keep up the rigor.. It’s why I read the blog. Also, I can guess why an ex-Vitesse (or any chip-co) guy hates NPs. They are a nightmare – only sell in trial gear and then get designed out by ASICs in volume.
As far as I know EZchip has never sponsored anything. You take Eli Fruchter out to breakfast and you pay. I picked them out of some 50 network processor companies six years ago and they are now number one at 10 gig.
Best,
George Gilder
What about Intel’s IXP2800?
Also, they appear as sponsors for this years and last years Telecosm conference. Didn’t they have to pay for that? Is Amedia paying this year?
LNOP has design wins which you do not mention. gilder does do financials now. He is an man of integrity and vision
Gilder is a genius. I am a subscriber.
He has been predicting about “technologies” and related companies that are going to win as technology paradigms change.
Sure many technologies have taken much longer to succeed in the market than expected. LNOP and WAVX are examples.
Non-engineers (I am a former Intel Santa Clara California engineer) don’t fully understand that Gilder is talking vision (and he’s spot on) and financial numbers in the initial stages are not very helpful.
One cannot understand or invest into such companies based on current quarterly or annul financial numbers.
I think LNOP will win big — wait one year to see the numbers as 90 design wins play out.
-Raj
They certainly play an important role don’t you think? Yours is the type of thinking that helped create the bubble.
Ultimately there must be a financial outcome. Otherwise, trading companies based on technical merits is an expensive business. Trading baseball cards is a much cheaper endeavor but provides the same satisfaction. Technology is only a means to an end.
I have no issue with Ez-Chip. I just feel that NPUs won’t work until several structural changes take place, none of which appear to be happening.
What does not appear to happening? Based on lnop’s recent price and predicition of sales in the coming quarter much appears to be happening
When Cisco builds their next generation Cat6k and Cat4k platform, I’ll get excited. For the sake of the Ez-Chip guys (who I recently got to meet, very nice folks) I would want nothing more to seem them be successful and break the stranglehold that Cisco has on the industry.
So, sure maybe revenue picks up, things get rolling etc. But NPU’s won’t be the $1BB business everyone wants them to be until Cisco rolls over. And that’s not happening (yet).
Well, can you say more about what you picked up at the conference? Your first mention of the LNOP guys was that they were very smart, and now they appear to have been downgraded to “very nice folks.” What’s up?
What about Jnpr, their quarterly revenues are north of $500 million, wouldn’t they like to “beat” csco for awhile, “system design can be hard or EZ”. I wouldn’t mind seeing 1% growing to 2% of those $500 Million quarterly revenues going toward NPU purchases from LNOP.
Guys, I am closing this thread. It is months old. I will be posting my notes from the EZ-Chip presentation tomorrow.
It would be nice to have 1% of the DRAM or Intel CPU market. But it just doesn’t work that way.
Trackbacks / Pingbacks