IBM Wins Most Patents In a Single Year For 2008

eldavojohn writes "You might have heard or felt that there is little left to patent these days but IBM begs to differ. They came in at over four thousand for the year of 2008. Now, this isn't a good metric to measure success or progress but for those of you who like to keep track: 'IBM said it earned 4,186 U.S. patents in 2008, more than triple the number of patents earned by rival Hewlett-Packard. Microsoft Corp earned 2,030 patents, while Intel Corp had 1,776 and Hewlett-Packard 1,424, according to the report, which compiled data from the U.S. Patent and Trademark Office. Korean electronics giant Samsung Electronics had the second-highest number of patents at 3,515.' You can find the original source of this study here as well as 2007's data and even 2006's data."

Comments

Weird

[Idiomatick]

IBM/Samsung make THINGS so that makes sense. How did MS get so many? They don't make any THING aside from xbox. You think the divide would be alot bigger than it is.

Re:Weird

[El Lobo]

1) They make THINGS as well: mice, keyboards, XBoxes, etc.

2) You don't need to make THINGS to get a pattent. You can patent almost anything these days. Software algorithms are especially important for such companies like Microsoft if they don't want to be eaten by patent trolls, which is so common these days.

Re:Weird

[GameMaster]

While it is true that, at least, some of MS's patents are for hardware, most of them are, probably, software related. As far as patent trolls are concerned, I think you have it backwards. There would be noe patent trolls to defend against if we didn't allow software patents in the first place. Software is, simply, a math notation and we used to have the brains to rule that people couldn't patent math.

Re:Weird

[mdf356]

The Federal Circuit recently ruled en banc that if your patent doesn't involve specific physical things, it's not actually patentable. So at the moment any software not tied to a specific hardware device is invalid again.

What Congress does with this will be interesting to see. They were taking up patent reform several times in the last session. Probably half or more of IBM's patents are on software methods.

Re:Weird

[GigaplexNZ]

There would be noe (sic) patent trolls to defend against if we didn't allow software patents in the first place.

Hardware patent trolls would cease to exist if software patents were removed?! Who knew?

On a more serious note, you can't really blame MS for this. They didn't decide software was patentable (cue conspiracy nuts), but since it is, they have to defend against the trolls.

Re:Weird

[Arthur Grumbine]

You can patent almost anything these days

Not anymore - "Almost Anything" was IBM's 3,621st patent of last year...

Re:Weird

[mabinogi]

No he has it perfectly correct.
Your statement is also correct, but they can both be true.

Yes, if there were no software patents then there would be no software patent trolls. There are, however, software patents, so Microsoft and other companies have to take out software patents to protect themselves from the resulting trolls.

Re:Weird

[Anpheus]

Sure, it involves a specific physical processor, and a method for achieving an operation more quickly using a particular set of physical electronic messages, "opcodes" if you will, sent to the processor.

That's what always confused me about people saying software patents aren't physical things. I know the "magic smoke" is a popular myth but surely these people don't think their computers run on fairy dust and little elves pushing bits around?

IMO, with judicious use of prior art and a blanket ban on rewriting Patent X so that it's now Patent X On The Internet or Patent X With A Cellphone, I would have no problem with patents.

Let's be honest, whoever manages to patent some of the first algorithms for quantum computers really does deserve some short-term exclusivity for their methodology. Stuff like Shor's Algorithm, that stuff blows me away.

Re:Weird

[Anonymous Coward]

Bullshit...IBM's patent collection is an albatross of opportunistic software patents.

Patent's you have to have big balls and expensive lawyers to even bother submitting.

Re:Weird

[Anonymous Coward]

doesn't IBM have big blue balls?

Re:Weird

[bob.appleyard]

What proection are they against trolls, though? They don't infringe any patents, so what use is developing a defensive portfolio against them?

Re:Weird

[tricore]

Why do they? Tons of people are working on it, and no doubt someone will do it. There's plenty of push without the additional payoff of exclusivity. I'm a professional programmer, I should add. The problem is, almost everything in software IS obvious about 1 year later. JPEG is totally "duh" for example, take the furrier transform, drop a few bits off all frequencies, convert back. Doing this drops the pieces of the image you are least likely to notice, it's some trivial mathematics, and anyone with rudimentary knowledge in the field could easily come up with it. I would be very hard pressed to find an algorithm that someone wouldn't likely come up with when working on the problem domain. 3DES, blowfish, SHA1, and a few of those algorithms, simply because they are fundamentally arbitrary. Note, I was specific, RSA (which was patented) is pretty obvious for example. Imagine if someone had patented UnionFind, BinarySearch, MedianFind, Quicksort, MergeSort, B-trees, Red-Black trees, scape-goat Trees, heaps, treeps, etc. ALL of these algorithms are fairly obvious once you start digging into that problem domain. The people who came up with them were brilliant, no question, but when you try and solve a certain type of problem you write a certain type of algorithm. I accidentally invented UnionFind once before I knew it existed. The only purpose of patents is to encourage people to invent things. I actually suspect the entire concept is antiquated now, but I don't know other domains well enough to be sure. In software it simply isn't necessary, people invent things daily, and any real product is sufficiently covered by copyright. Patenting software is like patenting a literary method or trope. "You can't write a novel where an airplane gets hit by lightning, that's MY idea!" or "You just used alliteration with a's, I patented that for the purpose of being hilarious, sorry"... what?

Re:Weird

[Anpheus]

And I could respond with an enormous, a truly mind-bogglingly large list of so-called "physical things" that were obvious in hindsight.

Hindsight bias, it's a tendency for all of us to think, "DUH! I should have thought of that!" All a number of mechanical devices have been patented over the years and of course there have been countless occasions when those devices were considered obvious... after the fact.

I read

[popeye44]

Somewhere that they were going to freely allow use on around 3000 of them? That's pretty righteous.

Re:I read

[jellomizer]

So patents are against the GPL except from the ones by IBM. Wow What is IBM pay to get FSF Support.

Re:I read

[Anonymous Coward]

The 3000 in the article is technical innovations, not patents - those are the sorts of things companies don't consider patentable, but publish them in case some other company might try to patent it - that way they can't be sued for the innovations use later on.

Re:I read

[jd]

They actually released some time back (2 years or so) over 500 patents for Open Source projects. Not sure how many have actually been used by Open Source folks, as most of the patents seemed to be hardware-related and, frankly, Open Cores isn't nearly as popular as Sourceforge. But a good effort none-the-less, in that they did at least make an effort. The 3,000, as others have noted, aren't patents but rather unpatentable ideas they can't market but can get cheap publicity with.

Re:I read

[thesp]

Exactly. In many cases, patent offices only conduct prior art searches on previously filed and published applications. Getting exactly what it is you do into prior-art-space is a useful defence against another person later alleging patent infringement on a subsequently-filed patent.

Doesn't maintaining patents cost money?

[blind biker]

I was talking the other day with colleagues that were pondering whether to patent an idea they had, or not. The counter-argument was that it cost substantial money to just maintain a patent - and the figures mentioned were several thousands of USD a year. IBM acquired 4000 patents in a single year. That makes me wonder how many patents are they maintaining nowadays? And if they pay, say, $5000 to maintain each, isn't that a substantial financial burden?

And if it isn't - shouldn't it be?

Re:Doesn't maintaining patents cost money?

[compro01]

The maintenance fee for a patent increases as the patent ages.

There's the filing, search, and examination fees when you apply for it (fees depends on what type of patent it is), then the issue fee if/when it is approved (again, variable depending on type), and reissue fees if you're reapplying when your previous application was rejected (variable), then maintenance fees due at 3.5 years ($980), 7.5 years ($2480), and 11.5 years ($4110) after the patent is granted.

The most of the fees are halved for "small entities".

If I'm adding correctly, the total cost of filing, acquiring, and maintaining a utility patent (assuming it gets through on the first try) is about $10,170 and a design patent is about $8,890.

Re:Doesn't maintaining patents cost money?

[Zakabog]

...And if they pay, say, $5000 to maintain each, isn't that a substantial financial burden?

And if it isn't - shouldn't it be?

That would probably hurt more than it would help. If we make maintaining patents cost prohibitive to a company like IBM, who does that benefit? If you've got a really good idea it shouldn't be too expensive to patent it, otherwise you're completely defeating the purpose of the patent system. Yes I know there are flaws in the system, but making it more expensive to patent things only helps the large companies that can afford it, not the small companies that might have some big ideas.

Re:Doesn't maintaining patents cost money?

[moderatorrater]

Let's not forget that they pay their employees (my uncle, in this case) nearly $1000 per patent they get granted for IBM. That's a cost of $4 million just to the employee that does it.

Re:Doesn't maintaining patents cost money?

[mdf356]

Don't forget the money for a patent attorney to draft claims, deal with non-final rejections, etc. The attorney's fees are around $20k per patent.

Re:Doesn't maintaining patents cost money?

[mdf356]

Last year it was $750 for each application filed, and another $500 on issue. If there are more than 4 inventors the $3000 and $2000 is split evenly.

Re:Doesn't maintaining patents cost money?

[Coldfusion97]

IBM is more than willing to pay it. When I worked there the company was fond of pointing out that they make around a billion dollars a year just from their patent portfolio.

Re:Doesn't maintaining patents cost money?

[Prof.Phreak]

If we make maintaining patents cost prohibitive to a company like IBM, who does that benefit?

How about letting the patent holder determine how much it is worth, and then -they- have to pay a... property tax (or something) on that every year to maintain it. Maybe waive the fee for the first year or two. The catch is that if they say it's worth $1billion, they have to pay tax on -that-. Yet if they say it's worth $1, their tax goes away, but any future lawsuits they initiate concerning the patent would be similarly worthless (ie: ``you're infringing on my $1 patent, pay up!'').

That would eliminate corps sitting on patents and not doing anything with them...

Re:Doesn't maintaining patents cost money?

[compro01]

I would presume with the amount of patents said companies deal with, they'd just have their own patent attorneys in house on salary.

Re:Doesn't maintaining patents cost money?

[mdf356]

All companies I'm aware of use outside counsel, for a simple reason: someone to blame (and sue) if something bad happens.

If a deadline is missed on a patent with significant business value, damages can be recovered. If an attorney has been suborned and deliberately sabotages the claims, damages can be recovered. You can't get damages back from yourself if an employee is at fault.

Re:Doesn't maintaining patents cost money?

[Anonymous Coward]

Yup, but how many companies do you know that have that many patents? with 20k per patent it would be 64 million just to patent lawyers per year, only counting the ones submitted but not counting anything extra. So I think we can pretty confident they don't pay that much at least. If they have lawyers outside the company I would think that they have a good deal for significantly reduced price tags.

Re:Doesn't maintaining patents cost money?

[swillden]

I'm sure IBM spends lots on patent maintenance, but they don't mind -- IBM makes over $1B per year in patent licensing revenues. Also, the figures you're talking about probably assume paying some outside lawyer an hourly fee, while IBM obviously has it's own staff of in-house patent attorneys.

Disclaimer: I work for IBM but don't speak for them, and my only interaction with the IBM patent attorneys has been when they refused my proposed patent because it was insufficiently novel.

Re:Doesn't maintaining patents cost money?

[Anonymous Coward]

It's not $20k/patent/year. A typical application spends years stuck in the PTO before anything possibly comes of it.

Re:Doesn't maintaining patents cost money?

[jlarocco]

And if they pay, say, $5000 to maintain each, isn't that a substantial financial burden?

Not really. The entire point of a patent is to allow a person or company to profit from their invention without other people ripping it off without doing their own research. The patent fees are almost like a benchmark. If the invention wouldn't be able to recover the cost of obtaining the patent it's really not even worth patenting. If they can't recover the patent fees, they're definitely not going to recover the researcher's salaries.

On the other hand, a company with as much money IBM probably patents a lot of things "just in case" it becomes useful to them later.

Re:Doesn't maintaining patents cost money?

[bh_doc]

Almost sounds like a kind of insurance.

Re:Doesn't maintaining patents cost money?

[John Hasler]

> The entire point of a patent is to allow a person or company to profit from their
> invention without other people ripping it off without doing their own research.

Not in the US. According to the Constitution the purpose of patehts is to promote progress in science and the useful arts:

"To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries;"

The courts have interpreted this as meaning that the point is to benefit society, not to protect an intrinsic right of an inventor to control his invention.

Re:Doesn't maintaining patents cost money?

[Draek]

If you've got a really good idea it shouldn't be too expensive to patent it, otherwise you're completely defeating the purpose of the patent system.

The current patent system is defeating the purpose of the patent system. What matters now is how to minimize future damages, and reducing the amount of patents being awarded, by *any* means, is a good thing in my book even if most of those that do pass end up on the hands of the Big Guys.

Plus, I ask you: where do you think the patent trolls' portfolios come from? hint: they don't buy them from Microsoft or Intel.

Re:Doesn't maintaining patents cost money?

[jlarocco]

Not in the US. According to the Constitution the purpose of patehts is to promote progress in science and the useful arts:

Obviously the long term goal is the overall progress of society. Equally obvious is that patents help achieve that goal by making it easier for inventors to profit from inventions without other people ripping off their idea.

Re:Doesn't maintaining patents cost money?

[TheLink]

Thing is the current patent system mostly rewards those who aren't really innovative!

The ones who are really innovative are like the Douglas Engelbart sort - invent something in the 1960s, then only in the 1980s and later people start to understand the use of _some_ of it.

Whereas people who invent "break eggs to make an omelette" get patents all the time.

As the technological and scientific fields avance and become more specialized it becomes harder and less likely for a patent examiner to be able to judge an application's worthiness within a reasonable amount of time.

In my opinion it is better for there to be Prizes for Innovation, which can be awarded after an invention is proven. There is no monopoly granted except that "you are the first recognized inventor of X". You get your reward from the prize money and the recognition.

It is easier to judge based on hindsight. You could have two classes of prizes (but many prizes) - the first class is awarded by experts in the field and the second awarded by members of the public.

BTW: If the pace of progress is supposed to be increasing (and marketing and distribution is supposed to be better), the protection terms for patents (and copyrights) should be getting shorter and not longer :).

Re:Doesn't maintaining patents cost money?

[MahJongKong]

Sounds good in theory but it's impossible to implement: nobody could get even close a decent estimate.

Re:Doesn't maintaining patents cost money?

[Anonymous Coward]

With respect, patent attorneys don't refuse or accept patents. That's the Patent Office's job. On the other hand, it might have been their professional opinion that your proposal would have required more cost and effort to get to grant than the likely worth of very narrow patent that would have resulted.

Re:Doesn't maintaining patents cost money?

[Jurily]

but making it more expensive to patent things only helps the large companies that can afford it, not the small companies that might have some big ideas.

Exactly. How about, say, 0,1% of last year's profit per year per patent?

Re:Doesn't maintaining patents cost money?

[notaspy]

Yes and no. Large companies generally use both in-house and outside patent counsel. The outside attorneys are retained for opinion work while the inside attorneys draft and prosecute the applications (speaking generally). IBM has something like 200 internal patent attorneys (from a quick search of the USPTO attorney/agent roster). Figure an average cost of 200k/year (salaries, benefits, overhead cost) per lawyer, and 10 patent applications per year, and the 20k/year/application might be just about right.

Then there are the costs of foreign applications. Every patent (application) is evaluated for potential importance, and a corresponding level of foreign filing is determined. Even a small breadth of foreign filing is expensive, and a large foreign filing can cost upwards of 100k or more.

And once again, the headline is moronic. How does IBM "win" patents? In a poker game?

IAAPA

Re:Doesn't maintaining patents cost money?

[Anonymous Coward]

In IBM's case it would be higher, as they actually give their employees a high fixed bonus per patent obtained (which also helps explain why they file so many of them).

Re:Doesn't maintaining patents cost money?

[jrady]

Ok, so lets assume for a second that IBM really wants to benefit the OSS movement by copylefting most of their Software Patents

(the fact that i believe that SW Patents are BS in the first place is another issue).

They file it, and maybe get it granted rather quickly.
In the US its 3.5 Years until they have to pay an annuity fee for another 3.5 years.
lets just assume, instead of paying this, they drop it.
the Patent is still on record as prior art.
Meaning that any Patent Troll, trying to re-patent this "invention" can be easily blown out of the water.
So: Mission accomplished - Software out in the open, no chance for Patent troll to cash in on it.

math FAIL

[Anonymous Coward]

'IBM said it earned 4,186 U.S. patents in 2008, more than triple the number of patents earned by rival Hewlett-Packard [...] 1,424.

Um, 1424 * 3 = 4272 > 4186 ...

Re:math FAIL

[AvitarX]

They probably rounded it for their calculations:

1000 * 3 = 3000 4000

Re:math FAIL

[AvitarX]

damn, 3000 < 4000

I previewed this time.

Re:math FAIL

[Gat0r30y]

Clearly, they must have used one of their fancy new patented algorithms to arrive at

more than triple the number of patents earned by rival Hewlett-Packard

Re:math FAIL

[madhatter256]

Yes, in a literal sense. However... 4186/1424 = 2.939

Round that into a whole number and you get 3.

Dun dun dunnnn!

Re:math FAIL

[Anonymous Coward]

It's still not more than 3 though...

Re:math FAIL

[rts008]

It was calculated in Excel, on a Pentium chip?

Re:math FAIL

[Anonymous Coward]

at least it's not OVER 9000

Not Surprising

[Anonymous Coward]

IBM employees get a bonus for filing patents (patent doesn't even need to be granted). And the process of filing is to submit your idea to a database and lawyers take care of the rest. They do filter what ones they will file. But there is definitely a lot of mud against the wall going on with that process. Also really helps to have a few filings come yearly review time.

Re:Not Surprising

[nategoose]

I hold a patent on the process of increasing performance review ratings via filing for patents.

Re:Not Surprising

[Kalriath]

That isn't even funny any more. Everyone knows you can't patent business processes.

Re:Not Surprising

[DirePickle]

That's because I have a patent on patenting business processes.

Re:Not Surprising

[Anonymous Coward]

I'm pretty certain from your description you DON'T work for IBM. Yes, the initial stages are quite simple (submit it to a database) but later stages (through peer review, search, prior-art review and justification, advance to filing status, review of all the reams of legalese from the lawyers so you don't miss claims and so on) take a sizable commitment from the employee. And you still have to do your day job or risk getting the chop :-).

No innovation left in the human race?

[Stile 65]

You might have heard or felt that there is little left to patent these days...

No, not so much. On the contrary, I think we're accelerating happily toward the Singularity and new inventions are very much part of that trend.

Wait a couple of years

[Anonymous Coward]

> IBM has most patents invalidated of any company for 2011

Can a [money] value be put on these patents?

[bogaboga]

What have IBM and the whole bunch of these companies had as monetary gains from these enormous patents? Sometimes I just do not get it.

Re:Can a [money] value be put on these patents?

[Endo13]

It reduces the chance someone will sue them.

Re:Can a [money] value be put on these patents?

[pines225]

Yes, a money value can be put on what they earn from patents, though experts disagree wildly on the figure. A commonly touted figure for IBM's annual patent royalty income is $1 billion (that's right, spelt with a "b") - basically, some analysts consider only the straight licence income, others figure that the patents add value to sales of product and other parts of the IBM balance sheet ... but nobody puts it under &100 million p.a. http://www.iam-magazine.com/blog/Detail.aspx?g=9be3f156-79b1-49f4-abf1-9bee7e788501

Re:Can a [money] value be put on these patents?

[freddy_dreddy]

There's a large misconception that patents are solely created to shield things off (protect) without discussion. Having worked for one of these corporations gave me an insight in what they're actually used for (although I can't generalise).
Patents in my field (micro-electronics) were used to negotiate access to other patents. Corp. A holds a patent which Corp. B needs: the two sit around the table (a number of times) and B has to offer something in return - access to B's patents in most cases.
All these things are negotiated behind closed doors, only a small fraction of the cases result in court cases, but in my experience the patent is then used in it's true sense: to protect R&D or potential thereof.
Yes: there are patents which just put a lock on an idea. But these grant the patent holder access to the patent portfolio of other companies. Such patents -in almost al cases- focus on a small amount of very well-chosen corporations, to get access to their portfolio. If they don't give in it normally ends in court, there are a lot of cases like this - notably the David vs Goliath cases. You could compare it to some extent with registering a domain and sitting back and waiting to see if they'll pay up.

Patents can be copylefted

[tepples]

So patents are against the GPL except from the ones by IBM. Wow What is IBM pay to get FSF Support.

A patent that is freely licensed for use in copylefted software does not violate the GPL. SELinux and PlusV patents are licensed this way, and so was On2's VP3 in the early days of the Theora project.

Re:Patents can be copylefted

[jellomizer]

That doesn't really seem fair.

We are OK with patents if you let us use it, mentality, if you don't then you are in the wrong. What about those small companies making software who can get killed with a lawsuit for stepping on an IBM patent accidentally.

It is like saying we believe in the freedom of speech for Us, for others we just don't care.

The GPL FSF are a very polarizing set of rules, which only suits itself. Just because the other guy is doing something you disagree with does it mean he deserve less rights.
If you are going to be vocal about patents you should say if it is freely and automatically licensed to anyone free or not. The patent is just proof that no one else can sue you.

Re:Patents can be copylefted

[LingNoi]

The free software foundation are pushing the agenda of free software being your only option available so I don't understand why you think it is somehow hypocritical. That is not to say they are succeeding however I think this is a pretty safe claim to make seeming as they founded the GNU project.

You could try and counter-act this point I made by saying that they just want to give the freedom of choice. If that was the case then why create such a stink about non-free software repositories? No, there motive is for the majority of computer users to be using only free software.

I'm not trying to debate the ethics of that, just trying to point out how it isn't hypocritical.

Re:Patents can be copylefted

[Draek]

We are OK with patents if you let us use it, mentality, if you don't then you are in the wrong. What about those small companies making software who can get killed with a lawsuit for stepping on an IBM patent accidentally.

Wrong. It's more like "software patents are a huge idiocy, but by letting us use them as we see fit you're minimizing the collateral damage, so you aren't as bad as the rest". And yes, small closed-source software companies *could* have trouble with these patents, but given that they also patent stuff on their own I find it hard to care.

If you are going to be vocal about patents you should say if it is freely and automatically licensed to anyone free or not. The patent is just proof that no one else can sue you.

Say that to IBM et al, the FSF has very little to do with their actions so you're aiming at the wrong place.

IBM is large

[drewzhrodague]

IBM is a very large company. They make and do all kinds of things, and lots of them are patentable. I'll bet there are people who get a bit of commission if they make a certain number of patents. Also, they do some very wonderful things that deserve a patent. Wasn't there one last week?

Re:IBM is large

[swillden]

I'll bet there are people who get a bit of commission if they make a certain number of patents.

As I recall, (I'm an IBM employee) it's a $5K bonus for your first patent, and, I think $2K per patent after that. There might be some limit on the number of smaller bonuses, but I don't think so.

Re:IBM is large

[Anonymous Coward]

I'll bet there are people who get a bit of commission if they make a certain number of patents.

As I recall, (I'm an IBM employee) it's a $5K bonus for your first patent, and, I think $2K per patent after that. There might be some limit on the number of smaller bonuses, but I don't think so.

It's $1500 for your first patent then $750 for every one after that. Then every 12 points you earn you reach a plateau and get $1250. If your idea doesn't get patent approval it might get a "publish" rating in which case you don't get money but do get a point.

Patent = 3 points
Publish = 1 point

You can't apply more than 6 points to a plateau. So to reach the plateau you can use points from all patents or a mix of patents/publishes.

Re:IBM is large

[fizzziks]

I am also an IBM employee. I have sort of the same numbers, however slight different. I am almost certain it is $1,500 for your first file. Then, for subsequent patent files, the price depends on the number of inventors for the patent. If you have 4 or less inventors, you receive $750 or $800 per inventor. If it involves more than 4 inventors, you receive $3000 divided amongst all of the inventors. Also, you receive patent 'points' as you file. The first and subsequent files are 3 points each. Once you hit 12 points, which is the plateau, you hit $1,200. IBM is a very very patent driven company. They have internal patent competition within the company. Hence, you promote internal competitiveness within your teammates, while also placing the entire company forward. IBM is a great company to work for if you want to work on cutting edge technology. I am not being biased here, but I've worked for Lockheed Martin and a few aircraft companies... and IBM thus far as been the most rewarding in the technical sense.

A lot of them would be deserved

[Zouden]

When IBM is doing things like increasing the resolution of MRI by a hundred million times, I'd say they must have earned a lot of their patents. They do much more research than HP or Microsoft.

Re:A lot of them would be deserved

[afidel]

Most of the basic research that has driven the increase in HDD density was done by IBM researchers.

Re:A lot of them would be deserved

[darkmeridian]

IBM scientists were also responsible for the excimer laser that is used for LASIK eye surgery. They were also the dudes who wrote out "IBM" with atoms. IBM gets a lot of patents but they also do a lot of basic scientific research to earn those patents. So it's not that they "win" a lot of patents, they "earn" a lot of patents.

Mathematical patent

[Anonymous Coward]

They should get a patent on the theorem showing 4,186 is more than three times 1,424. No prior art there, for sure.

Re:Mathematical patent

[jd]

The Pentium I wishes to disagree.

Uh oh

[Anonymous Coward]

Unfortunately I have a patent to count how many patents one has, so you all are infringing and as a licensing fee I want your firstborn or a pound of flesh

Re:Uh oh

[Anonymous Coward]

Will you settle for all the labor of my first born?

No, it's a new kind of math

[EmbeddedJanitor]

patented of course!

Re:No, it's a new kind of math

[rhyder128k]

I hereby patent a method of making money. The method involves an automatic process of patenting everything that I think that I might have a hope of patenting in the hope that the opportunity cost to my competitors in addition to any actual money made by the patents will add up to something greater than zero.

Same at many places

[EmbeddedJanitor]

This policy was also used in the last three companies I worked for. AFAIK, it is also used in HP, Apple, Intel and just about all similar companies.

Re:Same at many places

[Daniel Dvorkin]

Now if only these companies would start offering bonuses for finding prior art that invalidated competitors' patents, maybe we'd see an end to some of the patent insanity.

Re:Same at many places

[Anonymous Coward]

IBM for one takes prior art very seriously. They have a local peer review and a *very* good search department which is very hard to get suspect patents through to filing.
They treat patents the same way they treat software. Catch the bugs as early as possible since it gets more expensive to fix them at later stages.
Catch them locally and you've wasted a few hours of employee time. Catch them at the lawyers end of the pipeline and you've already expended vast amounts of money.

earned?

[frovingslosh]

I question the use of the word "earned" here. Paid a bunch of lawyers to get stuff past the patent examiners is one thing, actually earning a patent is quite another.

Re:earned?

[CannonballHead]

Kinda depends on who is filing, don't you think? It's one thing for a patent troll to file, it's another for IBM to file. IBM, last I checked, hasn't been involved in the whole get-patent-sue-other-company-for-money thing.

Re:earned?

[ClosedSource]

Why sue other companies when those companies are scared shitless of IBM and will payup? Lots of people pay protection money rather than get beaten up. That doesn't mean it's not a protection racket.

Re:earned?

[frovingslosh]

Just the fact that you're not seeing "IBM sues little guy who stood up against their patent claim and unlimited supply of lawyers" on Slashdot does not make all of the patents that they get legitimate (and I bet that more than a couple just might be software patents, and I would say the vast majority of those are far from legitimate and "earned").

I'm gonna win next year.

[geekmux]

I was just awarded the patent for being awarded the most patents in a year.

SUCK IT IBM, you'll be paying me next year!

Or

[deblau]

You could get the data directly from the USPTO here. They support full boolean searches on a large number of fields. For example, a search for "international business machines" as the assignee in the patent database shows that IBM has 52,781 patents (some of which are expired). Narrowing the search to those issued in the last year shows 4166 patents issued in 2008. The exact query I used was (an/"international business machines" and isd/20080101->20081231) without the parens. You can search patents and published applications -- IBM currently shows 27685 applications pending (and that's just those that have published since 2001).

Re:Or

[Anonymous Coward]

Some of those 27685 patent applications aren't actually pending anymore. AFAIK, there's no easy way to filter out the abandoned applications and issued patents from that number.

Are most of the patents in software?

[Anonymous Coward]

Japan and USA are the only countries supporting software patents. Since IBM is filing them in USA, are they really all practical applications like Creating a MRI With 100M Times the Resolution or mostly software patents? The article seems to imply that they don't file any software patents.

A prediction

[Anonymous Coward]

I predict that IBM will use the profits from this year's record number of government-granted monopolies to perpetuate their 16-year run of obtaining record numbers of government-granted monopolies.

Patenting patenting

[crazygamer]

They should patent winning so many patents in a year as to set a record, and then sue anyone who dares to cross them.

Triple?

[__aabvlw4075]

Since when is 4186 "more than triple" 1424?

"earned"?

[spasm]

I'm having a problem with the word "earned" in the original article. It implies work was done, beyond that needed to file the paperwork.

Re:"earned"?

[Anonymous Coward]

So how does this protect the work of the original scientists and artists? Corporations should not be allowed to hold patents. Only individuals. I mean seriously, what's the incentive to do research and patent it if your company can just fire you and continue to profit from your patent while you get nothing? I quit giving all my best and brightest work to my employers a long time ago. Now they get exactly what they pay for, nothing more.

How to turn an open source idea into a patent

[Anonymous Coward]

Open source project from 2004:
http://www.jchains.org

IBM Patent Application from 2006: http://aiw1.uspto.gov/.aiw?docid=us20060288401ki&PageNum=4&IDKey=0D84250DDE48&HomeUrl=http://www.uspto.gov

IBM should patent LayoffEveryAmericanNow (LEAN)

[Anonymous Coward]

What your IBM sales rep will never tell you (they're under orders not to discuss it), from www.pbs.org/cringely/pulpit

"LEAN is about offshoring and outsourcing at a rate never seen before at IBM. For two years Big Blue has been ramping up its operations in India and China with what I have been told is the ultimate goal of laying off at least one American worker for every overseas hire. The BIG PLAN is to continue until at least half of Global Services, or about 150,000 workers, have been cut from the U.S. division. Last week's LEAN meetings were quite specifically to find and identify common and repetitive work now being done that could be automated or moved offshore, and to find work Global Services is doing that it should not be doing at all. This latter part is with the idea that once extraneous work is eliminated, it will be easier to move the rest offshore.

All this is supposed to happen by the end of 2007, by the way, at which point IBM will also freeze its U.S. pension plan."

Math fail?

[Anonymous Coward]

Uh, I don't really think IBM's 4,186 patents is more than triple of HP's (3 x 1,424 = 4,272).