Silicon Investor (SI) -- The First Internet Community

STOCKTALK

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.

Technology Stocks : Son of SAN - Storage Networking Technologies -- Ignore unavailable to you. Want to Upgrade?

To: w2j2 who wrote (4423)	2/18/2002 12:22:34 PM
From: Logain Ablar	Respond to of 4808

Walter: Its not the customers but the OEM's that have to be concerned.

To: w2j2 who wrote (4423)	2/19/2002 1:41:29 AM
From: Gus	Read Replies (6) \| Respond to of 4808

The suit does include the mid-range 3800. That's the key target for McDATA since they want to displace Brocade at the OEMs at that product level. That's why they want to disable the frame filtering feature at the ASIC level and set BRCD back at least 1 year since the new ASIC and switches will be re-qualifying right alongside's Inrange's new ASIC and switches. MCDTA stretched to introduce 4 new switches in 4 straight quarters last year and managed to get only 8% of the fabric switch market or around $40M-$50M while maintaining its control of the director-switch market. It could have been better but an entangled Brocade allows them to jack that number up to the $100M to $200M right away. The first generation 12000 is a paper tiger. As many here already knew, it is just a bigger version of the poorly-received 6400 -- six 16-port switches prewired in a chassis. The 12000 is two 6400s prewired with some ornaments around the duct tape. That's a no-brainer decision between a 5th generation single stage director versus an unstable 1st generation multi-stage switch in a latency-sensitive environment. The 3800 started shipping in volume only in the recently completed January quarter so the timing is actually quite good. BRCD actually started talking about frame filtering early last year before McDATA's '236 patent was issued on May 15, 2001. Talking is not shipping, however. The 236 patent was filed on January 12, 1999. I believe that both McDATA and Brocade already knew that this showdown was coming since they probably butted heads at the USPTO with their competing frame filtering patent applications in the patent prosecution process. McDATA won so many times not just in frame-filtering primarily because it has a deeper foundation of patents. Brocade probably decide to gamble on a very loud jawbone strategy on the premise that mindshare will result in a de facto standard that will force the OEMs hands!?!?! To appreciate the very dynamic nature of these high-stakes patent strategies, consider that IBM's patent royalty business went from less than $90M a year in 1990 to more than $1B a year in 1999 before it decided to take what was then a pure royalty collection business to another level by ALSO using its vast patent portfolio to induce different types of alliances including the $3B supplier/cross-licensing agreement with EMC and the $15B supplier agreement with Dell in 1999. Lucent's recent actions on the security component of iSCSI may suggest that they are taking that same page out of the IBM turnaround playbook. IBM's patent business is currently somewhere in the $1.5B a year range. Brocade was interested in buying McDATA from EMC because of its patent portfolio and McDATA's very large ESCON installed base. The 14 or 15 patents that McDATA bought from HWP in 1997 were conceived when the Fibre Channel 80/20 open standard was still being baked so even then they represented an important foundation for other patents. That's the main reason why McDATA continues to rack up all those new patents at Brocade's expense despite its tactical decision to use Brocade ASICs to accelerate the validation of the 1Gbps Fibre Channel SAN market. It has the better foundation of patents. The closest parallel is probably the Unocal patent covering the designer fuel required by California regulators. Unocal was able to successfully prosecute its patent applications around the emerging standard so now everybody else has to pay them a royalty per gallon! In the case of McDATA, it is not going for the always unpopular taxman strategy since they already have the shipping director switches and fabric switches products to make the blocking strategy more rewarding. McDATA's patent strategy was formed during the strategizing that went on at EMC in the in the mid-90s after its patent battles with StorageTek in the early 90s. McDATA acquired HWP's CNO unit (14 or 15 patents) for around $5M cash in early 1997 with the approval of EMC which had previously acquired the patent portfolio of MTI Technology for around $30M over 5 years in early 1996. MTI got an additional $10M in 1999 as a result of the IBM and EMC cross-licensing agreement. MTI is a company founded by Ray Noorda of Novell fame. It did some pioneering work on networked storage that EMC is using as a foundation for newer patents. That article on iSCSI, for example, contains a reference to a 1997 EMC patent that could possibly cover key parts of the iSCSI standard. Among its many references or prior art is an MTI patent that was filed in 1990 and issued in 1995, presumably part of the 29 patents that EMC purchased in 1996. EMC also has another patent issued to Erez Ofer, its new head of software, covering the use of SCSI over the network and they probably have more in the pipeline. EMC co-chairs the iSCSI workgroup with Lucent and it also heads SNIA so it's going to be very interesting to see how EMC is going to play its iSCSI patents especially in multi-protocol (FC/iSCSI) storage networks. Remember that EMC already has a license to McDATA's patents and it also has a no-sue-for-5-years cross-licensing agreement with IBM, which owns the key patents to ESCON/FICON and is working just as hard as EMC and many others in filing patent applications around the emerging iSCSI standard. Presumably, all these OEMs want to commercialize their patent portfolio like IBM.