The invention is a difference engine, a new type of computer that produces a very specific output. The rules governing the difference engine are entirely novel, as is the output.

But the implementation (and monetization) of this ninvention would be as a virtual computer, running on existing chips, as opposed to requiring custom hardware, and running on top of existing operating systems.

  • Would such an invention be patent eligible?

Note: I'm not specifying any region, because I'm interested in answers relating to any given region.

  • By "virtual computer" do you really mean "computer software"? Software simulations of hardware are quite common. For instance you can run iPhone apps in simulation on a Macintosh. The simulation is modeling hardware with a completely different processor architecture. – Eric Shain Nov 29 '17 at 2:01
  • @EricShain but my understanding on software patents is that they need to improve the function of the computing device. This difference engine is isn't concerned with improving the way existing computers function, merely the output of its own processes. This virtual difference engine can be monetized as a product, because the process and output has commercial value, but that's the extent of the use. – DukeZhou Nov 29 '17 at 2:15
  • @DukeZhou It seems then that you're describing a mathematical algorithm, rather than a computer, would that be fair to say? That is, an algorithm in the sense of a set of rules to provide a certain output. Or to approach it from a different angle, since you say it could be done using conventional hardware, does that not imply that you could do it without a computer at all (even if impractically, by hand), such that it's neither software nor hardware, but mathematics? – Maca Nov 29 '17 at 2:34
  • @Maca It's a good point. This virtual difference engine could certainly be built as a physical computer, hardware only, or hardware and software, or be purely mechanical. But this is the case for all computers. I'm having trouble understanding why a microchip, for instance, would be patentable, since any chip can be emulated virtually. It's all just math and logic. – DukeZhou Nov 29 '17 at 15:44
  • 1
    I'm having trouble understanding what a difference engine is, but I suppose if you can fit it into a system claim with a few physical components it has a good chance. It's all about black-boxes instead of software, even though they can be implemented as software. – DonQuiKong Nov 30 '17 at 12:07

I'm going to take a crack at this. To preface, I am not a patent attorney so I am not suggesting my answer is aligned with any legal precedent. I do, however have several granted algorithm based patents. My take is that an algorithm by itself is not patentable as it is deemed abstract. My algorithm based patents all use an algorithm to solve a very specific problem. In my case taking data generated on a real-time PCR thermocycler from a biological sample and interpreting that data to result in a quantitative or qualitative assay determination. Someone might take inspiration from my patent and apply the same algorithm to a different problem and obtain a different patent. Thus, your "virtual difference engine" might be deemed an abstract algorithm whether or not it is implemented on a conventional computer. However, if you have a specific application of that algorithm to solve a practical problem, perhaps it might be patentable.

All this said, the patent attorneys I have used all use very particular claim language when referring to computing devices so I'm sure there is important legal precedent to consider. Also much of this is a moving target and it is possible that patents I acquired years ago may not be considered patentable now.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.