TechWorld reports that Graphics Properties Holdings, the post-bankruptcy incarnation of Silicon Graphics, is suing mobile device makers for infringement of a graphics-related patent. Smartphone manufacturers are no strangers to patent lawsuits, of course, but I think that this latest salvo is a good illustration of how the dual regimes of patent law and rent-seeking litigation do nothing but divert resources from actual innovation.
The patent in question, number 8,144,158, describes a “rendering circuit,” the novelty of which appears to lie in the use of floating-point rather than fixed-point data to create images, thereby allowing for more complexity and accuracy in the resulting images.
As an absolute layman in the field of graphics technology, all I want to point out in the actual patent is this: the “Background Art” section of the patent description points out that the advantages of operating with floating-point data are well-known, and that a large part of the reason why floating-point data was not previously used lay in hardware limitations. A good question, then (and one that the manufacturer’s lawyers are probably studying like crazy), is whether a graphics person would think it obvious to use floating-point data in the way the patent describes, given the existence of suitable hardware. If it would have been obvious, then the patent should be invalidated under Section 103 of the Patent Act. If not, then Graphics Properties, under the law, and assuming nothing else is wrong with the patent, is entitled to damages and/or possibly an injunction.
Either way there is a clear policy problem – the fact that low-quality patents enter the system and cause wildly objectionable allocations of resources (either through unjustified licensing fees paid to avoid litigation, or the litigation costs incurred in correcting the patent office’s mistake), or the fact that a company that doesn’t practice or in any way promulgate its invention has rights against a party that does give the public the benefit of the invention. The second problem would be mitigated somewhat, at least morally, if so-called “innocent infringers” were protected – but they aren’t.
Timothy B. Lee points briefly but eloquently to the moral argument against patents, and more importantly, sets that moral argument atop a strong empirical showing that figuring out the patent morass is essentially impossible in certain fields. Graphics Properties might have an enforceable patent – but Apple, Samsung, HTC, and others put the invention in your pocket. Enforcing the patent, if it’s valid, doesn’t promote progress, or magically force companies to innovate more – it removes that invention from the field, it forces practicing companies to spend money to license or attack the patent, or “innovate around” it – or simply stop using that presumably beneficial invention. Mark Lemley posits that innovation occurs despite patent law – because companies simply ignore it. Here, patent law is an agent of stagnation – it allows the ghosts of companies past to haunt companies that actually do things. As a society, we probably should not be empowering ghosts.
