Silverback says the chip is priced at around $75 per port, to be in line with Fibre Channel. Wait a minute here -- wasn't one of the key advantages of iSCSI supposed to be that it's cheaper than FC?
"We think iSCSI will ultimately have higher volume and lower cost structures than Fibre Channel," Kroesen says. In the meantime, he says, "the value proposition of iSCSI is the ability to leverage the existing infrastructure... The total solution and TCO [total cost of ownership] will be substantially less in an iSCSI implementation." Oh.
As far as customers, Silverback maintains it has already lined up some OEMs -- on both the host side and the target side -- which are aiming at the second half of 2003 for their first iSCSI product shipments, according to Kroesen. Unfortunately, Silverback couldn't tell us who these OEMs are yet. Interestingly, much of Silverback's development team, based in Rochester, Minn., hails from IBM Corp. (NYSE: IBM). Could Big Blue be working up a Silverback-based iSCSI system? Neither company would comment.
The TCP Offload Engine (TOE) segment is definitely a crowded one, and Silverback is hoping its performance results will help it stand out. Vendors developing components or cards in this area include Adaptec Inc. (Nasdaq: ADPT), Agilent Technologies Inc. (NYSE: A), Alacritech Inc., Astute Networks Inc., Intel Corp. (Nasdaq: INTC), iReady Corp., Trebia Networks Inc., and QLogic (see Trebia Ousts CEO, QLogic Does Storage Trois, Will Offload Chips Be Uploaded?, Vendors Chip Into IP Storage, Intel Wiggles Its TOE, and IReady Is on Its TOEs).
Silverback, founded in July 2000 and based in Campbell, Calif., has raised $17.3 million to date, and it's planning to announce its next round of funding next week. Investors include Gemini Israel Fund, JP Morgan Partners, Newbury Ventures, Pitango Venture Capital, and U.S. Trust Private Equity Capital (see Silverback Emerges From the Mist).