JP2006179021A - Logic synthesizer - Google Patents

Logic synthesizer Download PDF

Info

Publication number
JP2006179021A
JP2006179021A JP2006028387A JP2006028387A JP2006179021A JP 2006179021 A JP2006179021 A JP 2006179021A JP 2006028387 A JP2006028387 A JP 2006028387A JP 2006028387 A JP2006028387 A JP 2006028387A JP 2006179021 A JP2006179021 A JP 2006179021A
Authority
JP
Japan
Prior art keywords
layout
time
logic synthesis
logic
cell
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP2006028387A
Other languages
Japanese (ja)
Inventor
Hidetaka Minami
英孝 南
Takamitsu Yamada
孝光 山田
Yasutaka Tsukamoto
泰隆 塚本
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ricoh Co Ltd
Original Assignee
Ricoh Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ricoh Co Ltd filed Critical Ricoh Co Ltd
Priority to JP2006028387A priority Critical patent/JP2006179021A/en
Publication of JP2006179021A publication Critical patent/JP2006179021A/en
Pending legal-status Critical Current

Links

Images

Abstract

<P>PROBLEM TO BE SOLVED: To provide a logic synthesizer capable of preventing unnecessary load distribution and performing load distribution in consideration of the optimal cell arrangement and a wiring area at the time of layout in logic synthesis of LSI design. <P>SOLUTION: A room for improvement in layout remains at the time of logic synthesis by analyzing an RTL (s2), extracting a high fan-out network (s3) and inserting a buffer for a clock tree to be performed at the time of layout into the RTL to the network (s4) and the optimal layout in consideration of the cell arrangement, the wiring area is performed by a layout tool. In addition, since the library is analyzed (s6) and a cell with large drive capacity is not used at the time of logic synthesis (s7), the optimal layout by converging timing is performed by re-mapping a high drive cell to a place where timing is severe at the time of layout. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

本発明は、LSI(Large Scale Integration、大規模集積回路)設計の論理合成およびレイアウトに関する論理合成装置に関する。   The present invention relates to a logic synthesis device for logic synthesis and layout of LSI (Large Scale Integration) design.

従来のLSI設計手法は、論理合成の段階で、ファンアウト数の多い箇所に対しては、負荷分散のためバッファリングし、レイアウト工程へ進んでいた。   In the conventional LSI design method, at the stage of logic synthesis, a portion with a large number of fan-outs is buffered for load distribution and proceeds to a layout process.

また、論理合成の段階で、負荷が大きく処理タイミングの高速化が望まれる箇所に対しては、高ドライブセルをマッピングし、レイアウト工程へ進んでいた。   Also, at the stage of logic synthesis, high drive cells are mapped to locations where the load is large and processing speed is desired to be high, and the process proceeds to the layout process.

類似技術としては、日本電気株式会社から公開されている特開2000−231583号公報に記載されたものがある。これは、従来フロアプランを考慮した論理合成を行う際に、遅延調整のため論理セルを高ドライブセルに置き変えるという手法ではなく、仮フロアプランの情報からファンアウト数を考慮して調整用バッファを自動挿入することにより、遅延改善を行うというものである(例えば、特許文献1参照)。
特開2000−231583公報
As a similar technique, there is one described in Japanese Patent Application Laid-Open No. 2000-231583 published by NEC Corporation. This is not a technique of replacing logic cells with high drive cells for delay adjustment when performing logic synthesis in consideration of conventional floor plans, but adjusting buffer considering the number of fan-outs based on temporary floor plan information. Is automatically inserted to improve the delay (see, for example, Patent Document 1).
JP 2000-231583 A

しかしながら、上記従来の手法を用いても、レイアウト時の配置配線後に確実にタイミング収束している保証はない。また、論理合成時にファンアウト調整用のバッファを挿入するため、レイアウト時に配線効率が悪くなる可能性がある。   However, even if the conventional method is used, there is no guarantee that the timing is surely converged after the placement and routing at the time of layout. In addition, since a fan-out adjustment buffer is inserted at the time of logic synthesis, wiring efficiency may be deteriorated at the time of layout.

そこで本発明は、論理合成時に負荷分散のためのバッファを挿入するのではなく、レジスタトランスファーレベルのハードウェア記述言語:RTLを解析し、ハイファンアウトネットを抽出し、そのネットに対し、レイアウト時に行うクロックツリーのためのバッファをRTLへ挿入する。   Therefore, the present invention does not insert a buffer for load distribution at the time of logic synthesis, but analyzes a register transfer level hardware description language: RTL, extracts a high fan-out net, and outputs that net at the time of layout. Insert a buffer for the clock tree to be performed into the RTL.

また、論理合成に使用するライブラリを解析し、レイアウト後にタイミングバイオレーションが残ることを想定し、その際にタイミング調整ができる余地を残すためにドライブ能力の大きなセルは論理合成時には使用せず、レイアウトでの改善の余地を論理合成時に残す。   Analyzing the library used for logic synthesis and assuming that timing violations remain after layout, cells with large drive capacity are not used during logic synthesis to leave room for timing adjustment. Leave room for improvement in logic synthesis.

これにより、レイアウトツールでセル配置、配線領域を考慮した最適な負荷分散を行い、論理合成時の不要な負荷分散を防ぎ、配線が混雑してレイアウトできないという問題を解決する。   As a result, optimal load distribution in consideration of the cell arrangement and the wiring area is performed by the layout tool, unnecessary load distribution at the time of logic synthesis is prevented, and the problem that the wiring is congested and the layout cannot be solved.

また、タイミングの厳しい箇所には、論理合成時には高ドライブセルを使用していないので、レイアウト後に使用でき、高ドライブセルを再マッピングすることにより、レイアウト上で配線が回り込み、タイミングを収束することができなくなってしまうという問題を解決する。   In addition, high drive cells are not used for logic-synthesizing locations where logic synthesis is used, so they can be used after layout, and by re-mapping high drive cells, wiring can wrap around the layout and converge timing. Solve the problem of being unable to do so.

請求項1記載の発明は、上記課題を解決するため、論理設計の記述を解析する解析手段と、該解析に基づいてファンアウト数が所定の値以上となる箇所を抽出する抽出手段と、該抽出された箇所にレイアウト時に調整するためのクロックツリー用バッファを挿入する挿入手段と、を備えたことを特徴とするものである。   In order to solve the above problems, the invention according to claim 1 is an analysis means for analyzing the description of the logic design, an extraction means for extracting a portion where the fan-out number is a predetermined value or more based on the analysis, And an insertion means for inserting a clock tree buffer for adjustment at the time of layout in the extracted portion.

本発明によれば、RTLを解析し、ハイファンアウトネットを抽出し、そのネットに対し、レイアウト時に行うクロックツリーのためのバッファをRTL中へ挿入するので、論理合成時にレイアウトでの改善の余地を残すことができ、不要な負荷分散を防止し、レイアウト時に最適なセル配置、配線領域を考慮した負荷分散を行うことができる。   According to the present invention, an RTL is analyzed, a high fan-out net is extracted, and a buffer for a clock tree to be performed at the time of layout is inserted into the RTL, so there is room for improvement in layout at the time of logic synthesis. Therefore, unnecessary load distribution can be prevented, and load distribution can be performed in consideration of the optimum cell arrangement and wiring area during layout.

また、ライブラリを解析し、ドライブ能力の大きなセルは論理合成時には使用しないので、タイミングの厳しい箇所には、レイアウト時に高ドライブセルを再マッピングすることができ、タイミングを収束させた最適なレイアウトを行うことができる。   In addition, since the library is analyzed and cells with large drive capability are not used during logic synthesis, high drive cells can be remapped at the timing of layout, and optimal layout with converged timing can be performed. be able to.

以下、本発明の実施の形態について添付図面を参照しつつ説明する。   Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

本発明の処理フローを図1に示す。まず、設計者が作成したRTLをこの装置に読み込む(s1)。次に、読み込んだ情報をもとに、このRTLをVerilogのPLI(プログラマブル・ランゲージ・インターフェース)を用いて、RTLを解析(s2)し、ファンアウト数が多くなる記述箇所を検索する(s3)。ファンアウト数が多くなる記述例を、図2に示す。   The processing flow of the present invention is shown in FIG. First, the RTL created by the designer is read into this apparatus (s1). Next, based on the read information, this RTL is analyzed by using Verilog's PLI (Programmable Language Interface) (s2), and a description portion where the fan-out number increases is searched (s3). . A description example in which the number of fan-outs is increased is shown in FIG.

図2において、15行目からのcase文における論理では、sel[0], sel[1]信号はそれぞれ64×3個のマルチプレクサのセレクト信号となる(図3)。通常このまま論理合成を行うと、sel[0], sel[1]信号は、それぞれ64×3個のセルをドライブしなければならないため、フロアプラン等で見積もった負荷容量をもとに負荷分散され、バッファのツリーが形成される(図4)。   In FIG. 2, according to the logic in the case statement from the 15th line, the sel [0] and sel [1] signals are the select signals of 64 × 3 multiplexers (FIG. 3). Normally, if logic synthesis is performed as it is, the sel [0] and sel [1] signals must drive 64 × 3 cells, respectively, so the load is distributed based on the load capacity estimated by the floor plan. A tree of buffers is formed (FIG. 4).

ただし、このバッファリングは、レイアウトの配置によっては不必要な可能性がある。この負荷容量を正確に見積もるためには、実際にレイアウトを行なわないと判断できず、いかに事前にフロアプラン等を行い、負荷容量を見積もっていても、RTL修正等により再合成し、再レイアウトする度に、配置は変わってしまう可能性がある。   However, this buffering may be unnecessary depending on the layout arrangement. In order to accurately estimate the load capacity, it cannot be determined that the layout is not actually performed, and even if the floor plan is estimated in advance and the load capacity is estimated, it is re-synthesized by RTL correction and re-layout. Each time, the placement can change.

さらに、この不要なバッファリングにより、配線が増え、配線混雑度の問題が起こることがある。この配線混雑度の問題は、チップサイズ増大や、配線の回り込みによる遅延の増加を招く。   Further, this unnecessary buffering may increase wiring and cause a problem of wiring congestion. This wiring congestion problem causes an increase in chip size and an increase in delay due to wiring wraparound.

この問題を解決するため、本発明では、論理合成の段階では負荷分散させず、レイアウトで調整を行うように合成を行う。具体的には、図5に示すようにsel信号の出力にクロックツリーを行うためのバッファを挿入する(s4)。これにより、論理合成時に不要な負荷分散を行うことなくレイアウト工程へ進めることが可能になり、レイアウト時に実際に配置状況を考慮してファンアウト調整することが可能になる。また、クロックツリーにより、レイアウト時に優先的に最適なバッファ挿入が行なわれ、従来手法で起こっていた、不要に負荷分散したバッファチェーンが配線を混雑させるという問題も解決する。   In order to solve this problem, in the present invention, synthesis is performed such that adjustment is made in the layout without load distribution in the logic synthesis stage. Specifically, as shown in FIG. 5, a buffer for performing a clock tree is inserted into the output of the sel signal (s4). As a result, it is possible to proceed to the layout process without performing unnecessary load distribution at the time of logic synthesis, and it is possible to perform fanout adjustment in consideration of the actual arrangement state at the time of layout. In addition, the buffer tree preferentially inserts an optimum buffer at the time of layout, and solves the problem of unnecessarily load-distributed buffer chains congested with wiring, which has occurred in the conventional method.

次に、使用するテクノロジに対応した論理合成用のライブラリを読み込む(s5)。この際に、ライブラリを解析し(s6)、論理セルでドライブ能力の大きいセルは論理合成時に使用しないようにする(s7)。理由として、レイアウト後、遅延の問題が起こった際に、さらに大きなドライブ能力のセルをマッピングしなければならないことがある。ここで、もし論理合成時に最大のドライブ能力を持つセルを使用していると、それ以上改善することはできず対応が非常に困難となり、最悪の場合、再レイアウトということも考えられ、TATの大きな増加を招く。論理合成時にドライブ能力の大きいセルを使用しなければ、この問題を防ぐことができる。   Next, a logic synthesis library corresponding to the technology to be used is read (s5). At this time, the library is analyzed (s6), and a cell having a large drive capability as a logic cell is not used during logic synthesis (s7). The reason is that after the layout, when a delay problem occurs, a cell having a larger drive capability must be mapped. Here, if a cell with the maximum drive capability is used at the time of logic synthesis, it cannot be improved any further and it becomes very difficult to deal with, and in the worst case, it may be a re-layout. Incurs a large increase. This problem can be prevented if a cell having a large drive capability is not used during logic synthesis.

論理合成前に上記対応を行い、論理合成を行い(s8)、ネットリストを出力する(s9)。   The above correspondence is performed before logic synthesis, logic synthesis is performed (s8), and a net list is output (s9).

なお、上記実施の形態では、VerilogをHDL言語として例に挙げたが、その他の言語でも同様な解析は可能である。   In the above embodiment, Verilog is taken as an example of HDL language, but similar analysis is possible in other languages.

本発明に係る論理合成処理の一実施例を示すフローチャートである。It is a flowchart which shows one Example of the logic synthesis process which concerns on this invention. 一実施例のファンアウト数が多くなるRTL記述を示す図である。It is a figure which shows the RTL description which the number of fan-outs of one Example increases. 一実施例の回路図である。It is a circuit diagram of one Example. 一実施例の負荷分散された回路図である。FIG. 3 is a circuit diagram of load distribution according to an embodiment. 一実施例のクロックツリー化した回路図である。It is the circuit diagram which made the clock tree of one Example.

Claims (1)

論理設計の記述を解析する解析手段と、
該解析に基づいてファンアウト数が所定の値以上となる箇所を抽出する抽出手段と、
該抽出された箇所にレイアウト時に調整するためのクロックツリー用バッファを挿入する挿入手段と、
を備えたことを特徴とするLSI論理合成を行う論理合成装置。
An analysis means for analyzing the description of the logical design;
Extraction means for extracting a location where the fan-out number is equal to or greater than a predetermined value based on the analysis;
Insertion means for inserting a clock tree buffer for adjustment at the time of layout into the extracted portion;
A logic synthesis device for performing LSI logic synthesis, comprising:
JP2006028387A 2006-02-06 2006-02-06 Logic synthesizer Pending JP2006179021A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2006028387A JP2006179021A (en) 2006-02-06 2006-02-06 Logic synthesizer

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2006028387A JP2006179021A (en) 2006-02-06 2006-02-06 Logic synthesizer

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
JP2001111384A Division JP2002312411A (en) 2001-04-10 2001-04-10 Logic synthesizing device and method therefor

Publications (1)

Publication Number Publication Date
JP2006179021A true JP2006179021A (en) 2006-07-06

Family

ID=36732986

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2006028387A Pending JP2006179021A (en) 2006-02-06 2006-02-06 Logic synthesizer

Country Status (1)

Country Link
JP (1) JP2006179021A (en)

Similar Documents

Publication Publication Date Title
US5883808A (en) Logic circuit optimization apparatus and its method
CN109376467B (en) Clock tree layout flow method and clock tree deviation compensation device in integrated circuit
US8291364B2 (en) Automated digital circuit design tool that reduces or eliminates adverse timing constraints do to an inherent clock signal skew, and applications thereof
US8055884B2 (en) Method and apparatus for augmenting a pipeline with a bubble-removal circuit
US6910202B2 (en) Logic synthesis device and logic synthesis method
JP2007027841A (en) Design apparatus, method, and program for semiconductor integrated circuit
US6574781B1 (en) Design methodology for inserting RAM clock delays
JP2006146601A (en) Layout design method for semiconductor integrated circuit
US8578306B2 (en) Method and apparatus for performing asynchronous and synchronous reset removal during synthesis
US9646126B1 (en) Post-routing structural netlist optimization for circuit designs
JP2005123537A (en) Semiconductor device and its manufacturing method
JP2008305132A (en) Method for designing semiconductor integrated circuit and design support apparatus
US9965581B1 (en) Fanout optimization to facilitate timing improvement in circuit designs
US8453090B2 (en) System and method for optimizing logic timing
JP2006179021A (en) Logic synthesizer
US9710583B2 (en) Information processing apparatus, state machine dividing method, and computer-readable recording medium
US20080079468A1 (en) Layout method for semiconductor integrated circuit
JP4053969B2 (en) Semiconductor integrated circuit design apparatus and semiconductor integrated circuit design method
EP1564645A2 (en) Configurable memory system for embedded processors
US7962875B2 (en) Method, apparatus and program for designing circuits
JP2003256488A (en) Method for layout of lsi, program, and recording medium
US20230059055A1 (en) Routing layer re-optimization in physical synthesis
JP2008017389A (en) Design method for synchronous circuit having reduced emi noise
JP2006319162A (en) Method and program for clock tree generation
JP2009253756A (en) Layout generating method for clock distribution circuit, and semiconductor integrated circuitry

Legal Events

Date Code Title Description
A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20070123

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20070322

A02 Decision of refusal

Free format text: JAPANESE INTERMEDIATE CODE: A02

Effective date: 20070529