SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Strategies & Market Trends : Gorilla and King Portfolio Candidates

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Mike Buckley who wrote (18477)2/23/2000 11:19:00 PM
From: nick chacos  Read Replies (3) of 54805
 
I'm a long-time lurker of this thread who has also followed Ancor since '96. Having spent a bit of time in some ancor-specific chatrooms for the past year, I have come across a few scattered items which you might wish to consider in your evaluation of Ancor. THe problem is that I am a pseudotechie- know just enough to misunderstand most of what I hear and read about FC and SANs Disclaimer- I have a long position in Ancor, missed getting into Brocade at the ipo and have been afraid that an immediate and devastating correction would ensue the minute I did, so haven't. That said, there are several points I'd like to make which might help make clearer the relationship between Ancor and Brocade.

First, in response to your specific question about 64 vs 16 port switches. It is my understanding that as the complexity of the network gets greater, that which controls various functions requires more sophistication. Whereas in small SANs a server with an overseeing program might do the job, in these larger systems which have hundreds and up to potentially tens of thousands of objects attached to the system, a more efficient controller is required. The 64 and greater port switches are referred to as directors and, again, it is my understanding that they do much more than just switch network traffic. They monitor and optimize the traffic. Otherwise it would be hard to explain why a 16 port switch sells for about $1,000 per port while a 64 port switch sell for much more ( I believe in the $3-4,500/port range. Although I have read TRFM several times, particularly to get an insight into the Brocade/Ancor battle, I still have a difficult time distinguishing between continuous and discontinuous innovations in some cases. The above is a case in point- does the added complexity of a director make it a discontinuous improvement or just a more advanced, complex continuous improvement.

Some other disparate items I have heard or read, making them 2nd hand and therefore worth about what this costs.

1-Either in a phone conversation or at a recent analyst meeting,I don't remember which, the ceo of Ancor mentioned that he believes that the architecture of the Brocade switches might preclude them from being able to scale to larger switches than 16-32 ports.

2-Brocade uses an ASIC they buy from LSI which holds 2 ports. Ancor is about to introduce a 16-port ASIC in the "very near" future. My understanding is that about 1/3 of the cost of a switch resides in the ASIC. By placing a 16 port ASIC into a switch, a major cost reduction occurs. (Assume a 16 port switch sells for about $5,000. At around a 50% margin it costs about $2,500 to build. 1/3 is about $800 divided by 8 ASICs is about $100 per ASIC. Even if the new 16-port ASIC costs double ($200), the overall switch cost is now about $1,800 vs $2,400. If Brocade cannot increase their port density, they should be subject to potentially severe margin pressure.

3-While I agree that this is probably a royalty game, as I understand it, since there is probably not any proprietary standard, since standards are being developed by a committee, Ancor is working on an Infiniband switching device. Again, my understanding is that Intel will permit them to keep whatever IP they might create in the development process. At another industry meeting, an Intel engineer conceded that there is a possibility that Ancor could end up on the motherboard in a chip(s) through which all I/O would pass. Such an I/O chip would end up in every single item which would ever hope to be linked to the future Infiniband network. This suggests to me the possibility of proprietary standards which could move them to the primate arena.

4- Brocade had revs in '99 of almost $69MM. If their claim to 80% of the market is true, that puts the total market at around $86MM. The FC switch market is currently predicted to be $1,700MM in 2002 and $3,000MM in 2003, antough IBC is supposed to be about to reevaluate those estimates momentarily. I would contend that current market share is somewhat irrelevant in that context, since Ancor is to have a more dramatic sales ramp once Sun starts selling switches. Keep in mind, for example, that for Sun to be able to exercise all of the Ancor warrants (1.5 MM) to which they might be entitled, they need to spend at least $100MM on Ancor by 9/01.

5- The question asked about replacing switches between vendors is related to what they call interoperability. It is my understanding that currently, in a number of installations, Ancor switches and Brocade switches have been able to be replaced successfully, as long as they don't have to communicate with each other (a mixed switch network).

6- My understanding is that the McData/Brocade contract expired last December, although I have no further info on whether it was extended or not.

7- I believe that the Sun contract with Ancor will be exclusive due to the turnkey nature of the FC-SAN system that Sun will be announcing soon. My understanding is that Sun is about 15-20% of the FC SAN space, and is looking to go head-to-head with EMC (is this another gorilla war?). Ancor management has stated in the past that they expect to to grab 30-50% of the market. Ashok Kumar at Piper Jaffray, an analyst who seems to be very educated about this market segment believes that large SAN's will account for 50% or more of the switch market in a couple of years. This again seems to suggest that there will be a couple of roughly equal princes fighing it out.

8- Rumor is that Ancor switches have some performance specifications which make them much more suitable for larger SAN's and that Sun will soon make a public statement to this effect, although until there is an unbiased direct comparison at a qualified lab with published specs, we can only guess at this. If this is true, we might see Brocade own the smaller SAN market addressed through OEM's such as Compaq and Dell, whereas Ancor could lay claim to the larger enterprise SAN's through their OEM's with Sun, MTIC, Hitachi Data Systems, etc--maybe something analogous to desktop computers vs mini or mainframes.

Sorry for the longwinded post. I've been waiting a while for you all to discuss the Ancor/Brocade story in more than a superficial way so I could throw in my 2 or 3 cents worth.
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext