Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
The original Intel lawsuit specifically mentioned Scikit learn (an open source library) and the algorithm was quickly removed from that library. It is really up to HDVY to protect their IP. Having it publicly listed as open source and free to use is not a great way to protect that IP. It is up to them to fix this, and future lawsuits might be problematic for them because of this very public misinformation.
Alan
You will have to bring that up with both google and copilot
Alan
I just asked copilot about open source versus royalty payment machine learning algorithms:
copilot claims scikit-learn, tensorflow, pytorch, xgboost, lightgbm, and Keras are all open source (free to use) while SAS, Matlab, and IBM Watson require a royalty payment. I then specifically asked it about SVM-RFE and it said it is available as open source :-/
Interesting times. The Intel product line is stronger than ever with the crop of on time 2024 product releases. Lunar Lake, Arrow Lake, Gaudi 3, Sierra forest, and Granite Rapids are all great leading products. OTOH, the financials are crap. All are too expensive to build based on their selling prices.
Intel needs to downsize to fit their current revenue footprint of $50B to $60B, rather than the $70B to $80B they were sized for.
Intel_18A leading edge fab technology is looking very good for both Panther Lake and Clearwater forest introduction next year. The long term continues to look very rosy, but the manana mantra has worn very thing.
Tan came to Intel from Cadence, a maker of design tools. I don't think he had anything to do with AI.
The press on this is mixed, but according to his resignation letter: he was leaving the board because of a personal decision to “reprioritize various commitments” and that he remained “supportive of the company and its important work,” in a regulatory filing on Thursday.
I am not a fan of technical analysis and the "chart readers", but rely more on fundamental analysis and things underlying the company performance. That said, I think there is a fair amount of self fulfilling prophecy with TA. Based on the action over the past few days, what say you TA experts?
Intel recently updated Q2 guidance so it is very unlikely we get another update until Q2 earnings call. That said, it is possible some of the analysts will change their minds on the stock causing a move, most likely up.
Earnings Thursday PM may drive the stock up... we will see what they deliver.
Hey MB, Yeah, it seems iHUB caters to the small caps to a great extent.
Intel has had QOQ growth for a few quarters now with continued growth expected by most. I think Q1 is when we start seeing YOY growth reported which should be a real turning point on sentiment. It has been a rough couple years but the turnaround seems real and on track.
Happy Holidays, Merry Christmas, and whatever else we have to celebrate!
Alan
One thing that happens here is $1M to $10M verdicts are very common. However, the big money cases make the news so they appear much more common than they are. The number of lawsuits very large companies are dealing with is underestimated, and the results of those suits go unnoticed. The HDVY result did not even make it into the Intel 10Q, or even the news AFAIK. Any NPE with a patent, which number in the hundreds of thousands, can and will launch lawsuits against large corporations looking for some money. Companies that use patents to make products generally have broad cross license agreements with others that make products, so avoid the legal battles. NPE's have no need for such a cross license agreement so the only way to monetize the patent is through a lawsuit.
Alan
I have no data on how often law firms win contingency cases nor their criterion for accepting them. That said, it appears to me the HDVY legal team was not investing too much in this case, and I suspect the $2.25M is more than enough to cover expenses and give the attorneys a little extra for their effort.
I did find a reference from a few years ago that mentioned the median bench (judge issued) award is $1.9M, while the median for both juries, and settlements is closer to $10m. I suspect a legal team goes in expecting some number like this.
We really need HDVY to publish some financials so we get a better picture. HDVY may have paid some up front from their previous legal winnings, however when we see legal expenses on the earnings report we don't really know if they are vennwest or Intel
I believe he was more specific and qualified it as AI in the biomedical field. This may be due to the reality that they failed in securing much money outside the biomedical field with the Intel lawsuit.
Alan
It was clear to me, but I'll add a few more words.
In the original response Intel provided they stated that they did not use SVM-RFE so did not need a license. It is easy to prove something exists, but impossible to prove something does not exist. It was clear based on the evidence in the original lawsuit that in the past Intel has run the SVM-RFE algorithm. It was not clear if the algorithm ever made it into any Intel products. It is likely the settlement amount is what Intel offered and HDVY accepted, rather than what HDVY wanted. It is also likely that during discovery HDVY was not able to identify any Intel products that infringe the patent. This is not to say such products don't exist, just that HDVY was unable to prove such a product exists. It is likely the Intel payment is for past use of the algorithm, which was proven. It is unlikely the payment is meant to cover Intel usage of the patent in current or future shipping products.
Merry Christmas, and Happy Holidays to all! May 2024 bring us all happiness.
Alan
Keep in mind the patent on SVM expired a while ago. A way to think about it is:
Machine learning is one type of AI (artificial intelligence)
SVM is but one of many types of machine learning algorithms.
SVM-RFE (what the dispute is about) is but one of many types of SVM algorithms.
Alan
Intel's own published literature boasted of "their" revolutionary SVM-RFE technology
I have not seen this. Do you have a link, or article title for this evidence?
Alan
that is certainly possible (but I view as VERY unlikely) that the agreement included future payments from INTC to HDVY.
The whole hold some back until vennwest is settled strikes me as having the possibility of a redo on the vennwest suit due to bargaining in bad faith... but again, I'm not a lawyer.
Now we wait.
Intel appears to set a threshold of around $300M to consider an item "material" and reportable so $2.25M is not material for Intel. The 10-Q had the following wording:
Legal Proceedings
We are regularly party to various ongoing claims, litigation, and other proceedings, including those noted in this section. We have accrued a charge of $2.2 billion related to
litigation involving VLSI and a charge of $401 million related to an EC-imposed fine, both as described below. Excluding the VLSI claims, management at present believes that
the ultimate outcome of these proceedings, individually and in the aggregate, will not materially harm our financial position, results of operations, cash flows, or overall trends;
however, legal proceedings and related government investigations are subject to inherent uncertainties, and unfavorable rulings, excessive verdicts, or other events could occur.
Unfavorable resolutions could include substantial monetary damages, fines, or penalties. Certain of these outstanding matters include speculative, substantial, or indeterminate
monetary awards. In addition, in matters for which injunctive relief or other conduct remedies are sought, unfavorable resolutions could include an injunction or other order
prohibiting us from selling one or more products at all or in particular ways, precluding particular business practices, or requiring other remedies. An unfavorable outcome may
result in a material adverse impact on our business, results of operations, financial position, and overall trends. We might also conclude that settling one or more such matters is
in the best interests of our stockholders, employees, and customers, and any such settlement could include substantial payments. Unless specifically described below, we have
not concluded that settlement of any of the legal proceedings noted in this section is appropriate at this time.
I am not a lawyer nor accountant, but in the past when there has been a "big" win Intel has disclosed it as soon as documents are signed. It seems to me that if Intel knows they need to make a large payment, that is financially material so they must disclose it. Of course once the payment is made it is also disclosed in several different ways.
I generally believe anything is possible, but I put this one in the VERY unlikely category. Given there was already a monetary transfer, why would there be a second one? The argument that HDVY needed some money now to pay legal bills, and did not want a large amount shown until other litigation is settled does make sense... but I still view it as VERY unlikely.
Alan
The only problem with this line of thinking is that if Intel had a future obligation to pay $1B or so they would have been required to disclose it, and they have made no such disclosures.
Alan
The key message here is a year and a half after VLSI "won" $2B, they have to give it back.
Alan
$2.18B VLSI verdict overturned on appeal. I know some viewed this very large verdict, even though well outside the guidelines, as setting expectations. Intel won their appeal, and the case will go back to court.
What is RMTD? I do believe the $2.25M was called a settlement.
Alan
It appears my previous post was not clear, so lets make sure I use more words and make the timing more clear.
During the patent interference proceedings both HDVY and INTC submitted motions to the court. In those motions Intel submitted Li motion 4 regarding the unpatentability of the invention. HDVY also submitted a similar motion, known as Weston Motion 4. When the judge ruled on the interference claim (awarding the patent to HDVY), he chose to not rule on either Intel or HDVY motion 4 (unpatentability) but rather dismissed both motions.
Based only on INTC motion 4, HDVY included the following text in their original lawsuit:
Intel also tried to invalidate all of the HDC patents-in-suit, as well as sacrifice its own ‘077 patent
So given the text of Li motion 4, HDVY is drawing the conclusion that Intel attempted to invalidate the patent. We must then also conclude based on Weston Motion 4 that HDVY attempted to invalidate the patent.
Since the judge in the interference trial chose not to rule on patentability, that opened the door for Intel to attack HDVY on this issue, AFTER HDVY issued their lawsuit (which contained the above text) against Intel. Clearly once the lawsuit was launched, Intel ramped up their efforts to invalidate the patent, but that doesn't change HDVY's own language that the text of either Motion 4 was an attempt to invalidate the patent. Certainly the judge for the interference trial could have chosen to rule on motion 4 to invalidate the patent, or to affirm the patent, but he chose not to. Both Intel and HDVY had motions in the interference trial to invalidate the patent.
Alan
Sorry for the later reply, but the site has been giving me server errors
Zenos wrote: I think you mistakenly stated that HDVY claimed their patent was invalid (not so).
From the original interference proceedings:
Li Motion 4 – seeking judgment of unpatentability of the Weston claims
corresponding to Count 1, on the basis of 35 U.S.C. § 101. Id.
Which resulted in the following bombast from HDVY in the original lawsuit:
Intel also tried to invalidate all of the HDC patents-in-suit, as well as
sacrifice its own ‘077 patent
However, also from the interference proceedings:
Weston Motion 4 – seeking judgment of unpatentability under 35 U.S.C §101 as being directed to non-statutory subject matter.
My understanding and interpretation of the facts of the case are consistent with my observed outcome. I would suggest that rather than believing the observed outcome is incorrect, perhaps the interpretation and understanding of the facts needs to be revisited.
Alan
Thanks for the feedback! I recognize that we all have confirmation bias and see in the data what we expect to see. I often realize that leaves me blind to some realities.
That said, it was entirely possible that during the discovery phase HDVY would find a "smoking gun" of current Intel use of the patent. This could have led to the gold rush many of you were hoping for.
Intel clearly stated in their response to the lawsuit that they make no current use of the patent. Perhaps they were mistaken. Perhaps they knew they did and were hopeful nobody would find out. As combative and competitive as Intel can be, it is unlikely they would lie. OTOH, they do have something like 30,000 people writing software and it is certainly possible one of them incorporated SVM-RFE into some code somewhere. I suspect the odds of either Intel or HDVY uncovering this as vanishingly small.
Alan
It is not clear HDVY could have received an injunction for a couple of reasons.
HDVY had no competing product, so preventing Intel from shipping their product was of no financial benefit to HDVY. Usually injunctive relief is there to prevent unfair competition.
I don't think HDVY knew of (or at least had proof of) any current Intel products that they could issue an injunction against.
Alan
That all makes sense. However, in this post:
https://investorshub.advfn.com/boards/read_msg.aspx?message_id=173203678
you mentioned HDVY sent a letter to certain universities requesting a licensing's fee and potential royalties. I was curious about what those fees were expected to be. It is clear that once you enter the legal system, all things change.
Alan
It seems in legal wranglings words often have different or nuanced meanings compared to what we might think.
For instance, you need to be a bit careful about what is "claimed" in a lawsuit. Both HDVY and INTC claimed that the patent was invalid during the interference proceedings. All Intel / HDVY were doing is asking for legal rulings on issues. Their particular position/belief on any issue is irrelevant as a legal matter. All that matters is what the court rules.
I asked earlier if HDVY was able to collect on any of their previous requests for royalties that were mentioned. Does anybody know what royalty rate HDVY was asking for? Given HDVY had already established a royalty rate back in 2011, they could easily petition the court to establish an escrow account for Intel, and others who acknowledge they use the patent, to pay into until the interference was settled. I think you are right that Intel would not willingly agree to this; however not because it goes against their legal attempts but rather as a matter of Intel policy. Given HDVY had attempted to receive royalties, even after the interference was discovered, indicates that HDVY was not prevented from collecting royalties due to the interference.
Another important point is Intel's written response to the lawsuit, in which they state they do not need a license for the patent because they do not use the patent. Of course "do not use", is different than "have used". HDVY had already presented evidence that Intel had used the patent in the past, but had no evidence that Intel currently uses the patent. To this point, it would seem improbable a company would knowingly pay royalties to use the SVM-RFE algorithm when there are many comparable, if not better, algorithms available open source for free.
With all of that said, perhaps and escrow account would me meaningless since Intel claimed zero usage their royalty payments into the account would be zero until HDVY could prove current usage.
Here is the first thing that popped up regarding licensing, infringing, and escrow accounts:
https://webstorage.paulhastings.com/Documents/PDFs/medtronic-decision-client-alert.pdf
Alan
The interference proceedings certainly made it more difficult, but not impossible to collect royalties. One solution would be to establish a royalty rate, and have the funds go into escrow until the legal matter is settled. It would have been in HDVY's best interest to establish such a deal with Intel as soon as the dispute occurred. Intel and HDVY would have different ideas on what the royalty rate should be, which would make this much more complex.
Based on an earlier post, it appears HDVY did enter into some licensing agreements prior to the dispute. Do we know if they collected any royalties through this process, and if so what the royalty rate was?
On an unrelated note, I have heard two different versions of the "copy exactly patent" observation:
(1) Intel just brazenly copied the HDVY patent
(2) HDVY modified the text of their patent to match Intel's wording in order to force the interference proceedings.
Does anyone have any evidence supporting one or the other theories?
Alan
I see three options here...
(1) as you mentioned, gather the resources necessary to pursue other infringers. Note that while Intel was an obvious target since they attempted to patent the same invention, they are not a software company. I would see software companies as much more likely to use a machine learning algorithm.
(2) Instead of trying to pursue infringers themselves, they could hire a company that specializes in monetizing patents. This would be a lower cost and lower effort path, but perhaps with lower payback.
(3) they could sell the patent and close up shop.
It doesn't appear they have any assets or talents beyond the patent portfolio. I have heard that the "ticker symbol" is worth something for somebody attempting to incorporate and go public. Perhaps that can count as an asset as well, but I suspect not worth too much.
just my two cents worth,
Alan
I assumed (perhaps incorrectly) that there would be some court record of subpoenas and depositions.
It seems there are two cases that explain the result we have seen:
(1) they failed to come up with the resources to do the necessary discovery
(2) they did the discovery and came up empty
I have wandered into the land of two different penny stocks, because they came with "very positive news" visiting the Intel board. I had similar observations for both companies. Both started out with great ideas and some real products. They legitimately went public. Over time, their ideas failed to pan out into real revenue, and the financials began to degrade. As a result the "pumpers" show up in an attempt to get their money back for a failed investment, and perhaps there were a few "true believers". I don't think you can tell the true believers from the pumpers, but it doesn't really matter. I see these companies as different than "scammers" who have companies that never even intended to have real products, but rather were formed solely with the idea of fleecing investors.
Alan
I don't have much visibility into the HDVY financials, but I was under the impression they had maybe $1.5M from previous legal settlements. I assumed (perhaps incorrectly) that the lawyers for the INTC lawsuit were partially compensated from this, along with a contingency payment.
Of course it would be up to the legal firm to continue spending money for the potential of a higher future payout. It appears we have the answer to that one.
Alan
Absolutely the job of the law firm... but they are not going to do it for free. I suspect when it became clear Intel was not going to fold, the HDVY law firm let them know what would need to happen, and how much it would cost.
As an aside, I indeed had my office searched by the opposing lawyers at one point. They certainly do this, and it is usually unannounced.
I mean the evidence itself. I am not sure what you are referring to in "potential damage report". The fact that you call it "potential" seems to infer it is weak though. The original lawsuit paperwork did have enough evidence to get past summary judgement. OTOH, a trial with witnesses is a different thing.
By way of example, here are some of the things I remember from the original lawsuit paperwork. It is not at all clear to me who HDVY would call as witnesses and what they would say. They would need to sort through a lot of depositions to get to the right people with the right testimony.
Some of the potential testimony at trial that would have turned it into a strong case would be:
asked of the Intel authors of their SVM-RFE patent: "What products did you use to reduce the SVM-RFE patent to practice"
asked of Intel finance/business person: "how many units of product x did you sell, for how much revenue, selling price, and profit"?
Asked of the Intel author of the "mouse" paper: "Did you use SVM-RFE in developing a computer mouse? Which products incorporated the algorithm?"
asked of Intel finance/business person: "how many units of product x did you sell, for how much revenue, selling price, and profit"?
asked of the author of the paper that talked about running SVM-RFE on a Xeon processor.
"how many times did you run the algorithm?" What was Intel's return on your paper? How many additional Xeon units did they sell because of the paper? (perhaps they would need an outside marketing expert to testify to this).
Regarding the "intel indirect infringement" discussion we were having the other day:
What companies? What products did they make using SVM-RFE?
Will a person from that company testify that Intel "enabled" them?
"how many units of product x did you sell, for how much revenue, selling price, and profit"?
All of this testimony would be developed during discovery, and I have seen no evidence that HDVY did the work (nor could afford the work) necessary to develop this evidence.
During the discovery phase HDVY could depose Intel and others. They could also search Intel facilities. It is not clear how much of this was done. At trial, whatever they uncover during discovery could be presented as evidence.
HDVY did have some evidence that they presented in the original lawsuit filing. Perhaps this is all the evidence they could afford to collect. I know during the IPR's there was a small amount of expert testimony which could also be considered evidence for trial.
My opinion is the evidence presented in the lawsuit itself was pretty weak.
That is why I said "basically". Per HDVY, Intel was enabling/encouraging the usage of the algorithm which HDVY positioned as making it Intel's responsibility that others were using the algorithm without compensating HDVY.
Intel is not the only fish in the sea. They leave the Intel affair with a valid patent and evidence they will enforce it, even in court if required. Perhaps others who are using the algorithm will pay up with much less effort than was expended on Intel.
I have found this bit of the lawsuit (recently posted) quite amusing:
27. HDC accuses Intel of infringing its patents directly and indirectly through, for example, the use, sale, and marketing of "Intel AI-optimizing/machine learning processors," field programmable gate arrays, system on chips, and software deployed on "Intel/Intel-partnered computers" and other architectures
Basically, Intel was helping market the only product HDVY had, which was their patent. It was HDVY's responsibility to get royalties for the use of their patent.
It was a bit like suing microsoft for copyright infringement because people are using word to plagiarize.
I was surprised that there was no mention of royalty rates from either company. It would seem such a deal would have been a boon to HDVY, without much cost to Intel.
The notion that there is more to come from Intel is just wishful thinking; the matter is settled.