Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Businesses The Courts Apple Technology IT

Apple Lawsuit Says 'Stealth' Startup Poached Engineers To Steal Secrets (reuters.com) 35

Technology startup Rivos allegedly stole Apple's computer-chip trade secrets after poaching its engineers, Apple said in a lawsuit filed in California federal court. From a report: Apple's Friday lawsuit said Mountain View, California-based Rivos has hired over 40 of its former employees in the past year to work on competing "system-on-chip" (SoC) technology, and that at least two former Apple engineers took gigabytes of confidential information with them to Rivos. Rivos is a "stealth" startup that has largely avoided public attention since its founding last year.
This discussion has been archived. No new comments can be posted.

Apple Lawsuit Says 'Stealth' Startup Poached Engineers To Steal Secrets

Comments Filter:
  • That should make the case quite interesting!
    • by Alworx ( 885008 )

      objection: hearsay!

      • Isnt this what parents are for? Having all the knowledge is useless if the USPTO kicks your ass and gives your product back to the patent holder, or the winnings thereof.
        • parents?
        • In theory, yes. In practice, California has passed some "trade secrets" laws that make it so patents aren't required.

        • by Whateverthisis ( 7004192 ) on Monday May 02, 2022 @05:38PM (#62497792)
          Not everything is patentable, and not everything that is patentable should be patented. Let's stick with chip terminology since that is what's at stake here.

          The design of, let's say a new metal interface to connect transistors that results in some better performance between layers is patentable; that is an invention. More importantly, it is something you would want to patent, especially a group like Apple, because it is possible to analyze a chip layer by layer and identify what you did, but more importantly it's also enforceable because you could analyze a competitor's chip layer by layer and prove that it violates your patent.

          A process that might give you extremely tight quality control while introducing areas for improvement isn't really patentable. While this might be technically feasible and give you a quality edge over your competitors, the reality is you won't have the ability to tell when a competitor is using your technology because it doesn't express in the end result. A patent is only as good as you can tell when someone is violating your patent; if they're doing it in house it may not be seen in the end product and you can't enforce it. There are methods and processes that are patentable; for example the Bosch Process for deep reactive ion etching, is patentable (and was patented), although notably the patent was made available at a low license cost to enhance the industry as a whole, which Bosch benefitted from (sorry for the tangent).

          A system architecture is not patentable. It's just someone's design. Generally the way people do system on chip architectures can be patentable, but say Apple's unique chip design using those methods are not patentable. Notably this often comes out in the chip mask used for production purposes, and the mask is actually copyrightable, because a copy-right is an expression of an art, and copyrights have 99 years of protection unlike patents. However, you can't really copyright an unfinished design or architecture, but documents leading up to that are certainly internal knowledge.

          The other big form of IP that is seriously overlooked is why a company is doing something. In this scenario, Apple is projecting the market is going to move in X direction, and is thus making engineering decisions based on what they think the product will do. A design specification for example will have certain performance characteristics, and why it has those has everything to do with Apple's sense of where multi-billion dollar markets are going to evolve. Specs are not patentable or copyrightable, but given Apple's dominance in this position their specifications and the drawings on how they plan to meet those specifications absolutely provide intelligence on multi-billion dollar opportunities, intelligence that isn't patentable.

          And in some cases that market knowledge is more valuable than a patent. All a patent does is keep people from using your technology for a period of time; people often ascribe value to a technology because it's patented, but if I have a patent on a say a pencil made out of steel instead of wood, well that's not a valuable patent unless there is a market for it. The market knowledge tells you what tech and designs have value, leading to enormous financial opportunity. Someone using Apple's specs and designs means they are much less likely to go down an incorrect path and find a path that a market will pay for, and possibly beat them to the market as a small team (with an admittedly inferior product), or even worse predict where Apple is going and try to patent things in Apple's future, giving them leverage and forcing Apple to pay them for the patents they patented ahead of them.

          So TLDR; not all protected information is patentable, and plenty of non-patentable information has significant value.

        • Isnt this what parents are for? Having all the knowledge is useless if the USPTO kicks your ass and gives your product back to the patent holder, or the winnings thereof.

          Unfortunately, a LOT of Research in its preliminary stages isn't protected by Patents. . .

          Or by Parents, either!

          (Sorry; couldn't resist!)

        • Trade secrets, by definition, can't be patented. If somebody else independently comes up with the same idea, that's fair game. However if somebody steals it from you, then you can sue. Also the USPTO doesn't "give your product back" to the patent holder. At best, they might forbid you from selling it or importing it.

    • I f Fruit can tell that Gb were taken inappropriately, how come they couldn't stop it at the time? Or deal with it when the people handed in their notice?

      More like Fruit is hoping to use discovery to show malfeasance, and the rest is a guess.

      • by tlhIngan ( 30335 ) <slashdot&worf,net> on Monday May 02, 2022 @06:03PM (#62497840)

        I f Fruit can tell that Gb were taken inappropriately, how come they couldn't stop it at the time? Or deal with it when the people handed in their notice?

        More like Fruit is hoping to use discovery to show malfeasance, and the rest is a guess.

        Because having access to that information might be a part of their regular duties?

        I mean, I can copy over a project file to my computer for work purposes and it won't set off any alarms, nor should it because it's part of my job.

        However, if I proceed to take that file to my new job, then it goes from authorized use to unauthorized use. And the legal system requires the crime to happen first before it can be dealt with.

        So those employees could be copying the files as part of their jobs as authorized use. But they then have to walk off with it in order for it to start crossing the line to unauthorized use at which point Apple can sue over it. Apple cannot sue beforehand (there are many legitimate reasons why the files might be copied), nor do they want to impede the regular function of the company by accusing people of stealing information when it's required to do their jobs.

      • I f Fruit can tell that Gb were taken inappropriately, how come they couldn't stop it at the time? Or deal with it when the people handed in their notice?

        More like Fruit is hoping to use discovery to show malfeasance, and the rest is a guess.

        Because they probably don't do forensic audits on file accesses until after they start to suspect something,

        Duh.

  • Bringing Apple's M-series tech to RISC-V would be interesting.
    • My understanding is that ARM and RISC-V are both RISC architectures so some technology can be used. From my perspective, one of drawbacks of RISC-V is that it is very new and still has a lot development to go. Borrowing or stealing it from other designs would be one way to shortcut the development process.
      • Both being "RISC" is about as useful as a chocolate tea pot. MIPS, SPARC, and ARM are all RISC, but couldn't be more different.

        The only one with a passing semblance to RISC-V is MIPS, as the ancestry goes very vaguely along the lines of MIPS -> DLX -> RISC-V.

        • And RISC-V can never implement anything that Apple has done in their ARM chips?
          • by sl3xd ( 111641 )

            I'm not saying they can't do something similar. I'm saying "ARM and RISC-V are both RISC architectures" means exactly nothing to the entire discussion.

            Even RISC vs CISC is nearly meaningless - a PowerPC G3 (Reduced Instruction Set) processor has more instructions than its contemporary CISC Pentium II competitor. Many chip designers call this the "Post-RISC era" for a reason.

            Just because Apple has done something with ARM isn't necessarily a question of if RISC-V can do it -- but whether it should be done? It

      • by _merlin ( 160982 )

        The trouble is RISC-V isn't new. It's basically glorified MIPS, with none of the lessons learned in the last three decades applied. It's fine for embedded applications, but it doesn't scale up to big multi-issue supercalar architectures. There's a reason Libre-SOC went with PowerPC instead of RISC-V, and AArch64 is an even newer clean sheet design (it's not really ARM: it doubles the registers, makes PC an SFR, drops the scalar FPU register file, gets rid of most of the predicates, and many other changes

        • It's basically glorified MIPS, with none of the lessons learned in the last three decades applied.

          I laughed, but on a serious note- not entirely true- no more branch delay slots on RISC-V (thank fucking $DEITY)!

  • by VeryFluffyBunny ( 5037285 ) on Monday May 02, 2022 @04:03PM (#62497564)
    ...to win the case but to bankrupt a possible rival. A start up with Apple Inc.'s lawyers gunning for them? Gonna be hard to find backers in that case.
  • Likely, whomever has more money will win in Santa Clara County.
  • by aerogems ( 339274 ) on Monday May 02, 2022 @05:01PM (#62497732)

    If, and right now it's basically just Apple making an accusation, but if some of the poached employees did indeed copy proprietary files and took them with them, that's going to put Rivos in an uncomfortable position that companies like XTAL and Uber can probably relate to. Apple can't do much about what these people know if they decide to apply that knowledge at some other company, but they aren't entitled to take any files along with them to the new job. If they did that, Apple's going to nail them to the wall.

    • by sl3xd ( 111641 )

      Didn't the guy from Waymo get plastered to the wall for stealing Google's code and taking it to Uber? Something like 18 months in prison and Uber on the hook for $179 million?

  • by rgmoore ( 133276 ) <glandauer@charter.net> on Monday May 02, 2022 @05:13PM (#62497760) Homepage

    I would take all this stuff with a huge grain of salt, at least before Rivos has had a chance to say its piece. A key is that there's nothing illegal about hiring away key employees, even ones who know a lot about the stuff a company has been working on. California law makes it nearly impossible to prevent ordinary employees from going to work for a competitor, or for the competitor to target the employees.

    That said, the law does prohibit employees from taking their former employer's trade secrets to a competitor. How many trade secrets an employee can take in their head is a tough question, but bringing documents is clear cut. If the employees took Apple's documents as Apple alleges, they're in trouble. If Rivos asked or just encouraged them to take the documents, Rivos will be in trouble right there with them.

    • If the employees took Apple's documents as Apple alleges, they're in trouble. If Rivos asked or just encouraged them to take the documents, Rivos will be in trouble right there with them.

      I predict this is the last we will hear of Rivos or these engineers.

      Those Engineers probably oughta just save time and take down their LinkedIn Profiles now, and start learning how to drive Trucks!

Enzymes are things invented by biologists that explain things which otherwise require harder thinking. -- Jerome Lettvin

Working...