AmberWave Systems Corporation v. Intel Corporation

Good foundation. Now let me verify the settlement, AmberWave's corporate background, and downstream litigation/IPR history.

Now let me verify the patents' technical details and the downstream litigation/IPR history.

Case Overview — AmberWave Systems Corp. v. Intel Corp., No. 2:05-cv-00321 (E.D. Tex., Marshall Div.), transferred to and consolidated in Intel Corp. v. AmberWave Systems Corp., No. 1:05-cv-00301-KAJ (D. Del.)

The parties. AmberWave Systems Corporation was a Salem, New Hampshire, semiconductor-materials company — a classic MIT spin-out, though not a university or a litigation-only shell. Founded in 1998–99 by MIT materials-science professor Eugene Fitzgerald (widely described as the "Father of Strained Silicon") together with co-founders including Mayank Bulsara, it took an exclusive license to MIT technology, raised roughly $60–91 million in venture capital, and operated a ~30,000 sq. ft. R&D facility with roughly 20–23 employees (MIT SMA bio; Fitzgerald interview, MRS Bulletin, Jan. 2007; AmberWave testimony to House Small Business Committee, 2007). Its business model was pure IP commercialization — "researches, develops, patents and licenses advanced semiconductor materials technology, including strained silicon" (Intel/AmberWave settlement release, Mar. 5, 2007) — and by 2007 it claimed ~180 patents and licenses to seven chipmakers. So: a genuine technology developer and licensor with a royalty-based model (a "university spin-out licensor" rather than a classic PAE), suing the world's largest chipmaker. Intel Corporation (Santa Clara, CA; Delaware corp.) was accused in its capacity as the manufacturer of 90 nm Pentium processors. AmberWave's CEO Richard Faubert framed the suit as forced by Intel's refusal of a "commercially reasonable" license (EE Times, July 18, 2005; CNET, July 2005).

The accused technology and the patents. The accused products were Intel's 90 nm Pentium processors, which used strained-silicon engineering to boost carrier mobility as classical scaling ran out of steam (Intel admitted in its answer that "90nm Pentium® processors" were sold in the district — Doc. 37, Answer, ¶ 5). The patents in this case, per the case record: US 6,881,632 ("Method of Fabricating CMOS Inverter and Integrated Circuits Utilizing Strained Surface Channel MOSFETs," issued Apr. 19, 2005; inventors Fitzgerald and Nicole Gerrish) — a process claim to building a CMOS inverter/IC on a heterostructure of Si substrate / relaxed SiGe buffer / strained surface layer, with both n- and p-channel MOSFETs formed in the strained layer, using a compositionally graded buffer to accommodate lattice mismatch and put the silicon under biaxial tension (see the same-titled parent, US 6,649,480, abstract; Compound Semiconductor); and US 6,831,292 ("Semiconductor Structures Employing Strained Material Layers with Defined Impurity Gradients and Methods for Fabricating Same," issued Dec. 14, 2004; Appl. No. 10/251,424, filed Sept. 20, 2002, priority to Prov. 60/324,325 (Sept. 21, 2001); inventors Matthew Currie, Anthony Lochtefeld, Richard Hammond, and Fitzgerald; 62 claims) — a structure/method claim to a strained layer with a "distal zone" in which the impurity gradient is substantially zero, i.e., keeping the strained channel free of dopants that interdiffuse during high-temperature downstream processing (full text). A third patent, US 6,946,371 (facet-less epitaxially grown raised source/drain), was added to the consolidated Delaware dispute in late 2005 after issuing Sept. 20, 2005. Notably for non-infringement analysis: the asserted claims map most naturally onto global/biaxial strained-Si-on-relaxed-SiGe architectures, whereas Intel's 90 nm node was publicly reported as using local/uniaxial strain (SiGe source/drain for PMOS; see, e.g., Mistry et al., "Delaying Forever: Uniaxial Strained Silicon Transistors in a 90nm CMOS Technology," 2004 VLSI Symp.) — a mismatch that presumably animated Intel's invalidity and non-infringement defenses, though no court ever construed the claims. (My technical inference, not a documented ruling.)

Procedural posture and why it matters. AmberWave sent a cease-and-desist letter on May 9, 2005; Intel preempted with a declaratory-judgment action in the District of Delaware on May 17, 2005 (C.A. No. 1:05-c

Now let me dig into the Delaware consolidated action, the ITC/Federal Circuit parallel, and any PTAB or reexam proceedings on the two patents.

Let me pull the docket details and check for later PTAB/reexam activity on the two patents.

Scope note

The docket and the two substantive opinions establish the skeleton of this case clearly; the middle (discovery) is only partially visible from public sources, and the merits end was never judicially decided. Two structural points frame everything below:

1. The case the user-facing caption describes is the second-filed action. The lead case after transfer/consolidation is Intel's Delaware declaratory-judgment action, C.A. No. 05‑301‑KAJ (later 1:05-cv-00301-JJF-MPT), Intel Corp. v. AmberWave Sys. Corp. E.D. Tex. 2:05-cv-00321 was a transferor case that lived only ~3.5 months.
2. No Markman ruling, no summary-judgment ruling, no trial, no verdict, no JMOL/fee motion, and no appeal exist in this litigation. Settlement on 2007-03-05 preempted all of it. That is a finding, not a gap — I checked for each.

---

Chronological developments

2005-05-09/10 — AmberWave's notice letter (the trigger)

AmberWave's counsel Steven Frank sent Intel a letter (dated May 9 per Judge Jordan; May 10 per Intel's complaint) formally notifying Intel that its Prescott and Dothan 90nm processors infringed the '292 patent and the then‑pending '947 and '499 applications, and demanding license negotiations or cessation. Judge Jordan later characterized this as AmberWave "kick[ing] things off." (233 F.R.D. 416; Intel's DJ complaint)

2005-05-17 — Intel files first: Delaware DJ action on the '292 patent

Intel beat AmberWave to the courthouse by two months, filing a DJ action for non-infringement of US 6,831,292 in D. Del. before Judge Kent A. Jordan, expressly reserving rights to amend to add the '947/'499 applications upon issuance. Intel is a Delaware corporation and AmberWave is Delaware-incorporated, which made D. Del. a clean venue choice. (RPX docket summary; complaint)

2005-07-15 — AmberWave files the Texas case on the '632 patent

AmberWave sued in E.D. Tex., Marshall Division (Judge Leonard Davis), Case No. 2:05-cv-00321, asserting US 6,881,632 ("Method of Fabricating CMOS Inverter and Integrated Circuits Utilizing Strained Surface Channel MOSFETs," issued 2005-04-19) against the same 90nm Pentium products. This is the case in the caption. (Transfer Order, 2005 WL 2861476; CNET, July 2005)

Summer 2005 — Answer/counterclaim in Delaware

AmberWave answered Intel's DJ complaint and counterclaimed for infringement of the '292 patent, including a willfulness allegation. The willfulness paragraph is D.I. 11 ¶ 24 in the Delaware docket, cited by Judge Jordan in the Nov. 29 order. (233 F.R.D. 416) Precise filing date not verified from a primary source.

2005-09-02 — Intel's answer in the Texas case

Intel filed its Answer and jury demand in 2:05-cv-00321, admitting jurisdiction/venue facts but denying infringement, and pleading failure-to-state-a-claim defenses — effectively a merits denial plus a venue fight rather than a Rule 12(b) motion. (Intel's Answer)

2005-09-20 — The '371 patent issues, and the second race to the courthouse

The '499 application matured into US 6,946,371 ("Methods of Fabricating Semiconductor Structures Having Epitaxially Grown Source and Drain Elements"). Per Judge Jordan's account, AmberWave deposited a '371 infringement complaint in the E.D. Tex. drop box at 5:50 a.m. Central; Intel filed its '371 DJ complaint in Delaware (C.A. 05-682-KAJ) at about 8:30 a.m. Eastern — roughly two hours later. AmberWave thus "won" that race by clock time and lost it by docket priority.

2005-09-26 — Intel moves to supplement (D.I. 30, D. Del.)

Rather than fight the Texas '371 case, Intel moved under Rule 15(d) to supplement its Delaware complaint to add DJ claims of non-infringement and invalidity on the '371 patent.

2005-11-01 — Transfer granted (the key merits-adjacent ruling)

Judge Davis granted Intel's Motion to Transfer (D.I. 17), 2005 WL 2861476, applying the first-to-file rule to different patents between the same parties on the same accused products. Holding: substantial overlap "does not require that the core issues be identical, but that the two actions will involve closely related questions or subject

Press coverage of AmberWave v. Intel

The filing wave (July 2005). The case drew a burst of trade-press attention at filing, framed almost uniformly as a small IP house taking on the chip giant. EE Times reported on 2005‑07‑18 that AmberWave filed in Marshall, Texas on the '632 patent while filing a Delaware counterclaim on the '292 patent, and that the suits "arose after Intel refused to negotiate a commercially reasonable license agreement"; it quoted CEO Richard Faubert saying "we have no choice but to defend our intellectual property rights." InformationWeek ran essentially the same wire copy that day, and ZDNet published a version headlined "AmberWave files patent suit against Intel" that placed the dispute in context — noting it was "the latest blow to hit Intel, which only last month was slapped with an antitrust claim filed by rival chipmaker AMD." CNET was the only outlet in that wave to get a substantive Intel response on the record: spokesman Chuck Mulloy said "Intel does not believe in the validity of AmberWave's claims and we are planning a vigorous defense," and characterized Intel's Delaware suit as "a pre-emptive request … to judge the merits of the dispute without penalty." Compound Semiconductor went deeper technically than the wires, identifying Fitzgerald and Gerrish as the '632 inventors and explaining the graded SiGe buffer/biaxial-tension mechanism — though it blurred the procedural posture by stating flatly that "Intel's latest Pentium processors … infringe US patent numbers 6,881,632 and 6,831,292," when the '292 patent was actually being asserted as a counterclaim in Delaware. Tech Monitor added the company-profile gloss that AmberWave was a privately held Salem, N.H. licensor with "20 employees" that had "raised more than $66m in venture capital financing since it founded in 1998."

The quiet middle. Once the venue fight began, mainstream and legal-trade coverage largely dried up. I found no Reuters, Bloomberg Law, Ars Technica, The Verge, IPWatchdog, Patently-O, JD Supra or EFF Deeplinks article on this litigation in any of my searches, and nothing in the accessible press reporting on Judge Davis's November 1, 2005 first-to-file transfer or on the September 2005 race to the courthouse over the '371 patent. Law360's archive does list a headline — "AmberWave Sues Intel Over Strained-Silicon Technology" — on its reprints/rights page (Law360), but the article body is paywalled and I could not open or date it, so I cannot characterize its contents. The only substantive contemporaneous treatment of the transfer ruling I located is in the secondary legal databases (Casetext, Casemine), which reproduce the opinion rather than report on it. Practically speaking, between mid‑2005 and March 2007 the case lived in the docket and in academic/policy commentary, not in the press.

The settlement (March 2007). Coverage resumed on the deal. EE Times reported on 2007‑03‑05 that Intel "has settled all patent infringement suits pending with AmberWave" and "received a license to all AmberWave patents and patent applications either existing today or filed during the agreement's 10-year term," with Intel making license payments over the term and all other terms undisclosed. CNET carried the same essentials the same day, and Semiconductor Today (2007‑03‑08) and Solid State Technology/Semiconductor Digest (2007‑03‑06) followed. The Digest piece is the most analytically useful: it flagged that the license covering future applications was "intriguing," and noted pointedly that "[t]here was no explanation whether/how much of AmberWave's IP includes local strain vs. global strain, or if the 10‑year deal includes future nodes" — i.e., the trade press noticed the same doctrinal gap (Intel's 90 nm parts used local/uniaxial strain) that would have driven the merits. ITNews and TechPartner (both by Tom Sanders, 2007‑03‑06/07) added the licensing-portfolio context that AmberWave had publicly disclosed a deal with LG Siltron but few others. No outlet anywhere reported a dollar figure — not a damages demand, not a settlement value. Every account I found attributes the silence to the confidentiality clause in Intel's March 5, 2007 press release, which is the primary source all of the settlement stories tracked almost verbatim.

Where accounts differ, or go beyond the record. Three divergences worth flagging. (1) The DJ filing date: CNET, EE Times, Tech Monitor and ZDNet all state Intel filed its declaratory-judgment action on May 25, 2005; Intel's own DJ complaint and Judge Davis's transfer opinion put it at May 17, 2005 (Casetext). The press date appears to be a propagated error — prefer the docket. (2) Outcome characterization: a Nashua Telegraph feature by David Brooks (2008‑01‑20) referred to "when AmberWave won its suit against Intel" and said that under proposed apportionment rules "its settlement would undoubtedly have been a lot smaller." That is a local-paper gloss on a settled case that produced no judgment of infringement or validity; no other outlet described it as a win, and the court record contains no such finding. (3) Retrospective industry framing: an EDN column by Ron Wilson (2007‑06‑22) described AmberWave's model as mining "—the word 'trolls' also comes to mind—" university research and stated that AmberWave "did some initial development, and licensed the idea to, among others, Intel," recasting the adversary as a licensee. That framing was reinforced by AmberWave's own Washington activity: general counsel Bryan Lord used the case as a case study in opposing the 2007 Patent Reform Act, in testimony to the House Small Business Committee (2007‑03‑29) — i.e., the party most invested in the case's significance was itself the loudest voice characterizing it, which is worth weighing when reading any of the settlement coverage.

Sources

• AmberWave files patent suit against Intel - ZDNET
• Intel hit with chip tech lawsuits - CNET - Skip to content
• Amberwave Systems Corp. v. Intel Corporation - From Casetext: Smarter Legal Research
• AmberWave sues Intel citing latest Pentium chips - News - Loading
• Intel and AmberWave Systems settlement strained silicon patent infringement - News
• Electronics News & Analysis - EE Times
• Intel settles patent dispute with AmberWave Systems
• AmberWave Systems Corporation's Responses to Intel's Sixth Set of Interrogatories, U
• Intel's Responses and Objections to AmberWave's Second Set of Interrogatories, U
• AmberWave is a small firm focused on research and development of advanced materials
• Semiconductor structures employing strained material layers with defined impurity gradients and methods for fabricating same - JP2005504436A (en) * | 2001-09-21 | 2005-02-10 | アンバーウェーブ システムズ コーポレイション | A semiconductor structure using a strained material ...
• Free Letter - District Court of Delaware - Delaware

Good leads. Let me pull the actual dockets and confirm the appearances.

Let me open the full docket sheets for both the Texas case and the transferred Delaware case.

Counsel of Record for AmberWave Systems Corporation

Framing note (procedural wrinkle worth flagging up front). Because the two actions were consolidated, "counsel for the plaintiff" is not a single coherent list. In the first-filed Delaware action (Intel Corp. v. AmberWave, C.A. No. 1:05-cv-00301-KAJ), Intel was the plaintiff and AmberWave the defendant/counterclaimant; in the transferred Texas case (No. 2:05-cv-00321-LED, which became D. Del. C.A. No. 1:05-cv-00837-MPT), AmberWave was the plaintiff. AmberWave's legal team was the same across both, and it was a two-tier structure: an out-of-state lead firm (Irell & Manella LLP) plus Delaware and Texas local counsel.

Lead / trial counsel — Irell & Manella LLP (Los Angeles, Century City; 1800 Avenue of the Stars)

Source: AmberWave's approved pro hac vice applications on the E.D. Tex. docket, as reflected in the PlainSite docket for No. 2:05-cv-00321; Delaware PHV motion and of-counsel signature blocks (D.I. 15; D.I. 129, June 27, 2006; D.I. 74, Dec. 26, 2005).

Delaware local counsel — Morris, Nichols, Arsht & Tunnell LLP (Wilmington, DE)

The Nov. 29, 2005 memorandum order in 05-301 confirms the lineup from the bench's perspective: "Jack B. Blumenfeld, Leslie A. Polizoti, Morris, Nichols, Arsht & Tunnel, Wilmington, DE, for defendant" (against Young Conaway for Intel) — CourtListener.

Texas local counsel — Law Office of T. John Ward, Jr., P.C. (Longview, TX)

In-house

Bryan Lord, AmberWave's Vice President of Licensing and Finance and General Counsel, was the company's senior in-house lawyer (AmberWave fact sheet). I found no docket entry showing him entering an appearance; treat him as the client-side contact rather than counsel of record.

---

Caveats and open items

• No sealed or missing appearances for plaintiff's side. The AmberWave appearances are all documented in public docket entries and signature blocks; the only sealed items I encountered were substantive (e.g., a sealed sur-reply on the transfer motion), not appearance-related.
• The Delaware docket in 05-301 lists AmberWave's counsel as "defendant." If you need a strict "plaintiff-side" list, use the Texas docket (2:05-cv-00321) and the transferred 05-837 filings; Blumenfeld's Dec. 8, 2005 notice is the operative plaintiff-side Delaware appearance.
• Contradiction/loose end to flag: Giza's published bio cites the AmberWave–Intel matters as "No. 1:06-cv-638, 1:06-cv-429 (D. Del.), 9:06-cv-157 (E.D. Tex.)." Those numbers do not match the case numbers in our metadata (05-301 / 05-837 / 05-321), and I could not verify them; they likely refer to later 2006 follow-on AmberWave–Intel suits over additional patents, or are simply garbled in the bio. Do not rely on them without a PACER check.
• I did not find evidence of any second outside firm (e.g., a separate IP boutique) sharing the AmberWave representation beyond Irell, Morris Nichols, and the Ward office.

Strong leads. Let me pull the actual docket sheets and appearances for both cases, and verify firm affiliations.

Now let me pin down the Texas-side appearances and verify each attorney's background.

Intel's counsel of record — AmberWave Sys. Corp. v. Intel Corp., No. 2:05-cv-321-LED (E.D. Tex.), and Intel Corp. v. AmberWave Sys. Corp., No. 1:05-cv-301-KAJ/-JJF-MPT (D. Del., consolidated)

Bottom line. Intel used a two-tier structure throughout: a single national/IP lead firm — Simpson Thacher & Bartlett LLP (Palo Alto) — paired with a different local firm in each forum. In Marshall, the local/attorney-in-charge was Michael E. Jones of Potter Minton (Tyler); after the November 1, 2005 transfer, Potter Minton disappears from the record and Young Conaway Stargatt & Taylor LLP (Wilmington) carried the Delaware local-counsel role, with Simpson Thacher as of counsel. Not one Intel in-house lawyer appears as counsel of record in any document I found. For contrast (and to avoid confusion, since both firms' names appear throughout this docket): Morris Nichols Arsht & Tunnell and Irell & Manella were AmberWave's counsel, not Intel's.

Summary table

Texas phase (2:05-cv-321-LED, Judge Leonard Davis, Marshall)

The operative document is Intel Corporation's Answer, Doc. 37, dated September 2, 2005, signed:

"By /s/ Michael E. Jones, Attorney-In-Charge, Texas Bar No. 10929400 … POTTER MINTON, 110 North College, 500 Plaza Tower, Tyler, Texas 75702 … OF COUNSEL: George M. Newcombe (CA Bar No. 202898) … SIMPSON THACHER & BARTLETT LLP, 3330 Hillview Avenue, Palo Alto, California 94304 … Attorneys for Defendant Intel Corporation."

So the Texas allocation was: Potter Minton as local counsel and attorney-in-charge of record, Simpson Thacher as the substantive/IP team. Jones also signed the transfer motion papers and supplied the supporting declaration, and he recurs as Intel's East Texas local counsel in later Marshall matters — consistent with Potter Minton's role as a standard East Texas local-counsel shop for out-of-state defendants.

Docket anomaly worth flagging: the Doc. 37 image carries two file stamps — "Case 1:05-cv-00837-***-MPT Document 16 Filed 09/02/2005" (D. Del.) and "Case 2:05-cv-00321-LED Document 37 Filed 12/05/2005" (E.D. Tex.). The September 2 signature date is consistent with the Delaware stamp, but the Texas stamp date (December 5, 2005) post-dates the November 1 transfer, which is chronologically odd. The most likely explanation is that a Delaware-filed pleading was later entered on the Texas docket, but I cannot confirm the docketing history from a free source; treat the Texas entry date as unverified.

Delaware phase (1:05-cv-301-KAJ → -JJF-MPT, Judges Jordan, then Farnan/Thynge)

Intel was plaintiff and counterdefendant here (the May 17, 2005 DJ action on the '292 patent, later supplemented to add the '371 patent). Its filings are uniformly signed by the Young Conaway trio with a Simpson Thacher "Of Counsel" block:

• D.I. 30, Intel Corporation's Motion for Leave to Supplement Complaint (Sept. 26, 2005): "YOUNG CONAWAY STARGATT & TAYLOR LLP / Josy W. Ingersoll (No. 1088) / John W. Shaw (No. 3362) / Karen E. Keller (No. 4489) … Attorneys for Plaintiff and Counterdefendant Intel / Of Counsel: George M. Newcombe, Simpson Thacher & Bartlett LLP."
• D.I. 194, Notice re subpoenas to Dinesh Mehta and Nicole D. Gerrish (Nov. 16, 2006): signed by Ingersoll, Shaw and Keller; "Of Counsel: George M. Newcombe, Patrick E. King, Simpson Thacher & Bartlett LLP."
• AmberWave's certificates of service (e.g., Feb. 28, 2006; Mar. 28, 2006) list Intel's side as Ingersoll/Shaw/Keller at Young Conaway and Newcombe/Ostrow/King at Simpson Thacher, plus LaRocca and Anderson in New York.

AmberWave's side — for disambiguation only — was Morris Nichols Arsht & Tunnell (Jack B. Blumenfeld, No. 1014; Leslie A. Polizoti, No. 4299) with Irell & Manella of counsel (Morgan Chu, Amir A. Naini, David I. Gindler, Jason G. Sheasby), as shown by the August 5, 2005 pro hac vice motion (D.I. 15) and later notices.

What I could not verify

• No Intel in-house counsel of record. I found no Intel legal-department attorney signing any filing or entering an appearance. (The Intel personnel named in AmberWave's May 9, 2005 cease-and-desist letter — Brad Greenwald, Reza Afghan, Farshid Adibi-Rizi, Reed Stahler of Intel's IP licensing group — were business/licensing contacts, not counsel of record, and should not be listed as such.)
• No formal pro hac vice orders located for Newcombe, King or Ostrow in either court, and no separate "notice of appearance" for the Simpson Thacher lawyers in Texas. Their status in Texas rests on the "Of Counsel" designation in the Answer rather than an admission order I could see. It is possible admissions were handled administratively or the documents are simply not in the free mirrors.
• Not sealed. Everything above comes from publicly available, ECF-stamped pleading images and a published opinion (Intel Corp. v. AmberWave Sys. Corp., 233 F.R.D. 416 (D. Del. 2005)). I saw no indication that appearances in this case were filed under seal; the settlement terms were confidential, but the appearances were not.
• Biographical detail is thin. I was unable to run the firm-bio checks I would normally use to characterize each lawyer's broader patent practice, so I have deliberately not attributed specific prior cases to any of them. Jones's East Texas patent-defense practice, Ingersoll's Delaware IP litigation practice, and Newcombe's role as Simpson Thacher's Palo Alto IP litigation partner are consistent with everything in this record, but the only hard, sourced facts I have are their titles, bar numbers, firm addresses and the documents they signed in this case. Keller and Shaw later left Young Conaway to found Shaw Keller LLP — I believe that is right, but I could not confirm it in this research pass and it should be verified before publication.
• Contradiction with the case metadata, flagged: the metadata lists the case as "terminated 2005-11-01," which reflects only the Texas docket's transfer date. The consolidated Delaware proceeding (1:05-cv-301) remained open until March 5, 2007, and RPX shows a last docket entry of October 8, 2008 (RPX Insight). Related Intel-side filings also appear under D. Del. C.A. No. 05-682-KAJ (the '371 DJ action) and a 05-837 docket — counsel there appears to be the same Young Conaway / Simpson Thacher team, but I could not open those dockets to confirm.