IP Library › Granted Patent US 12,737,118
Granted Patent B2
US 12,737,118 · App. 19/172,441 · Granted Sep 15, 2026

Photonic memory fabric for system memory interconnection

Inventors: David Lazovsky (Los Gatos, CA); Philip Winterbottom (San Jose, CA); Martinus Bos (San Jose, CA)
Assignee: Sicily Merger Sub II, Inc.
G06F3/0611G02B6/12019G02B6/1225G02B6/4246G06F3/0655G06F3/0679H04B10/695H04B10/801H04Q11/0062
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,737,118
App. No.
19/172,441
Filed
Apr 7, 2025
Granted
Sep 15, 2026
Kind
B2
Art Unit
2634
USPC
398/49
Abstract

One embodiment is a method that includes generating a request for a data item in a memory, obtaining the data item from the memory with a photonic interface, sending the data item to a fabric using a transmit unit of the photonic interface, and routing the data item through a portion of the fabric coupled to the memory, the portion of the fabric including one or more additional transmit and receive units between the photonic interface and a destination receive unit.

Claims (36)

1 . A system-in-package, comprising:

a computing resource; and

an interface associated with the computing resource, the interface comprising:

a first transmit unit for sending data from the computing resource via a first optical link, wherein the first transmit unit comprises a driver connected to a modulator, and a serializer, the driver being disposed in a photonic-integrated circuit (PIC) and connected to the serializer, wherein the serializer provides an output to the driver;

a first receive unit for receiving data sent to the computing resource via a second optical link, wherein the first receive unit comprises a photodetector a deserializer;

at least one additional transmit unit for sending data from the computing resource via a first one or more additional optical links of a bidirectional channel and at least one additional receive unit for receiving data sent to the computing resource via a second one or more additional optical links of the bidirectional channel; and

a transmit bonding engine for splitting the data sent from the computing resource across the first optical link and the first one or more additional optical links and a receive bonding engine for assembling received data sent to the computing resource across the second optical link and the second one or more additional optical links.

2 . The system-in-package of claim 1 , wherein the first optical link comprises a first waveguide from the modulator to a second receive unit associated with a second computing resource.

3 . The system-in-package of claim 2 wherein the second optical link comprises a second waveguide from a second transmit unit associated with the second computing resource to the photodetector.

4 . The system-in-package of claim 3 , further comprising a third transmit unit for sending data from the computing resource via a third optical link wherein the first transmit unit comprises a second driver connected to a second modulator, and a second serializer,

wherein the third optical link comprises an optical pathway that uses an optical interface that guides light from a waveguide into a fiber.

5 . The system-in-package of claim 1 , wherein the computing resource is one or more of a memory element or a processing element.

6 . The system-in-package of claim 1 , wherein the serializer converts an electronic message received in a form of parallel data into a signal suitable for driving the modulator, and wherein the deserializer converts a received message back into parallel data.

7 . The system-in-package of claim 1 , wherein the modulator is selected from the group consisting of an electro-absorption modulator (EAM), a micro-ring resonator, a ring modulator, a Mach-Zender interferometer (MZI), and a quantum confined stark effect (QCSE) electro-absorptive modulator.

8 . The system-in-package of claim 1 , wherein the computing resource is one or more of a NAND Flash memory, a solid-state drive (SSD) memory, a NOR Flash memory, a CMOS memory, a thin film transistor-based memory, a phase change memory (PCM), a storage class memory (SCM), a magneto-resistive memory (MRAM), a resistive RAM, a DRAM, an HBM, a DDR-based DRAM, or a DIMM memory.

9 . The system-in-package of claim 1 , wherein the computing resource includes a combination of at least one memory element that provides access to memory traffic and a processing element that provides access to compute traffic.

10 . The system-in-package of claim 1 , wherein the first receive unit further comprises:

a gain control to normalize a signal level of a signal containing the data sent to the computing resource via the second optical link; and

a slicer to extract a bit-stream that is provided as an input to a de-serializer for converting the data sent to the computing resource into parallel data.

11 . A system-in-package, comprising:

a first computing resource;

a second computing resource; and

a fabric coupled to the first computing resource and the second computing resource, the fabric comprising:

a first interface associated with the first computing resource, the first interface comprising:

a first transmit unit for sending data from the first computing resource via a first optical link of a bidirectional channel, the first transmit unit comprising a first driver connected to a first modulator, the first driver and the first modulator being implemented in a photonic integrated circuit (PIC), and a first serializer in a first electronic integrated circuit (EIC), the first driver being connected to the first serializer, wherein the first serializer provides a first output to the first driver; and

a first receive unit for receiving data sent to the first computing resource via a second optical link of the bidirectional channel, the first receive unit comprising a photodetector in the PIC, and a first deserializer in the first EIC;

at least one additional transmit unit for sending data from the first computing resource via a first one or more additional first optical links of the bidirectional channel;

at least one additional first receive unit for receiving data sent to the first computing resource via a second one or more additional first optical links of the bidirectional channel;

a transmit bonding engine for splitting the data sent from the first computing resource across the first optical link and the first one or more additional optical links; and

a receive bonding engine for assembling received data sent to the first computing resource across the second optical link and the second one or more first additional optical links;

a second interface associated with the second computing resource, the second interface comprising:

a second transmit unit for sending data from the second computing resource via the second optical link of the bidirectional channel, the second transmit unit comprising a second driver connected to a second modulator, the second driver and the second modulator being implemented in the PIC, and a second serializer in a second EIC, the second driver being connected to the second serializer, wherein the second serializer provides a second output to the second driver; and

a second receive unit for receiving data sent to the second computing resource via the first optical link of the bidirectional channel, the second receive unit comprising a second photodetector in the PIC, and a second deserializer in the second EIC.

12 . The system-in-package of claim 11 , wherein each computing resource of the first computing resource and the second computing resource is one or more of a memory element or a processing element.

13 . The system-in-package of claim 11 , wherein each modulator of the first modulator and the second modulator is selected from the group consisting of an electro-absorption modulator (EAM), a micro-ring resonator, a ring modulator, a Mach-Zender interferometer (MZI), and a quantum confined stark effect (QCSE) electro-absorptive modulator.

14 . The system-in-package of claim 11 , wherein each computing resource of the first computing resource and the second computing resource is one or more of a NAND Flash memory, a solid-state drive (SSD) memory, a NOR Flash memory, a CMOS memory, a thin film transistor-based memory, a phase change memory (PCM), a storage class memory (SCM), a magneto-resistive memory (MRAM), a resistive RAM, a DRAM, an HBM, a DDR-based DRAM, or a DIMM memory.

Assignments (2)
MERGER Recorded Feb 10, 2026
From: CELESTIAL AI INC.
To: SICILY MERGER SUB II, INC.
Reel/Frame 074721/0610 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 8, 2025
From: LAZOVSKY, DAVID; WINTERBOTTOM, PHILIP; BOS, MARTINUS
To: CELESTIAL AI INC.
Reel/Frame 070768/0084 →
Continuity (3)
Continuation 17903455 · Sep 6, 2022
Provisional Application 63321453 · Mar 18, 2022
Related Publication 20250258606A1 · Aug 14, 2025
References Cited (79)
US 5541914A · Krishnamoorthy et al. · 1996 [cited by applicant]
US 6908856B2 · Beyne et al. · 2005 [cited by applicant]
US 7734191B1 · Welch · 2010 [cited by examiner]
US 8307021B1 · Dhanoa et al. · 2012 [cited by applicant]
US 9354039B2 · Mower et al. · 2016 [cited by applicant]
US 9946539B1 · Temam et al. · 2018 [cited by applicant]
US 10054737B2 · Kobrinsky et al. · 2018 [cited by applicant]
US 10268232B2 · Harris et al. · 2019 [cited by applicant]
US 10365447B2 · Mekis et al. · 2019 [cited by applicant]
US 10970809B1 · Seiler · 2021 [cited by applicant]
US 11157807B2 · Abel et al. · 2021 [cited by applicant]
US 11475367B2 · Lazovich et al. · 2022 [cited by applicant]
US 12061978B2 · Zalevsky et al. · 2024 [cited by applicant]
US 12353006B2 · Winterbottom et al. · 2025 [cited by applicant]
US 20020118713A1 · Shirai et al. · 2002 [cited by applicant]
US 20080067677A1 · Lin · 2008 [cited by examiner]
US 20090026607A1 · Huebner · 2009 [cited by examiner]
US 20100148210A1 · Huang et al. · 2010 [cited by applicant]
US 20100266295A1 · Zheng et al. · 2010 [cited by applicant]
US 20110069963A1 · McLaren et al. · 2011 [cited by applicant]
US 20130037948A1 · Samoilov et al. · 2013 [cited by applicant]
US 20140008757A1 · Ramachandran et al. · 2014 [cited by applicant]
US 20140073134A1 · Farooq et al. · 2014 [cited by applicant]
US 20140103520A1 · Kirby et al. · 2014 [cited by applicant]
US 20140203175A1 · Kobrinsky et al. · 2014 [cited by applicant]
US 20150249073A1 · Herrmann · 2015 [cited by applicant]
US 20150341119A1 · Fincato · 2015 [cited by examiner]
US 20160268163A1 · Nakamura · 2016 [cited by applicant]
US 20160285581A1 · Mickelson et al. · 2016 [cited by applicant]
US 20170115458A1 · Mekis et al. · 2017 [cited by applicant]
US 20170141142A1 · Leobandung et al. · 2017 [cited by applicant]
US 20170237226A1 · Johnson et al. · 2017 [cited by applicant]
US 20190012295A1 · Yinger et al. · 2019 [cited by applicant]
US 20190014341A1 · Kondo et al. · 2019 [cited by applicant]
US 20190067260A1 · Koyama et al. · 2019 [cited by applicant]
US 20190067900A1 · Bhattacharya et al. · 2019 [cited by applicant]
US 20190089466A1 · Li · 2019 [cited by examiner]
US 20190236049A1 · Vantrease et al. · 2019 [cited by applicant]
US 20190305027A1 · Qian et al. · 2019 [cited by applicant]
US 20190333905A1 · Raghunathan et al. · 2019 [cited by applicant]
US 20190391811A1 · Garegrat et al. · 2019 [cited by applicant]
US 20200209655A1 · Roth et al. · 2020 [cited by applicant]
US 20200219865A1 · Nelson et al. · 2020 [cited by applicant]
US 20200234099A1 · Wang et al. · 2020 [cited by applicant]
US 20200280173A1 · Gao et al. · 2020 [cited by applicant]
US 20200327403A1 · Du et al. · 2020 [cited by applicant]
US 20210202562A1 · Chang et al. · 2021 [cited by applicant]
US 20210215897A1 · Epitaux · 2021 [cited by examiner]
US 20210225708A1 · Lee et al. · 2021 [cited by applicant]
US 20210256360A1 · Nam · 2021 [cited by applicant]
US 20210266200A1 · Yang et al. · 2021 [cited by applicant]
US 20210271020A1 · Islam · 2021 [cited by examiner]
US 20210319076A1 · Komuravelli et al. · 2021 [cited by applicant]
US 20210373895A1 · Shahim et al. · 2021 [cited by applicant]
US 20220012013A1 · Sebastian et al. · 2022 [cited by applicant]
US 20220171605A1 · Willcock · 2022 [cited by applicant]
US 20220171829A1 · Snelgrove · 2022 [cited by applicant]
US 20220222174A1 · Garegrat et al. · 2022 [cited by applicant]
US 20220293820A1 · Fitzgerald et al. · 2022 [cited by applicant]
US 20220309336A1 · Minkin · 2022 [cited by applicant]
US 20230006417A1 · Von Malm · 2023 [cited by applicant]
US 20230152667A1 · Miscuglio et al. · 2023 [cited by applicant]
US 20230309353A1 · Jin et al. · 2023 [cited by applicant]
US 20230344518A1 · Kalman et al. · 2023 [cited by applicant]
US 20240013041A1 · Pleros et al. · 2024 [cited by applicant]
US 20250253300A1 · Aggarwal · 2025 [cited by applicant]
CN 1484877A · 2004 [cited by applicant]
CN 106068579A · 2016 [cited by applicant]
JP 2019523932A · 2019 [cited by applicant]
JP 2022543366A · 2022 [cited by applicant]
KR 20130007063A · 2013 [cited by applicant]
KR 20170140297A · 2017 [cited by applicant]
KR 20190137835A · 2019 [cited by applicant]
Chittamuru, Sai Vineel Reddy et al., “BiGNoC: Accelerating Big Data Computing with Application-Specific Photonic Network-on-Chip Architectures”, IEEE Transactions on Parallel and Distributed Systems, IEEE, May 8, 2018, … [cited by applicant]
Foulk et al.; “Broad Temperature Operation and Widely Tunable High Dynamic Range High-Speed Amplified Electroabsorbtion Modulator”, IEEE Photonics Technology Letters, vol. 17, No. 10, Oct. 2005, pp. 2191-2193. [cited by applicant]
T. Luo et al., “DaDianNao: A Neural Network Supercomputer”, IEEE Transactions on Computers, May 30, 2016. [cited by applicant]
Hendry, et al., “Analysis of Photonic Networks for a Chip Multiprocessor Using Scientific Applications,” published at 2009 3rd ACM/IEEE International Symposium on Networks-on-Chip, May 2009, 10 pages. [cited by applicant]
Ke Wen et al., “Silicon photonic memory interconnect for many-core architectures”, 2016 IEEE High Performance Extreme Computing Conference (HPEC), Sep. 13, 2016, 7 pages. [cited by applicant]
Zhang Yu et al., “Foundry-Enabled Scalable All-to-All Optical Interconnects Using Silicon Nitride Arrayed Waveguide Router Interposers and Silicon Photonic Transceivers”, IEEE Journal of Selected Topics in Quantum Elect… [cited by applicant]