IBM countersues Neon over zPrime accelerator
More lawyers feast on the mainframe largesse
With mainframe revenues off sharply and likely to be so until the System z11 mainframes ship much later this year, IBM can ill afford to look the other way as Neon Software peddles its zPrime tool for offloading mainframe workloads to much cheaper specialty engines on Big Blue's mainframes.
So it has called out the lawyers and countersued Neon Software, which sued Big Blue for anticompetitive practices back in December.
IBM tried a soft approach to zPrime at first, warning customers interested in zPrime they may be in violation of their software licensing agreements after the tool was announced at the end of June last year. The zPrime tool was starting to get traction in the fall, when zPrime 1.2 came out in early November. At that time, Neon had 14 customers using zPrime in production, 50 had tested it, and another 200 were getting ready to do evaluations, and 1,500 had contacted the company to get more information.
This is a lot of interest in a mainframe base that has somewhere between 6,000 and 7,000 customers worldwide, according to estimates made by Neon. And so, Neon decided that IBM's FUD had gone on long enough and it sued Big Blue in the US District Court for the Western District of Texas, located in Austin, for unfair business practices and anticompetitive behavior.
Neon's essential argument in its lawsuit is that IBM's contracts with mainframe customers have no provisions that restrict the use of zPrime to move various jobs running on the z/OS operating system and its expensive regular engines to the System z Application Assist Processor (zAAP) specialty engine, which is used to run offloaded Java and XML processing, or to the System z Integrated Information Processor (zIIP), which accelerates DB2 databases by running offloaded DB2 functions. The barn door was open, and the horse was out, and IBM is trying to seal off the exit.
IBM's answer to the suit and counterclaim, filed on 27 January in the same Austin court, cut right to the chase scene in the first paragraph. "This case is about Neon's attempted hijacking of IBM's intellectual property. Neon's business model expressly depends upon Neon inducing IBM’s customers to violate their agreements with IBM.
"In this respect, it is no different than that of a crafty technician who promises, for a fee, to rig your cable box so you can watch premium TV channels without paying the cable company. Even if it could be accomplished technically, it is neither lawful nor ethical."
IBM's countersuit alleges that Neon Software's business violates both federal and state law, that Neon has tortuously interfered with contracts between IBM and its customers, and that Neon has breached its own agreements with Big Blue. The filing goes on to suggest that Neon Software has violated IBM's copyrights because the use of zPrime makes unauthorized copies of IBM's mainframe systems software.
And, here's one you don't see often in the mainframe racket: IBM alleges that the zPrime tool violates the Digital Millennium Copyright Act "by trafficking in software designed to facilitate infringement and circumvent technological measures in IBM's mainframe computer systems." Finally, IBM's lawyers say that Neon Software violates the Lanham Act by taking money owed to Big Blue for software licenses.
The new twist in the suit is that IBM is claiming that zPrime is not just using the switch-to and dispatcher at the heart of the z/OS operating system to redirect z/OS jobs illegally from general purpose engines to zIIPs and zAAPs - something it was designed to do by IBM, but for very specific workloads, and not the ones zPrime is moving over, according to the suit - but is also being used to activate general purpose mainframe capacity. No one El Reg has ever talked to from Neon Software has ever said zPrime could do this or suggested that this was a goal of the tool.
According to the countersuit, the switch-to and dispatcher software is a block IBM put into mainframes that restricts the copying of mainframe systems software from the caches of general purpose mainframe engines to the cache memories on the zIIPs and zAAPs; these access controls also prevent copied programs from accessing services back on the general purpose mainframes.
IBM is seeking compensatory damages, punitive (triple) damages, and lost license fee revenue recovery in its countersuit, and says the suit is "not about stifling innovation" or a "purported 'monopoly'," as Neon Software has called it. The lawyers can, and may, argue about who is stifling what, but it is completely asinine to say that IBM does not have a monopoly on mainframes that run its systems software.
IBM says that Neon's lawsuit has no merits and should be dismissed, and is seeking an injunction to stop Neon from selling the zPrime tool. ®
But many people says that IBM has in practice, monopoly on Mainframe market. You mean this is not true? Which other Mainframe vendors are out there, selling Mainframes?
I think this is the first time I agree with you! :o) It sounds reasonable that IBM will buy this company (as before) and then continue their monopoly.
Lamborghinis maybe, but you could probably trademark Lamgorhinis yourself.
The monopoly on 360/370 family mainframes is, as Henry Wertz points out, not enforced by IBM. The plug compatible manufacturers simply left the market. It was as though everybody but Ford stopped making cars.
I've heard it expressed that the 370 was the first open computer platform. IBM published the 370 Principals of Operations (POO), which documented the complete workings of the CPU, channel structure and I/O processors that enabled the whole PCM market, exploited by Amdahl, Fujitsu and Hitachi, to name just a few.
Yes, IBM has a monopoly on mainframes
but not on computers. Like Audi has a monopoly on Lamgorhinis but not on supercars.