# US 2013/0252379 A1 · Method for making high-speed ceramic modules with hybrid referencing scheme for improved performance and reduced cost

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2013/0252379 A1
- Document ID: US20130252379A1
- Status: Published application
- Kind: A1
- Application number: 13/621,242
- Filed: 2012-09-15
- Published: 2013-09-26
- Earliest claimed: 2009-12-04
- Assignees: IBM Corporation
- Applicants: IBM Corporation
- Inventors: Wiren D Becker; Jinwoo Choi; Tingdong Zhou
- Web page: https://patentgrep.com/patent/US20130252379A1

## CPC

- H10W 20/031: {of conductive parts of the interconnections}
- H05K 1/0253: {Impedance adaptations of transmission lines by special lay-out of power planes, e.g. providing openings (H05K1/0251 takes precedence)}
- H10W 42/267: {Patterned shielding planes}
- H10W 44/20: at high-frequency [HF] or radio frequency [RF]
- H10W 70/635: {Through-vias}
- H10W 70/65: Shapes or dispositions of interconnections
- H10W 70/685: comprising multiple insulating layers
- H05K 3/4629: {laminating inorganic sheets comprising printed circuits, e.g. green ceramic sheets}
- H05K 2201/09309: Core having two or more power planes; Capacitive laminate of two power planes
- H05K 2201/09681: Mesh conductors, e.g. as a ground plane
- H05K 2201/09727: Varying width along a single conductor; Conductors or pads having different widths
- H05K 2201/10674: Flip chip
- Y10T 29/49155: Manufacturing circuit on or in base
- H10W: GENERIC PACKAGES, INTERCONNECTIONS, CONNECTORS OR OTHER CONSTRUCTIONAL DETAILS OF DEVICES COVERED BY CLASS H10
- H05K: PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- Y10T: TECHNICAL SUBJECTS COVERED BY FORMER US CLASSIFICATION

## Abstract

A multi-layered ceramic package comprises: a signal layer with identified chip/device area(s) that require a supply of power; and a voltage power (Vdd) layer and a ground (Gnd) layer disposed on opposite sides directly above or below and adjacent to the signal layer and providing a first reference mesh plane and a second reference mesh plane configured utilizing a hybrid mesh scheme. The hybrid mesh scheme comprises: a full dense mesh in a first area directly above or below the identified chip/device area(s); a half dense mesh in a second area that is above or below the edge(s) of the chip/device area; and a wider mesh pitch in all other areas The Vdd traces are aligned to run parallel and adjacent to signal lines in those other areas. Wider traces are provided within the mesh areas that run parallel and adjacent to signal lines.

## Claims

### Claim 1

In a fabrication system for a multi-layer ceramic module having at least one signal layer with one or more devices/chips that require power, a method for providing adequate power to the one or more devices/chips while reducing the amount of metal utilized within the module and reducing far end (FE) noise within the module, the method comprising:
identifying one or more high power usage (HPU) areas within the signal layer at which the one or more devices/chips that require power are to be situated;
providing a full dense mesh plane of traces within first areas of a voltage power (Vdd) layer and a ground (Gnd) layer that are adjacent to the HPU areas within the signal layer to support power deliver to the devices/chips on the signal layer; and
providing a low density mesh plane of traces in other areas of the Vdd layer and Gnd layer that are not adjacent to or proximate to the HPU areas within the signal layer.

### Claim 2 (depends on 1)

The method of claim 1, wherein the providing a low density mesh plane further comprises providing only Vdd traces and Gnd traces that are parallel to adjacent signals lines running on the signal layer.

### Claim 3 (depends on 1)

The method of claim 1, wherein:
identifying the one or more HPU areas comprises differentiating in a Vdd reference plane and in a Gnd reference plane (a) a first HPU area that is directly adjacent to the one or more HPU areas of the signal layer from (b) a second area that is proximate to a perimeter edge of the one or more HPU areas of the signal layer and (c) a third area on the signal layer that is away from the first area and the second area and not proximate to the one or more HPU areas; and
said method further comprises providing a half dense mesh of traces within second areas of the Vdd layer and Gnd layer that are adjacent to and proximate to the perimeter edge of the HPU areas.

### Claim 4 (depends on 1)

The method of claim 1, further comprising:
providing a larger width of the Vdd and Gnd traces running parallel to and adjacent to signal lines in the other areas that are not proximate to the one or more HPU areas, wherein the larger width provides wider Vdd and Gnd traces, which reduce the IR drop on the respective traces.

### Claim 5 (depends on 4)

The method of claim 4, further comprising providing a matched impedance at a driver side of one or more signal lines to reduce near end (NE) crosstalk.

### Claim 6 (depends on 1)

The method of claim 1, further comprising:
reducing an amount of metal loading at areas within the Gnd and Vdd layers that are not adjacent to or proximate to the HPU areas by increasing a mesh pitch between traces on the Vdd layer and Gnd layer in areas that are not directly adjacent to or proximate to the one or more HPU areas, wherein the increasing of the mesh pitch provides a substantial reduction in the far end crosstalk and better bus performance.

### Claim 7 (depends on 1)

The method of claim 1, further comprising:
optimizing a width of the signal lines and a dielectric material thickness to obtain a desired impedance of the signal lines; and
increasing the width of the Vdd and Gnd reference traces that are above, below, and adjacent to the signals lines to reduce noise coupling.

### Claim 8 (depends on 1)

The method of claim 1, further comprising:
disposing a plurality of vias in spaces provided by the Vdd mesh plane and Gnd mesh plane with each via of the plurality of vias providing one of a Vdd connection or a Gnd connection;
coupling to surrounding Vdd vias and Gnd vias the signal lines that are not proximate to the HPU areas to close the return path; and
running parallel Vdd traces and Gnd traces that are adjacent to the signal lines, wherein providing the parallel, adjacent Vdd and Gnd traces reduces discontinuity on a return path and results in less far end (FE) crosstalk.

### Claim 9

A computer program product for use in fabrication and design of high speed ceramic packages comprising:
a computer readable medium; and
program code on said computer readable medium that when executed within a data processing device, said program code provides the functionality of:
identifying one or more high power usage (HPU) areas within the signal layer at which the one or more devices/chips that require power are to be situated;
providing a full dense mesh plane of traces within first areas of a voltage power (Vdd) layer and a ground (Gnd) layer that are adjacent to the HPU areas within the signal layer to support power deliver to the devices/chips on the signal layer; and
providing a low density mesh plane of traces in other areas of the Vdd layer and Gnd layer that are not adjacent to or proximate to the HPU areas within the signal layer.

### Claim 10 (depends on 9)

The computer program product of claim 9, wherein:
said program code for identifying the one or more areas further comprises code for differentiating in a Vdd reference plane and in a Gnd reference plane (a) a first HPU area that is directly adjacent to the one or more HPU areas of the signal layer from (b) a second area that is proximate to a perimeter edge of the one or more HPU areas of the signal layer and (c) a third area on the signal layer that is away from the first area and the second area and not proximate to the one or more HPU areas; and
said program code further comprises code completing the functionality of providing a half dense mesh of traces within second areas of the Vdd layer and Gnd layer that are adjacent to and proximate to the perimeter edge of the HPU areas.

### Claim 11 (depends on 9)

The computer program product of claim 9, said program code further comprising code for completing the functionality of:
reducing an amount of metal loading at areas within the Gnd and Vdd layers that are not adjacent to or proximate to the HPU areas by increasing a mesh pitch between traces on the Vdd layer and Gnd layer in areas that are not directly adjacent to or proximate to the one or more HPU areas, wherein the increasing of the mesh pitch provides a substantial reduction in the far end crosstalk and better bus performance.

### Claim 12 (depends on 9)

The computer program product of claim 9, said program code further comprising code for:
optimizing a width of the signal lines and a dielectric material thickness to obtain a desired impedance of the signal lines;
reducing an amount of metal loading at areas within the Gnd and Vdd layers that are not above or below the device/chip in the signal layer, by reducing a width of the Vdd and Gnd traces to a second width; and
increasing the width of the Vdd and Gnd reference traces that are above, below, and adjacent to the signals lines to reduce noise coupling.

### Claim 13 (depends on 9)

The computer program product of claim 9, said program code further comprising code for:
optimizing a width of the signal lines and a dielectric material thickness to obtain a desired impedance of the signal lines; and
increasing the width of the Vdd and Gnd reference traces that are above, below, and adjacent to the signals lines to reduce noise coupling.

### Claim 14 (depends on 9)

The computer program product of claim 9, wherein the program code for providing a low density mesh plane further comprises:
providing only Vdd traces and Gnd traces that are parallel to adjacent signals lines running on the signal layer; and
providing a larger width of the Vdd and Gnd traces running parallel to and adjacent to signal lines in the other areas that are not proximate to the one or more HPU areas, wherein the larger width provides wider Vdd and Gnd traces, which reduce the IR drop on the respective traces.

### Claim 15 (depends on 14)

The computer program product of claim 14, said program code further comprising code for providing a matched impedance at a driver side of one or more signal lines to reduce near end (NE) crosstalk.
