Question on Anand's Banias article

[IntelUser2000]

If you go to Anand's review about Centrino
here: http://www.anandtech.com/mobile/showdoc.html?i=1800&p=3

Anand says that there was a technology that Intel was afraid to patent, because the danger of competition picking them through patent filing. Can't the competition reverse-engineer the Pentium M processor and find some of the technologies? Why did then Anand say Intel was afraid to patent some of the technologies in Banias? Isn't patent safer then not doing patent? Am I missing something here?

 

[CTho9305]

If you patent it while you're doing the design, the competition can get to work implementing it. If you wait until the product has been released to the market, even if the competition COMPLETELY reverse engineers it, it will still be 2-3 years before they can get a competing product to market.

 

[IntelUser2000]

Really? That's it? Nothing more?

 

[CTho9305]

Originally posted by: IntelUser2000
Really? That's it? Nothing more?

I was just speculating... but it seems reasonable... processors have very long design cycles, so if your competitors have to wait for a release to start copying, they'll be over a year behind you.

 

[Mark R]

Yup, that's it basically.

When you patent something, the day you submit the application the patent office publish your design to the general public. Anyone who is interested can find out what your design does, and more importantly exactly how it does it.

The patent affords you a complete monopoly over your design - if someone sells a product, they must make sure that it doesn't infringe your patent (that's why the patent is published). However, your monopoly only covers the precise design that is stated in your patent. The risk, therefore, is that if someone can produce an equivalent design (with different internal workings) then they are free to use it however they want.

Not patenting your invention means that anyone is free to reverse engineer your design, and build an equivalent design - at which point you are powerless to do anything about it. But you get a head start on your competitors, and that may be all that you care about.

Of course, the other risk with not patenting your invention is that someone else patents it before you release it. As I understand it, until you publish your design (either by patenting, releasing your design, or otherwise) someone else can patent it first, leaving you SOL.

 

[CTho9305]

Originally posted by: Mark R
Of course, the other risk with not patenting your invention is that someone else patents it before you release it. As I understand it, until you publish your design (either by patenting, releasing your design, or otherwise) someone else can patent it first, leaving you SOL.

Unless you can prove prior art, right? If you had a processor a month from tape-out and someone filed a patent for technology in it, you'd have very strong evidence they weren't first.

 

[Mark R]

Originally posted by: CTho9305

Unless you can prove prior art, right? If you had a processor a month from tape-out and someone filed a patent for technology in it, you'd have very strong evidence they weren't first.

It's been a while since I investigated patents (I sort of got to the stage of realising my 'invention' wasn't patentable, and then gave up). And this problem concerned me at the time - I never found a solution to it.

The key point is that an invention can only be 'prior art' if it has been published. From this I conclude that if the design is not available to the public, whether or not it is in use in a confidential setting (e.g. in your scenario above), then it is *not* prior art.

I welcome clarification on this point, if anyone knows the answer.

 

[andyman7]

i dont think it HAS to be published to qualify for prior art, but it is definately the most common way of proving prior art
according to my textbook on IP, it says that there are two main ways of proving prior art: publications and "testimony and statements of skilled practitioners and experts"
it also goes on to say that prior art doesnt just consist of a, b, and c....its like so many other laws that can be interpreted in many different ways and depending on the situation, some thing qualify as prior art and some dont

 

[buleyb]

Yes, the key to proving prior art is outside independent 'expert' testimony supporting it, or officially dated documents. The old stand-by of mailing your design to yourself and never opening it does hold true (because the post office postdate is held to be a legal truth). Outside noteries, financial audits, etc can all help prove prior art.