US 9310855

I'll search for authoritative information on US 9,310,855, including any 2026 CAFC activity.

Let me verify whether 9,310,855 itself appears in any litigation or 2026 Federal Circuit docket, since my first searches surfaced the Valtrus portfolio generally but not this specific patent.

Let me check specifically for any 2026 Federal Circuit docket involving this patent.

US Patent 9,310,855 — Summary

Search scope note: I searched for the specific number 9,310,855 (US13/703,911) and did not auto-correct or substitute any similar numbers. I found strong evidence of district-court litigation involving this patent, but no Federal Circuit docket from 2026 that specifically concerns the '855 patent. Details and caveats below.

---

Bibliographic data (from the patent front page / Google Patents)

Abstract (verbatim): Flexible data center systems and methods of deployment are provided. A flexible data center (100, 200) including T rows of server racks (106, 206, 306) can be deployed by constructing a number B of blocks (101, 201). Constructing each block (101, 201) can include constructing from one to a number P of perimeter structures (104, 204, 304) each housing up to a number R of rows of server racks (106, 206, 306). Constructing each block (102, 201) can include constructing a connecting structure (102, 202) connected to the perimeter structures (104, 204, 304), the connecting structure (102, 202) housing operations monitoring equipment (108, 208) for the server racks (106, 206, 306). A total integer number T/R of perimeter structures (104, 204, 304) can be constructed for the flexible data center (100, 200). At most one perimeter structure (10, 204, 304) houses less than R rows of server racks (106, 206, 306). B is equal to an integer number (T/R)/P. At most one block (101, 201) includes less than P perimeter structures (104, 204, 304).

---

Plain-language overview of the independent claims

Claim 1 — Method of deploying a flexible data center (T rows of server racks)

1. Build "blocks" modularly. Construct a number B of blocks on a site. Each block = 1 to P "perimeter structures" (each holding up to R rows of server racks) plus a "connecting structure" attached to those perimeter structures that houses the operations-monitoring equipment for the racks.
2. Follow a fill-before-expand math rule. Build a total of T/R perimeter structures (rounded up), where each perimeter structure consumes up to some quantity of power. Constraints: at most one perimeter structure holds fewer than R rows, B = (T/R)/P (rounded up), and at most one block has fewer than P perimeter structures. In practice this means you fully fill existing structures/blocks before adding new ones.
3. Power design + validation. Select a number of power conditioner units per perimeter structure according to a desired power quality, validate that selection with an energy-optimization evaluation based on the site's climate, and connect the units.
4. Cooling selection + validation. Provide a menu-driven process for choosing a cooling-unit type from a predetermined set, select the type according to the site climate, validate it with an energy-optimization evaluation, and attach the selected cooling units to the exterior of each perimeter structure.

Note: the granted claim 1 is narrower than published application claim 1 (US2013/0094136 A1) — the power-conditioner validation and menu-driven cooling-selection/validation limitations appear to have been added during prosecution.

Claim 8 — The flexible data center itself (apparatus claim)

A data center with T rows of server racks, comprising:

• B blocks on a site, each with 1 to P perimeter structures (each housing up to R rows of racks) and a connecting structure housing operations-monitoring equipment — with the key added requirement that the perimeter structures retain functionality independent of the connecting structure (i.e., the connecting structure is not a single point of failure);
• the same T/R, "at most one structure < R rows," B = (T/R)/P, and "at most one block < P structures" relationships;
• a number of cooling units on the exterior of each perimeter structure, the type being particular to the site climate; and
• a number of power conditioner units on the exterior, the type being particular to a desired power quality and the site climate.

Notable dependent claims: claim 2 (P = 4, "butterfly"/"H" configuration); claim 9 (closed lists of cooling-unit and UPS types); claim 11 (B = 1, P = 4 → 3.2 MW IT power, 400 kW network/comms, PUE < 1.28); claim 12 (perimeter structures exclude a power distribution unit).

---

Litigation / CAFC 2026 search

• District court (confirmed, this specific patent): 9,310,855 has been asserted in E.D. Tex. cases by Valtrus Innovations Ltd. The Google Patents "family has litigation" entries link E.D. Tex. cases 2:24-cv-00361, 2:24-cv-00139, 2:24-cv-00142, 2:24-cv-00259, 2:24-cv-00534, 2:24-cv-00535, and 2:25-cv-00323. CourtListener records confirm the '855 patent appears as an exhibit and in a claim chart in Valtrus v. NTT Data Services, LLC, No. 2:24-cv-00361 (E.D. Tex., filed 2024-05-14), and a CyrusOne counterclaim (Dkt. 66 in the consolidated E.D. Tex. action) seeks a declaratory judgment of non-infringement of the '855 patent. Aggregator summaries also place the '855 patent in Valtrus v. Digital Realty Trust (2:24-cv-00139) and Valtrus v. TierPoint (2:25-cv-00323), where the parties reportedly dismissed with prejudice in November 2025 with no public validity ruling. I did not verify these aggregator descriptions against primary dockets.
• Federal Circuit 2026: The only 2026 CAFC activity I found for Valtrus is Appeal No. 2024-1945, Google LLC v. Valtrus Innovations Ltd., with oral argument audio posted March 4, 2026 (cafc.uscourts.gov). Based on the underlying N.D. Cal. opinion, that appeal concerns the search-engine patents ('704 and '604), not the '855 patent. I found no 2026 CAFC docket, appeal, or order specifically addressing US 9,310,855. If such an appeal exists, it was not surfaced by my searches — treat this as a "no evidence found," not a definitive negative.

Uncertainty flags: (1) The 2032-02-29 adjusted expiration is as listed by Google Patents and would reflect a term extension beyond the nominal 20-year term from the 2010-07-12 filing date — I have not independently verified the PTA/PTE calculation. (2) Current assignee/standing is contested in the Valtrus litigation (courts have found Valtrus lacks "all substantial rights," with Key Patent Innovations Limited joined as an interested party), so the operational "owner" of enforcement rights is legally nuanced.

I'll search for litigation involving US 9310855. Let me start with several targeted queries.

Let me dig deeper into the specific cases.

Known Litigation Involving U.S. Patent No. 9,310,085

Important framing note: I searched litigation-specific sources (Stanford NPE Litigation Database, Unified Patents, RPX Empower, CourtListener/RECAP, PACER Monitor, Google Patents litigation links, PatSnap, and PTAB filings). All results point to a single patent-assertion campaign: Valtrus Innovations Ltd. (with Key Patent Innovations Ltd. as beneficial owner/co-plaintiff) asserting a data-center patent portfolio — including the '085 patent — against data-center operators in the Eastern District of Texas, Marshall Division, before Judge Rodney Gilstrap. The Google Patents litigation record for this patent lists exactly seven related E.D. Tex. case numbers; I treat those as authoritative for this patent and do not extrapolate to other Valtrus/Gilstrap cases that may assert different patents.

Also note: the patent record shows the family is subject to litigation (Darts‑ip family ID 45469719), and later Valtrus matters were routed into MDL No. 3190 (E.D. Tex.). I did not find any Federal Circuit appeal or Supreme Court proceeding specific to the '085 patent, and no IPR specifically challenging the '085 patent.

---

Cases listed on the '085 patent record (E.D. Tex.)

---

Notable contextual details confirmed in the record

• Owner/plaintiff identity: Valtrus Innovations Ltd. is described in its own pleadings as successor-in-interest to an HPE patent portfolio. The '085 patent's assignment chain on the record is: Hewlett-Packard Development Co. → Hewlett Packard Enterprise Development LP (2015) → OT Patent Escrow, LLC (2021) → Valtrus Innovations Limited (Feb. 2021).
• Intervenor: Vertiv Corporation, the accused equipment supplier (e.g., Liebert cooling / iCOM control systems), intervened across multiple of these cases asserting contractual indemnity obligations to its customers, and separately filed a declaratory-judgment action that was consolidated into case 2:24-cv-00361.
• '085-specific assertion: The clearest '085-specific infringement pleading I found is in 2:24-cv-00534 (CyrusOne — "flexible data centers," DFW1) and 2:24-cv-00142 (Dawn/Evoque — asserted independent claim 8 of the '085 patent). PatSnap also lists US 9,310,855 B2 among the patents asserted in the CyrusOne and TierPoint matters.
• No outcome on the merits in any case: every disposition I located was a joint/consensual dismissal with prejudice (with "bear own costs" terms), i.e., no judicial finding on infringement, validity, or damages.

---

Things I could not confirm (stated explicitly rather than guessed)

• Exact filing dates for 2:24-cv-00139 and 2:24-cv-00535. Sources show early‑2024 filing for the -00139 case; I did not obtain a reliable single date for either.
• Whether the '085 patent was formally asserted in the lead case 2:24-cv-00361 (as opposed to merely being listed in the consolidated docket). One PTAB paper indicates the '085 patent was "no longer at issue."
• Any later (2025–2026) Valtrus cases — e.g., Valtrus v. STACK Infrastructure (2:26-cv-00287) and Valtrus v. FifteenFortySeven Critical Systems Realty (2:26-cv-00820, transferred into E.D. Tex./MDL 3190) — do not, on the records I reviewed, list the '085 patent; those appear to assert other patents (e.g., '287, '682).
• Any IPR, PGR, or reexamination challenging claims of the '085 patent. The Valtrus-related PTAB petitions I found target other patents (e.g., IPR2025-00668 on the '277 patent; IPR2025-01157 on the '967 patent).
• Any appeal to the Federal Circuit or petition to the Supreme Court involving the '085 patent.

Sources consulted: Google Patents litigation record for US 9,310,855 (https://patents.google.com/patent/US9310855/en); Stanford NPE Litigation Database (https://npe.law.stanford.edu/patent/9310855); RPX Empower (https://litigation.rpxcorp.com/litigation/txedce-[230090](/patent/230090)-valtrus-innovations-v-ntt-data-services); CourtListener/RECAP dockets for 2:24-cv-00361 (docs #83, #96, #117, #151); PACER Monitor for 2:26-cv-00820; PTAB petition filings (ptacts.uspto.gov); PatSnap litigation summaries for Valtrus v. TierPoint, Valtrus v. Digital Realty Trust, and Valtrus v. CyrusOne.

If you need primary-source verification of any single row (particularly the filing dates for -00139 and -00535), I'd recommend pulling the PACER/RECAP docket sheets directly; I flagged those two dates because I could not pin them down confidently.

I'll verify the structured "no proceedings" block against web sources before finalizing.

Let me run a few more targeted checks for any IPR naming this patent number.

One more pass to be thorough on PTAB dockets and any ex parte reexamination activity on this patent.

PTAB Proceedings — US 9,310,855 B2 ("Flexible data center and methods for deployment")

Proceedings overview

No AIA trial proceedings are on file for US 9,310,855. The USPTO ODP structured block reports zero IPR, PGR, or CBM proceedings, and I could not corroborate any petition against this patent through web search — so the total count is 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials. Bottom line for a defendant: there is no PTAB determination of any kind to lean on. All 12 issued claims (1–12) are untested at the Board, the patent has never been narrowed by an AIA trial, and a defense built on "the PTAB already killed this claim" is unavailable here. The absence of any petition is itself the headline: this patent has been in force since 2016-04-12 and has now been asserted in multiple district court suits without ever drawing a single IPR.

No proceedings to report

The structured "PTAB proceedings on file" block is the canonical list and it is empty for this patent. I searched for IPR/PGR activity indexed to the patent number, to the current owner of record (Valtrus Innovations Ltd), and to the sibling Valtrus data-center patents, and found no petition directed at the '855 patent. Per the task constraints I will not invent a proceeding number to fill this section — there is nothing to list. If a petition surfaces in a later ODP ingest or a PTAB E2E docket search (https://ptacts.uspto.gov/), it should be treated as newly filed.

Adjacent PTAB activity — different patents, useful context only

To be explicit: none of the proceedings below challenges US 9,310,855. They are cited only because they show how the same patent owner and the same defendants have behaved at the Board, which informs how a petition against the '855 patent would likely be handled.

• IPR2025-00414 — SAP America, Inc. / SAP SE v. Valtrus Innovations Ltd. (US 7,152,182). Filed by SAP; petition challenges claims 1–18 under pre-AIA § 103. This is Valtrus's data‑redundancy patent, not the '855 patent.
• IPR2025-00419 / IPR2025-00420 — SAP v. Valtrus (US 7,936,738 and US 6,899,244 respectively). Both petitions drew a substantial Fintiv / § 314(a) fight, with SAP arguing the parallel E.D. Tex. actions were at an advanced stage.
• IPR2025-00667 — Vertiv Corporation v. Valtrus Innovations Ltd. (US 6,854,287). Vertiv (the cooling-equipment OEM that indemnifies the data-center defendants) asserted § 315(a)(2) and RPI/privity arguments premised on its indemnity of the accused customers.
• Litigation-analysis note: Unified Patents has publicly described the Valtrus campaign as having generated "a dense landscape of district court litigation, IPRs, declaratory judgement and reexamination activity," and its members-only report includes invalidity and ex parte reexamination work product for one Valtrus patent. I found no public indication that the '855 patent is that patent. Treat this as a lead to verify, not a confirmed '855 reexam.

Strategic summary

Claim status: everything is untested. Claims 1–12 of US 9,310,855 are exactly as issued — independent claims 1 (method of deploying a flexible data center, including the menu-driven cooling-unit selection and power-conditioner validation steps) and 8 (the flexible data center apparatus), plus dependents 2–7 and 9–12. Because no IPR was ever instituted, there is no cancelation of any claim, no substitute claim, and no FWD. Any assertion of claims 1–12 today must be met in district court or in a new petition — there is no PTAB record to import as collateral estoppel.

Estoppel landscape: essentially empty, but so is the petitioner runway. Section 315(e)(2) estoppel binds only petitioners who were parties to an instituted IPR; since nobody has been, no defendant is estopped and no prior-art ground has been "used up" at the PTAB. The constraint is different: § 315(b) time-bar. Valtrus (with Key Patent Innovations Ltd.) filed the '855-related complaints in E.D. Tex. beginning in January 2024 (e.g., 2:24-cv-00139 and 2:24-cv-00142) and continuing through 2024–2025 (2:25-cv-00323; the NTT Data Services lead case 2:24-cv-00361). For any defendant served in those waves, the one-year § 315(b) window has closed, which is a likely reason no IPR ever issued. A defendant served now who still has runway would additionally face the Office's 2025 "settled expectations" discretionary-denial practice: with the patent more than six years past grant (issued 2016-04-12; today is 2026-09-29), a fresh petition faces a real risk of summary denial on discretionary grounds without any merits review.

Pattern signals. Same-petitioner multiples exist across the Valtrus portfolio (SAP filed a cluster of IPRs in the IPR2025-00414/‑00419/‑00420 range; Vertiv filed its own against the '287 patent), but nothing against the '855 patent. The patent owner has not had to defend a single PTAB appeal for this patent, so there is no CAFC docket to report. The most consequential non-PTAB development for a defendant is the standing fight: in the E.D. Tex. action, the court found Valtrus transferred substantial rights to Key Patent Innovations and does not satisfy 35 U.S.C. § 281 (KPI had to be joined); and in the N.D. Cal. Google action, the court dismissed Valtrus's case without leave to amend for lack of constitutional standing (https://www.govinfo.gov/content/pkg/USCOURTS-cand-5_25-cv-07063/pdf/USCOURTS-cand-5_25-cv-07063-1.pdf). Several '855-adjacent suits have also ended in dismissal with prejudice (TierPoint, No. 2:25-cv-00323; Digital Realty, Nos. 2:24-cv-00139 and 2:24-cv-00535), and the JPML centralized the Valtrus data-center campaign into MDL No. 3190 in the Eastern District of Texas (https://www.jpml.uscourts.gov/sites/jpml/files/MDL-3190-Transfer_Order-7-26.pdf).

Recommended next steps

1. State the negative plainly in any opinion or defense memo: "No PTAB proceeding has ever been filed against US 9,310,855; the patent is un-narrowed and every claim is presumed valid." A demand letter citing claims 1–12 remains fully live — there is no cancelled claim to point to.
2. Before spending on a petition, check the § 315(b) clock. If your client or its privy was served more than one year before the intended filing date, the petition is dead on arrival. If you are within the window, budget for a discretionary-denial fight under the Office's settled-expectations practice given the 2016 grant date.
3. Because the '855 patent may be IPR-foreclosed in practice, build the invalidity case in district court, where system/prior-use art, § 112 (written description and enablement), and § 101 are all available and are not subject to IPR's patents-and-printed-publications limit. Vertiv's invalidity contentions in the parallel actions (asserting § 112 ¶ 1 enablement and written-description theories against the Valtrus cooling patents) are a useful template.
4. Verify standing and ownership before the merits. The § 281 / constitutional-standing rulings in the Valtrus campaign are a threshold, potentially case-dispositive issue: confirm the current chain of title (HPE → OT Patent Escrow → Valtrus Innovations Ltd) and whether Key Patent Innovations must be joined as a plaintiff.
5. Monitor the MDL and ODP. Track MDL No. 3190 in E.D. Tex. and re-run the ODP PTAB proceedings query periodically; if a petition against the '855 patent is filed, the key milestones will be the POPR (3 months), the institution decision (statutory 3 months after POPR, extended by the 2025 bifurcated discretionary process), and the FWD (statutory 1 year from institution, extendable to 18 months).

Confidence note: I am confident that no PTAB proceeding on this patent appears in the structured ODP data. My web corroboration was substantive but not exhaustive of PTAB E2E; if a recently filed, not-yet-indexed petition exists, it would not appear in the sources I could reach. Nothing above should be read as a statement about claim outcomes, because there are none.

I'll reconstruct this ownership chain. First, let me gather the assignment records, litigation data, and NPE-directory entries I need to ground the analysis.

Good — I have the escrow/trust structure and litigation. Let me pin down the recorded correspondents and reel/frame details, plus the inventors' employer context.

I have the escrow chain. Let me verify the recorded correspondents and the litigation dates.

Ownership & Assignment Reconstruction — US 9,310,855 B2

Patent: "Flexible data center and methods for deployment" · App. 13/703,911 (371(c) national stage of PCT/US2010/041734) · Filed 2010-07-12 · Issued 2016-04-12 · Status: Active, adjusted expiration 2032-02-29 (4th-yr fee paid 2019-09-23; 8th-yr fee paid 2023-10-04)

Primary verification link: USPTO Assignment Center (search patent number) / legacy Patent Assignment Search

---

Inventors

Basis for employer attribution: All seven inventors executed a single assignment to Hewlett-Packard Development Company, L.P. with signing dates from 2010-07-09 to 2010-07-12 — i.e., simultaneously with, or days before, the 2010-07-12 PCT filing date (recorded at Reel 029475/0187). Simultaneous signing on the filing date is the standard employment-work-product pattern. I did not independently verify each inventor's job title, so treat "HP employee" as strongly indicated rather than confirmed from a primary HR source.

Unusual patterns: None found. There is no evidence in the record of inventors departing the original assignee within 12 months of filing, and no individual (non-employer) assignments, no inventor-retained rights, and no reversionary interests. The chain shows one clean bulk inventor→HP assignment; nothing that resembles the pre-fire-sale inventor-exodus pattern.

---

Original assignee

Hewlett-Packard Development Company, L.P. (later Hewlett Packard Enterprise Development LP after the November 2015 HP split).

• Primary line of business: Computer hardware, enterprise servers/storage, and data-center infrastructure — an operating industrial company, not an IP holding vehicle.
• Product embodying the claims: HP was actively commercializing modular/factory-built data-center products in this window (e.g., the HP Performance Optimized Datacenter "POD" line), and this patent's disclosure reads on that product family. I could not confirm from the record a specific shipped product that maps to the issued claims, so I flag this as "likely commercialized, unverified."
• Current status: Operating. HPE remains a publicly traded operating company (NASDAQ: HPE). No bankruptcy, dissolution, or insolvency event is associated with the original assignee.
• Note on Google Patents' "Current Assignee" field: it lists both "Valtrus Innovations Ltd" and "Hewlett Packard Enterprise Development LP." The HPE entry is stale/an inaccurate rendered field; HPE divested this patent in 2021 (see below).

---

Assignment timeline

Four recorded assignments are on the patent's face. Two of them (the 2021 pair) were bulk recordings covering a large HPE portfolio, so the same conveyance can appear under more than one reel/frame depending on which recording instrument a given patent's chain cites — that discrepancy is flagged below rather than smoothed over.

---

2010-07-09 → 2010-07-12 (executed) / recorded 2012-12-14 — Reel 029475/0187

• Conveyance: Assignment of assignors' interest
• Assignor: Kfir L. Godrich; Peter Gross; Scott A. Wilson; James Zuppan; Christopher W. Kurkjian; Gregory K. K. Young; Peter F. Gmiter (individually)
• Assignee: Hewlett-Packard Development Company, L.P., Texas
• Correspondent: Not recoverable from the record I can access. The Google Patents legal-events entry does not expose a correspondent, and I did not locate the original cover sheet. I am not going to invent one.
• Context: Ordinary employment assignment of inventors' rights to the operating employer, executed at filing.

---

2015-10-27 (effective) / executed 2015-11-09 / recorded 2015-11-09 — Reel 037079/0001

• Conveyance: Assignment of assignor's interest
• Assignor: Hewlett-Packard Development Company, L.P.
• Assignee: Hewlett Packard Enterprise Development LP
• Correspondent: Not recoverable from the record I can access.
• Context: Internal corporate reorganization — the November 2015 HP Inc./Hewlett Packard Enterprise split. No consideration-driven transfer of ownership between unrelated parties; a pure intra-group name/entity migration.

---

2021-01-15 (executed) / recorded 2021-01-26 — Reel 055269/0001

• Conveyance: Patent Assignment, Security Interest, and Lien Agreement (assignment into escrow)
• Assignor: Hewlett Packard Enterprise Development LP (Texas LP) and Hewlett Packard Enterprise Company (Delaware corp.), jointly as "Assignor"; signed by Brett Alten, SVP, DGC, Chief IP Counsel (Jan 20, 2021)
• Assignee: OT Patent Escrow, LLC, Delaware LLC, 200 West Madison, 37th Floor, Chicago, IL 60606; signed by Joel Lutzker (Jan 21, 2021)
• Correspondent: Richard J. Botos — Botos Churchill IP Law LLP, 430 Mountain Avenue, Suite 401, New Providence, NJ 07974 (signed /Richard J. Botos/, May 24, 2024) on the assignment cover sheet that lists both Reel 055269/0001 and Reel 056157/0492 in the chain of title. See the recurrence flag under NPE signal #3.
• Context: Transfer-to-asserter, first leg. HPE conveyed the patent into a Delaware escrow vehicle under a Patent Escrow Agreement dated 2021-01-15, under which OT Patent Escrow "hold[s] in escrow for later assignment to Purchaser" (Valtrus Innovations Limited, Ireland). HPE expressly back-reserved "all rights granted under the Assigned Patents to any and all parties (including Assignor) prior to or concurrent with January 15, 2021," and took a security interest and lien in the assigned patents to secure the escrow vehicle's performance. This is a monetization divestiture with retained pre-existing licenses — not a bankruptcy sale.

---

2021-02-01 (effective) / recorded 2021-02-24 — Reel 055403/0001 (see reel discrepancy note)

• Conveyance: Assignment of assignors' interest
• Assignor: OT Patent Escrow, LLC
• Assignee: Valtrus Innovations Limited, company no. 669289 (Ireland), The Glasshouses GH2, 92 Georges Street Lower, Dun Laoghaire, Dublin A96 VR66
• Correspondent: Same cover-sheet cluster as above — Richard J. Botos, Botos Churchill IP Law LLP appears on the recorded chain-of-title sheet citing Reel 056157/0492 for this transfer.
• Reel discrepancy (flagging, not reconciling): Google Patents' legal-events record for US 9,310,855 shows the 2021-02-24 HPE→Valtrus-linked recording at 055403/0001, whereas Valtrus's Managing Director Angela Quinlan declares in a sworn PTAB exhibit that the portfolio acquisition appears "at Reel/Frame 56157/0492 and 55269/0001," and the Botos cover sheet likewise cites 055269/0001 and 056157/0492. The most likely explanation is that the same bulk portfolio assignment was recorded across more than one reel/frame tranche (a February 2021 recording and a later confirmatory/May 2021 recording — Quinlan states the portfolio was "acquired" in May 2021). Both citations are reported above rather than one being substituted for the other.
• Context: Transfer-to-asserter, second leg — escrow vehicle releases title to the designated purchaser.

---

Non-recorded instrument that materially changes the economics (flagged because it drives the NPE analysis):

2021-01-15 — Declaration of Trust (Valtrus → Key Patent Innovations Limited, "KPI"), executed the same day as the HPE purchase agreement and signed by Angela Quinlan for both companies. Under it, Valtrus holds the patents as trustee and KPI (its sole parent) is the beneficiary, with KPI holding the right of first refusal to enforce, directing any enforcement action, and controlling alienation. In litigation (SAP, Google) defendants established that Valtrus's total assets are reported as ~€1, that KPI pays Valtrus's bills directly, that Valtrus remits 100% of licensing receipts to KPI, and that Valtrus and KPI share the same Dublin office, the same five directors (Quinlan, Gibbons, Lyons, Riley, Seaman), the same phone line, and the same laptop. Several courts have held Valtrus lacks constitutional/statutory standing to sue in its own name. This trust instrument is the strongest single piece of evidence about the true beneficial owner.

Related family filings (for completeness): WO2012008945A1 (PCT, ceased), GB2503965B (active), CN102971721A (active/pending), DE112010005636T5 (withdrawn). The PCT/GB/CN/DE counterparts share the same 2010-07-12 priority.

---

Timeline diagram

timeline
    title Ownership of US 9310855
    2010 : PCT application filed by HP
         : Inventors assign rights to HP Development Co
    2013 : US national stage published
    2015 : Internal reorg to HPE Development LP
    2016 : Patent issued US 9310855
    2021 : HPE transfers to OT Patent Escrow LLC
         : Escrow transfers to Valtrus Innovations Ltd
         : Valtrus declares trust for Key Patent
    2024 : First assertion of the patent
         : Valtrus sues Digital Realty in EDTex
    2025 : TierPoint case dismissed with prejudice

---

NPE / troll-pattern signals

1. Shell-entity transfer — PRESENT (strong).
Reel 055269/0001 (2021-01-26) and the linked 2021-02-24 recording move the patent out of an operating industrial company (HPE) into a two-hop chain terminating at a single-purpose Irish entity. Concrete corroboration beyond the naming: (a) Valtrus's own audited financial statements state the company "was set up solely to acquire this patent portfolio and does not beneficially hold any assets or liabilities," with intangibles booked in the parent; (b) Valtrus's balance sheet reports ~€1 total assets; (c) the intermediate OT Patent Escrow, LLC (Delaware, 200 W Madison, 37th Floor, Chicago) is a transitory escrow conduit that held title for ~29 days. The intermediate is a classic pass-through: the assignment instrument itself recites it will "hold in escrow for later assignment to Purchaser." Unverified observation, offered cautiously: the escrow vehicle's Chicago address (200 West Madison, 37th floor) is the office location I associate with the IP merchant bank Ocean Tomo, LLC, and its signatory is Joel Lutzker — I could not confirm that connection from the assignment record, so treat it as a lead to verify, not a finding.

2. Known asserter in the chain — PRESENT.
Current owner Valtrus Innovations Limited is catalogued as a Patent Asserter / Non-Practicing Entity ("Acquired patents") in the Stanford NPE Litigation Database (e.g., patent 9310855 case list), and Unified Patents publishes a dedicated members-only litigation analysis of the Valtrus program ("Valtrus Innovations Limited, backed by Irish patent monetization platform Key Patent Innovations, has rapidly developed multiple parallel assertion programs"). IAM/Docket Navigator count the KPI-affiliated NPEs at 87 filed cases. Valtrus is not one of the legacy names on the list supplied in the brief (Acacia, Marathon, IV, Wi-LAN, Mosaid, Vringo, MPHJ, etc.) — it is a newer-generation financed assertion platform surfaced by Unified Patents, which satisfies the "any entity surfaced by Unified Patents or RPX as a high-frequency plaintiff" prong.

3. Repeat correspondent across the chain — PRESENT (with a precision caveat).
The same correspondent — Richard J. Botos, Botos Churchill IP Law LLP, 430 Mountain Avenue, Suite 401, New Providence, NJ 07974 — is the correspondent of record on the assignment cover sheet that lists both chain links (Reel 055269/0001 HPE→OT Patent Escrow, and Reel 056157/0492 OT Patent Escrow→Valtrus). So one attorney/firm spans the two 2021 links of this chain. Caveat: that cover sheet was signed May 24, 2024 — I cannot determine from the available record whether Botos Churchill was also the correspondent on the original 2021 cover sheets or was retained later for a confirmatory chain-of-title recording. I also did not find a Unified Patents / RPX / Patent Progress listing naming Botos as a correspondent on an NPE assertion register, so I am not asserting he appears there. Basis for "present" is recurrence across links within this chain, not outside-list matching.

4. Cascading transfers — PRESENT.
Two consecutive assignments completed within 29 days (recorded 2021-01-26 and 2021-02-24), preceded by a trust declaration the same day as the purchase agreement (2021-01-15), and with the intermediate and terminal entities sharing a single recording correspondent. That is the chained-LLC/cascade pattern, compressed into under a month.

5. Pre-litigation transfer — NOT PRESENT.
The ownership chain settled in Jan–Feb 2021. The earliest assertion of the '855 patent is E.D. Tex. 2:24-cv-00139 (Valtrus v. Digital Realty Trust), filed in 2024, and Valtrus's overall campaign opened with Valtrus v. Verizon on 2023-09-27 (2:23-cv-00445). The transfer therefore preceded first assertion by roughly 2.5–3 years, not within 6 months. No venue/standing-motivated last-minute assignment is visible for this patent.

6. Bankruptcy fire-sale — NOT PRESENT.
HPE was a solvent, publicly traded operating company at the time of the 2021 divestiture; the assignment was a negotiated portfolio sale (Valtrus's own financial statements describe acquiring the portfolio "for approximately [currency] 32 million" in January 2021). Nothing on the record ties this patent to a Chapter 7/11 proceeding.

7. Privateering — UNCLEAR.
On one hand this looks like a classic operating-company→NPE divestiture that now generates assertions against data-center/colocation operators, cloud providers, and their suppliers (Digital Realty, Evoque/Dawn, TierPoint, SAP, Google, Cloudflare, FedEx, Starbucks, Home Depot, Lumen). On the other, the record shows a sale for consideration with retained back-rights ("subject to all rights granted... including Assignor"), not an agency where HPE directs or shares in Valtrus's recoveries. I found no SEC filing or Patent Progress/EFF coverage showing HPE directing the campaign or sharing proceeds, so I cannot mark this present. Unclear — no dispositive evidence either way.

8. Defensive aggregator (anti-NPE) — NOT PRESENT.
The chain terminates at Valtrus Innovations Limited / Key Patent Innovations Limited (asserting entities) — not RPX, AST, LOT, Unified Patents, or OIN. The patent is live and actively asserted, not neutralized.

---

Verdict

NPE — high confidence.

Four signals are present, at least three of them strong. The chain runs HPE (operating) → OT Patent Escrow, LLC (Reel 055269/0001, recorded 2021-01-26 — a transitory Delaware escrow conduit) → Valtrus Innovations Limited (recorded 2021-02-24; also cited as Reel 056157/0492), with a same-day Declaration of Trust naming Key Patent Innovations as beneficial owner — an instrument that has caused multiple federal courts to find Valtrus lacks standing to sue in its own name and that leaves Valtrus with ~€1 in reported assets. Valtrus is a catalogued NPE/acquired-patent asserter in the Stanford NPE database and the subject of a Unified Patents litigation analysis, and the same correspondent (Richard J. Botos, Botos Churchill IP Law LLP) spans both 2021 chain links. The transfer-to-asserter conclusion rests on the recorded reel/frame entries and the beneficial-ownership trust, not on the entity names alone; the absent signals (bankruptcy, pre-litigation timing) are affirmatively refuted by the dates, and privateering is genuinely unresolved.

Verify: USPTO Assignment Center — search "9310855" and confirm reels 029475/0187, 037079/0001, 055269/0001 (and whether 055403/0001 vs. 056157/0492 governs the 2021 OT→Valtrus recording), and pull the correspondent field on each cover sheet to test the Botos recurrence claim.

---

Note on scope: I did not fabricate any reel/frame, correspondent, or assignment. Two correspondents (for the 2012 and 2015 recordings) could not be recovered from accessible sources and are marked as such rather than guessed.

I'll research US patent 9310855 and its prior-art citations. Let me run several targeted searches.

Let me gather details on the most relevant cited references.

Prior-Art Analysis — US 9,310,855 B2 ("Flexible data center and methods for deployment")

1. Patent identification (interpreted literally)

Sources: https://patents.google.com/patent/[US9310855B2](/patent/US9310855B2)/en ; https://www.freepatentsonline.com/9310855.html ; https://www.courtlistener.com/docket/68526786/1/6/valtrus-innovations-ltd-v-ntt-data-services-llc/

Important framing. The citation list for US 9,310,855 is an International Search Report / examiner citation list, and the overwhelming majority of these documents are 35 U.S.C. § 103 (obviousness) references cited as background, not anticipatory § 102 references. Under § 102 a single reference must disclose every element of a claim as arranged. None of the 25 patent citations, standing alone, appears to disclose the full combination claimed — in particular the scaling rules of claims 1/8 (a total integer number T/R of perimeter structures, at most one perimeter structure holding fewer than R rows, B = integer (T/R)/P, at most one block holding fewer than P perimeter structures) together with the menu‑driven, climate‑validated selection of cooling units and power‑conditioner units. My § 102 characterizations below are therefore stated with that reservation; where a reference is better characterized as § 103 art, I say so.

Note on § 102 subsection (pre‑AIA application, filed 2010-07-12): most references are § 102(b) art (published/patented more than one year before filing). US 2011/0240265 A1 is different — see item 25.

---

2. Claim framework used for the mapping

• Claim 1 (method): constructing B blocks; each block = 1…P perimeter structures (each holding up to R rows of server racks) plus a connecting structure housing operations monitoring equipment; total integer T/R perimeter structures; each perimeter structure consumes up to a quantity of power; at most one perimeter structure < R rows; B = integer (T/R)/P; at most one block < P perimeter structures; selecting a number of power conditioner units per desired power quality and validating them via an energy optimization evaluation against site climate; menu‑driven selection of a cooling unit type per climate, similarly validated; connecting the cooling units to an exterior of each perimeter structure.
• Claims 2–7 (dependent): butterfly/"H" configuration of P=4 with structural independence from the connecting structure (2); cool‑air pressurization / hot‑aisle return ducting (3); backup generation sized to power consumption (4); later addition of a perimeter structure on demand (5); vertical stacking of blocks (6); adjacent blocks not structurally interconnected (7).
• Claim 8 (apparatus): the corresponding data‑center structure (B blocks; 1…P perimeter structures; connecting structure; perimeter structures functional independently of the connecting structure; the same T/R, R, P, B rules; exterior cooling units whose type is climate‑particular; exterior power conditioner units whose type is particular to desired power quality and climate).
• Claims 9–12 (dependent): enumerated cooling‑unit and UPS module types (9); placement above/adjacent to a full block (10); B=1, P=4, 3.2 MW IT + 400 kW network power at PUE < 1.28 (11); no power distribution unit in the perimeter structures (12).

---

3. Reference-by-reference analysis (25 cited patent documents)

A. Modular/containerized data‑center architecture (most material)

A1. US 7,278,273 B1 — "Modular data center" — Whitted & Aigner, Google Inc.
Filed 2003-12-30; granted 2007-10-09. (Examiner-flagged.) Discloses a central office/interconnecting module 12 interconnecting four computing modules 20, plus power‑and‑cooling modules 14, all in intermodal shipping containers (Google Patents, US7278273B1). Of all citations this is structurally closest to the "connecting structure connected to P (=4) perimeter structures" architecture of claim 1/8 and to claim 2's P=4 requirement.

• Potentially relevant to: claims 1, 2, 8.
• § 102 assessment: Not anticipating. It does not disclose the T/R, R, P, B counting rules, the "at most one perimeter structure < R" / "at most one block < P" constraints, the exterior power‑conditioner step, or the menu‑driven, climate‑validated cooling selection. Best treated as § 103 art.

A2. WO 2007/139560 A1 / US 2008/0055846 A1 (Clidaras) / CN 101501599 A — "Modular computing environments" — Google Inc.
Priority 2006-06-01; published 2007-12-06 (US 2008-03-06; CN 2009-08-05). Discloses a connection hub having multiple docking areas supplying power, data network interface, and cooling‑fluid supply/return, and multiple shipping containers each adding an incremental amount of computing capacity via a docking component (Google Patents; CN 101501599 A Chinese text). This is the clearest teaching of a "connecting structure" to which multiple "perimeter structures" dock.

• Potentially relevant to: claims 1, 2, 8.
• § 102 assessment: Not anticipating the independent claims (no T/R‑P‑B scaling rules, no power‑conditioner selection/validation, no menu‑driven cooling selection), but it is the most important § 103 reference for the connecting‑structure + perimeter‑structure concept.

A3. US 2009/0229194 A1 — "Portable modular data center" — Advanced Shielding Technologies Europe S.I.
Priority 2008-03-11; published 2009-09-17. Discloses modular containers with expandable walls, in‑row cooling units, and — notably — FIG. 23, a vertical configuration of multiple portable modular data centers, plus FIG. 22, an interconnected "park or complex" of 20‑/40‑ft containers (US 2009/0229194 A1 PDF; rpxcorp record).

• Potentially relevant to: claims 1, 6, 8, 10.
• § 102 assessment: Not anticipating (no numeric modularity rules; no validated climate‑based equipment selection). § 103 art for the vertical‑stacking and multi‑container concepts.

A4. US 2011/0240265 A1 — "Container based data center solutions" — American Power Conversion Corp. (Schneider)
Priority 2010-04-06; published 2011-10-06 (family EP 2556201 A2). Discloses containerized data centers; a building of equipment racks served by a utility shipping container; an air‑handling shipping container vertically mounted to an exterior wall drawing hot exhaust and venting cool air; cooling containers with chiller/free‑cool modules; and a scalability figure (FIG. 5) (Google Patents, US20110240265; patents-review summary).

• Potentially relevant to: claims 1, 2, 3, 5, 8.
• § 102 assessment: This is the one citation that is § 102(e)‑type art (priority earlier than 2010-07-12 but published after — pre‑AIA § 102(e)). It is not a full anticipation of claims 1/8, but it is the strongest single § 103 reference against the "exterior cooling unit mounted to the structure" and "add modules as demand grows" concepts.

A5. US 2008/0055846 (see A2), US 2010/0136895 A1 — "Data center and methods for cooling thereof" — Turner Logistics, priority 2008-08-19, published 2010-06-03 — modular data‑center building with integrated cooling.

• Relevant to: claims 1, 3, 8. § 102: no; § 103 backdrop art for a modular data‑center building.

B. Cooling units, airflow and cooling‑mode selection

B1. US 2007/0076373 A1 — "Data center cooling system" — James Fink. Priority 2003-03-19; pub. 2007-04-05. CRAC/air‑distribution cooling. Relevant to claims 3, 8. § 102: no.
B2. US 2004/0227435 A1 — "Rack enclosure" — Neil Rasmussen (APC). Filed 2003-05-13; pub. 2004-11-18. Rack enclosure with an exhaust‑air duct. Relevant to claims 3, 8. § 102: no.
B3. US 2004/0240514 A1 — "Air re‑circulation index" — Bash et al. Filed 2003-05-29; pub. 2004-12-02. A metric quantifying hot‑air recirculation — supports the energy‑optimization/efficiency theme of claim 1. Relevant to claims 1, 8. § 102: no.
B4. US 2005/0235671 A1 — "Upgradeable, modular data center cooling apparatus" — Belady, HP. Filed 2004-04-22; pub. 2005-10-27. (Examiner‑flagged.) Modular, upgradeable cooling. Relevant to claims 1, 8.
B5. US 2006/0065000 A1 — "Redundant upgradeable, modular data center cooling apparatus" — Belady, HP. Filed 2004-04-22; pub. 2006-03-30. (Examiner‑flagged.) Adds redundancy to the upgradeable cooling apparatus — supports the N/N+1/2N aspects referenced in the spec and claim 4's backup‑generation concept. Relevant to claims 1, 4, 8. § 102: no.
B6. US 2008/0029250 A1 — "Warm Water Cooling" — Andrew Carlson. Priority 2006-06-01; pub. 2008-02-07. (Examiner‑flagged.) Liquid cooling. Relevant to claims 3, 8. § 102: no.
B7. US 2008/0158818 A1 — "Motherboards with Integrated Cooling" — Google Inc. Priority 2006-12-29; pub. 2008-07-03. Server‑level integrated cooling. Relevant to claim 3. § 102: no.
B8. US 7,525,207 B2 — "Water‑based data center" — Google Inc. Priority 2007-02-26; granted 2009-04-28. Data center sited on water (sea‑water cooling). Relevant to claims 1 (site selection), 8. § 102: no.
B9. US 2008/0232064 A1 — "Cooling system for information device" — Fujitsu Ltd. Priority 2007-03-22; pub. 2008-09-25. (Examiner‑flagged.) Electronic‑equipment cooling. Relevant to claim 3. § 102: no.
B10. US 8,469,782 B1 — "Data center air handling unit" — Switch Communications Group, LLC. Priority 2007-06-14; granted 2013-06-25. (Examiner‑flagged.) Large external/rooftop air‑handling units serving a data hall — relevant to claim 8's "cooling units connected to an exterior." § 102: no; this family is a significant § 103 reference for exterior AHU placement.
B11. US 7,864,530 B1 — "Changing data center cooling modes" — Exaflop LLC (Google). Priority 2007-09-28; granted 2011-01-04. (Examiner‑flagged.) Switching cooling modes (e.g., economizer vs. mechanical) based on conditions — this is the closest art to claim 1's menu‑driven selection of a cooling‑unit type and to the "energy optimization evaluation according to climate." Relevant to claim 1. § 102: no (no perimeter/connecting structure or power‑conditioner elements).
B12. US 7,630,795 B2 — "Method and air‑cooling unit with dynamic airflow and heat removal adjustability" — IBM. Priority 2008-02-15; granted 2009-12-08. (Examiner‑flagged.) Dynamic airflow control. Relevant to claim 3. § 102: no.
B13. US 2011/0100045 A1 — "Data Center Cooling" — Exaflop LLC (Google). Priority 2009-11-02; pub. 2011-05-05. (Examiner‑flagged.) Modular hot‑aisle‑capture cooling units / control of plenum pressure. Relevant to claims 3, 8. § 102: no.

C. Monitoring, capacity planning, and operations

C1. US 2007/0187343 A1 and WO 2007/095144 A2 — "Storage rack management system and method" — American Power Conversion Corp. Priority 2006-02-10; pub. 2007-08-16 / 2007-08-23. (US publication examiner‑flagged.) Rack‑level environmental/power monitoring and management. This is the citation most on point for claim 1/8's "operations monitoring equipment." Relevant to claims 1, 8. § 102: no (device is a rack management system, not the claimed block architecture).
C2. US 2009/0113323 A1 — "Data center operation optimization" — [Microsoft Corp.](/litigations/by-plaintiff/Microsoft%20Corp.) Priority 2007-10-31; pub. 2009-04-30. (Examiner‑flagged.) Optimizing data‑center operation (power/thermal efficiency). Relevant to claim 1's energy optimization evaluation. § 102: no.
C3. WO 2009/154623 A1 — "Capacity planning" — Hewlett‑Packard Development Co., L.P. Priority 2008-06-19; pub. 2009-12-23. HP's own capacity‑planning disclosure — directly relevant to the scalability/planning philosophy (claims 1, 5, 8), and notable as applicant's own prior art. § 102: no.

D. Stacking and physical deployment

D1. US 2010/0170277 A1 — "System and Method For Vertically Stacked Information Handling System and Infrastructure Enclosures" — Dell Products L.P. Priority 2008-10-31; pub. 2010-07-08. (Examiner‑flagged.) Vertically stacked IT/infrastructure enclosures — the closest art to claim 6 (stacking a second block above the first) and claim 10 (placement above/adjacent a full block). Relevant to claims 6, 10. § 102: no as to whole claims.

E. Remaining cooling/containment citations (background)

E1. CN 101501599 A — see A2 (Google, Modular computing environments).
E2. US 2008/0055846 A1 — see A2 (Clidaras, Modular Computing Environments).

---

4. Bottom line

1. No cited patent is a clean § 102 anticipation of any independent claim (1 or 8). Anticipation would require a single reference to recite, as arranged, (i) the block/perimeter/connecting‑structure architecture, (ii) the integer scaling rules (T/R, R, P, B, and the "at most one" constraints), (iii) exterior power‑conditioner selection validated by climate‑based energy optimization, and (iv) menu‑driven, climate‑validated cooling‑unit selection. None does. The references function as § 103 obviousness art, individually or in combination.
2. Most relevant / most material art (ranked):
   1. US 7,278,273 B1 (Google) — central interconnecting module + four modules in containers → claims 1, 2, 8.
   2. WO 2007/139560 A1 / US 2008/0055846 A1 / CN 101501599 A (Google, Modular computing environments) — docking connection hub + capacity‑adding containers → claims 1, 2, 8.
   3. US 2011/0240265 A1 (APC, Container based data center solutions) — exterior‑mounted cooling container, utility container, scalability → claims 1, 2, 3, 5, 8; the only § 102(e)‑type reference.
   4. US 7,864,530 B1 (Exaflop, Changing data center cooling modes) — climate/condition‑based cooling‑mode selection → claim 1.
   5. US 2007/0187343 A1 (APC) — rack/environmental monitoring → claim 1/8 "operations monitoring equipment."
   6. US 2010/0170277 A1 (Dell) — vertical stacking → claims 6, 10.
3. Where the novelty likely resides (and why the citations fall short): the claimed arithmetic modularity (T, R, P, B with the "at most one under‑filled unit" constraints), the structural independence of perimeter structures from the connecting structure (claims 2, 8), and the validated, menu‑driven, climate‑specific selection of cooling and power‑conditioning equipment (claim 1). These elements are not disclosed by the cited references, which is presumably why the claims issued over them.

Caveats / limits of this analysis. (a) I was able to verify full text/claims directly for only some references (e.g., US 2011/0240265; US 2009/0229194; US 7,278,273; WO 2007/139560/CN 101501599); for the remaining citations my characterization rests on the official titles, abstracts, and bibliographic data retrieved from Google Patents/FreePatentsOnline and should be confirmed against each reference's full specification before any § 102 assertion. (b) The patent is under active assertion by Valtrus Innovations in multiple E.D. Tex. actions, so validity contentions (including any § 102/§ 103 grounds) are being litigated; that is a legal question I cannot resolve. (c) Patent numbers and identifiers above are reproduced literally; none was auto‑corrected.

I'll ground this analysis in the prior-art references listed on the patent's own face and the Google Patents citation section, verifying key references before building combinations.

Let me verify the remaining key references, especially those touching the cooling-selection/climate limitations and modular stacking.

Obviousness Analysis — US 9,310,855 B2 under 35 U.S.C. § 103

Patent: US 9,310,855 B2, "Flexible data center and methods for deployment"
Appl. No.: 13/703,911 (national stage of PCT/US2010/041734)
Filing/priority date: 2010-07-12 → pre-AIA § 103 governs (AIA first-inventor-to-file applies only to applications filed on/after 2013-03-16). KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), is therefore controlling precedent.
Independent claims analyzed: 1 (method) and 8 (apparatus); dependents 2–7 and 9–12.

This analysis builds on the earlier summary (claim scope, file history note that the power-conditioner validation and menu-driven cooling selection limitations were added during prosecution, and the litigation/standing posture) and does not repeat it. No contradiction was found between the earlier sections and the record reproduced on the Google Patents page.

---

1. Legal framework applied

A § 103 rejection requires: (a) scope and content of the prior art; (b) differences between the prior art and the claims; (c) level of ordinary skill; and (d) where appropriate, objective indicia. Graham v. John Deere Co., 383 U.S. 1 (1966).

Because the claimed subject matter is a system/architecture built from known modular data-center elements plus a capacity-planning rule and a climate-based equipment-selection step, the following KSR lines are directly on point:

• Combination of familiar elements according to known methods, yielding no more than predictable results, is obvious. KSR, 550 U.S. at 416.
• "Obvious to try" — where there is a finite number of identified, predictable solutions, and the design incentive is known, a claim to one of them is obvious. Id. at 421.
• Design choice / optimization of a recognized parameter is obvious absent criticality. In re Boesch; In re Aller; In re Applied Materials, 692 F.3d 1289 (Fed. Cir. 2012) (optimizing a result-effective variable).
• Market/competitive pressure is a legitimate motivation. KSR at 418–19.

Level of ordinary skill in the art (proposed): a bachelor's degree in mechanical or electrical engineering (or equivalent) plus 2–4 years of experience in data-center design, construction, or facilities engineering, including containerized/modular data centers, HVAC load selection, and power distribution. The problems addressed (PUE, scalability, capital cost, climate-driven cooling) are all classic facilities-engineering problems; the '855 disclosure is essentially a design playbook rather than a technical advance.

---

2. The prior art of record (from the patent's own face / citation list)

All references below appear in the "Patent Citations (25)" and "Non-Patent Citations (13)" sections of the page, i.e., they were before the examiner. Effective dates all precede the 2010-07-12 priority date except where flagged.

Priority-date caveat on US 2011/0240265 (APC): it published 2011-10-06, after the '855 priority date. It appears in the '855 citation list, so it evidently qualified as prior art on the record (most plausibly via pre-AIA § 102(e) on its 2010-04-06 priority/PCT filing). If APC '265's 102(e) date were successfully challenged, the cooling-module-type and exterior-mounting limitations would need to be carried by '273, Dell '277 and the NPL instead — the grounds below are constructed so that APC '265 is helpful but not indispensable.

---

3. Grounds of rejection

Ground 1 (primary) — § 103 over US 7,278,273 + US 2010/0136895 + US 2010/0170277, optionally further in view of US 2011/0240265

Why they combine: All three are in the same field (designing/constructing data centers), address the same problem the '855 specification identifies (scalability, capital efficiency, and bridging containerized and brick-and-mortar approaches), and are expressly mutually aware — Dell '277 discusses and critiques Google's '273 architecture by name. KSR at 417 ("the background knowledge of a person of ordinary skill" and the "interrelated teachings of multiple patents").

Claim 1 chart:

Ground 2 — § 103 over Belady '671 / Belady '000 in view of '273 and Turner '695

Belady frames the data center explicitly as alternating rows of racks and cooling units that can be upgraded modularly; '273 supplies the block/interconnect architecture; Turner supplies "each cell having a predefined power capacity" and enthalpy-based climate logic. This ground is attractive because Belady is HP's own prior art (same assignee lineage), which weakens any argument that the '855 inventors were in an unforeseeably different design space. Belady also teaches the airflow pattern of claim 3 (false ceiling separating warm return from the racks, warm air ducted above the racks) and, in the '000 continuation, the redundancy aspects invoked by the '855 spec's N/N+1/2N discussion.

Ground 3 — § 103 over Dell '277 in view of '273 and Turner '695 (targeting claims 6, 7, 8)

Dell '277 directly teaches claim 6's vertical stacking ("the shipping containers are arranged in a stacked configuration so that cooled air and exhausted air are exchanged through aligned vents formed in the ceiling and floor of stacked shipping containers") and claim 8's independence requirement (separate containers, each operable, with infrastructure in a separate enclosure). The motivation is expressly recited in Dell's own background: greater computing density with more efficient infrastructure resource allocation, and avoiding the need for a module to include cooling for "a variety of different climates."

For claim 7 (second block adjacent without connecting structures tying the blocks together), '273's self-contained modules plus APC '265's containerized layout supply the teaching; the motivation is fault isolation — the same rationale the '855 specification itself offers ("a complete loss of the connecting structure … the connecting structure is not a single point of failure"). A claimed benefit that the specification ascribes to the arrangement is not itself a basis for non-obviousness where the arrangement is known.

Ground 4 — § 103 over the climate/energy NPL (Cho et al.; Di Yuhui et al.; Kumar) in view of Turner '695 and APC '265 (targeting the prosecution-added limitations)

The limitations that were added during prosecution — (i) validating the power-conditioner count by an energy-optimization evaluation keyed to site climate, and (ii) the menu-driven climate-based cooling-type selection with energy-optimization validation — are the only limitations with any non-trivial content, and they are squarely met by the NPL the examiner cited:

• Di Yuhui et al. demarcates climate zones for applicability of evaporative cooling — the climatic-selection step.
• Cho et al. performs a regional energy-efficiency viability analysis of alternative data-center cooling systems — the energy-optimization validation.
• Kumar evaluates air-side economizers for data-center thermal management — the alternative cooling modality.
• Turner '695 supplies the control analogy with its enthalpy comparison, and APC '265 supplies the selection set ("chiller module or a free cool unit").

Combining a known climate-map (Di Yuhui) with a known energy-viability study (Cho) and a known modular cooling catalog (APC '265 / the '855 specification's own enumerated list) to produce a "menu-driven validated selection" is the paradigm of KSR's "combination of familiar elements according to known methods" yielding a predictable result.

---

4. Claim-by-claim vulnerability summary

---

5. Motivation to combine — consolidated

1. Common field and common problem. Every reference is a data-center design/construction/cooling reference; each addresses scalability, capital cost, modularity, or energy efficiency — the exact problems the '855 background recites.
2. Express cross-awareness. Dell '277 criticizes and distinguishes Google '273 by name, showing a person of ordinary skill was already combining/contrasting these modular approaches.
3. Finite, predictable solution space. The number of ways to array N racks into K modular enclosures with climate-appropriate cooling is small; KSR's "obvious to try" applies.
4. Business/economic drivers. The '855 specification itself frames the invention as a cost/time-to-market solution ("less than $7 million per megawatt"; "24 to 36 weeks"). Cost- and schedule-driven design changes are the classic "market forces" motivation of KSR.
5. Admissions in the specification. The '855 spec (a) treats the block/perimeter arithmetic as a utilization-efficiency rule, (b) states the menu-driven process can be implemented by conventional machine-readable instructions on a tangible medium, and (c) recites N/N+1/2N power topologies and conventional cooling-unit types as known options. These are party admissions usable against the claims.

---

6. Where a non-obviousness argument could still gain traction

Being candid about the limits of this analysis:

• The specific conjunction of arithmetic + menu + dual validation. No single reference of record clearly discloses all of (i) the fill-before-expand rule, (ii) a menu-driven cooling-type selection, and (iii) an energy-optimization validation of both the power-conditioner count and the cooling type, all keyed to the site climate. A patentee would argue these were added to overcome art and reflect a genuine, non-routine design methodology. The counter is Ground 4 (NPL) plus KSR "combination of familiar elements" and the specification's own admission that the menu process is a generic configurator.
• "Menu driven process" as an interface. A patentee may argue the phrase imports a specific tool/UX, not merely a design method. The examiner's own cited NPL (Di Yuhui, Cho, Kumar, IBM Rx) suggests otherwise and appears to be the reason those references were cited.
• Priority-date risk on APC '265. As flagged, if its pre-AIA § 102(e) date is challengeable, the cooling-module-type and exterior-mounting limitations must rest on '273 + Dell '277 + the NPL. That weakens — but does not defeat — Grounds 1 and 4.
• Claim 11's numerical parameters. If the patentee can show the 1.28 PUE threshold is critical (i.e., an unexpected or non-routine result), claim 11 gets stronger. Nothing in the specification as reproduced demonstrates criticality, so I assess this as unlikely.

---

7. Verification limits and flags

• WO 2009/154623 ("Capacity planning," HP) is cited on the face of '855, but I could not retrieve its specification text before my search budget was exhausted. I use it only as corroboration that capacity planning for data centers was a recognized discipline, not for any specific teaching. Any rejection relying on it should re-verify its disclosure.
• US 2011/0240265 (APC) carries a post-'855-publication date; its prior-art status rests on the examiner's citation and a likely § 102(e) date. Verify before relying on it as the sole source for a limitation.
• I did not retrieve full texts for every secondary reference (e.g., US 2009/0229194, US 7,864,530, US 2011/0100045, US 2009/0113323, US 7,630,795, US 8,469,782); the characterizations above are drawn from the title/assignee entries on the '855 citation list plus the snippets retrieved, and should be confirmed claim-chart-by-claim-chart before filing.
• No contradiction with the earlier-generated sections; this analysis is consistent with (and extends) the note that the power-conditioner and menu-driven-cooling limitations were prosecution additions.
• Source for the authoritative patent text and citation list: US9310855B2 on Google Patents.

Bottom line: Independent claims 1 and 8, and dependent claims 2–7 and 9–12, are vulnerable to § 103 rejection on the record's own art — principally US 7,278,273 (Google) in view of US 2010/0136895 (Turner Logistics), US 2010/0170277 (Dell), and US 2011/0240265 (APC), with the climate/energy-validation limitations met by the cited non-patent literature (Cho et al.; Di Yuhui et al.; Kumar). The claims most likely to survive a well-pleaded KSR ground are only those keyed to a critical numerical outcome (claim 11) — and even that is doubtful absent evidence of criticality.