会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 1. 发明授权
    • Fault modeling for state retention logic
    • 状态保留逻辑的故障建模
    • US08296703B1
    • 2012-10-23
    • US12339958
    • 2008-12-19
    • Krishna ChakravadhanulaSteven L. GregorBrion L. KellerVivek Chickermane
    • Krishna ChakravadhanulaSteven L. GregorBrion L. KellerVivek Chickermane
    • G06F9/455G06F17/50
    • G11C29/56008G06F2217/78
    • A method for modeling state-retention logic includes: specifying a circuit that includes an arrangement of circuit elements, wherein a portion of the circuit is organized into a power domain with a power-domain control for effecting power variations within the power domain, and the power domain includes a state-retention cell that includes a retention element with a retention-element control for saving state-retention-cell values in the retention element during power variations in the power domain; determining one or more pattern faults for detecting defects in state-retention operation of the circuit by associating circuit element values with values for the power-domain control or the retention-element control; and saving one or more values for the one or more pattern faults.
    • 一种用于建模状态保持逻辑的方法包括:指定包括电路元件布置的电路,其中电路的一部分被组织成具有功率域控制的功率域,以实现功率域内的功率变化,并且 功率域包括状态保持单元,其包括具有保持元件控制的保持元件,用于在功率域的功率变化期间在保持元件中保持状态保持单元值; 通过将电路元件值与功率域控制或保持元件控制的值相关联来确定用于检测电路的状态保持操作中的缺陷的一个或多个模式故障; 并为一个或多个模式故障保存一个或多个值。
    • 4. 发明授权
    • Built-in self test system and method for two-dimensional memory redundancy allocation
    • 内置自检系统和二维内存冗余分配方法
    • US06907554B2
    • 2005-06-14
    • US10249817
    • 2003-05-09
    • R. Dean AdamsThomas J. EckenrodeSteven L. GregorGary S. Koch
    • R. Dean AdamsThomas J. EckenrodeSteven L. GregorGary S. Koch
    • G11C29/44G11C29/00
    • G11C29/4401G11C29/44G11C2029/0401
    • A built-in self test system (124) and method for two-dimensional memory redundancy allocation. The built-in self test system is adapted to allocate two redundant columns (116) and one redundant row (120) to an embedded memory (104) as needed to repair single cell failures (SCFs) within the rows (108) and columns of the memory. The self-test system includes a left-priority encoder (136), a right-priority encoder (140), and a greater-than-two detector (144). The left-priority encoder encodes the location of the first SCF most proximate the most-significant bit of the corresponding word. The right-priority encoder encodes the location of the first SCF most proximate the least-significant bit of the corresponding word. The greater-than-two detector determines whether a word contains more than two SCFs. If the greater-than-two detector detects that a word contains more than two SCFs, the built-in self test system identifies the corresponding row as being a must-fix row, since the number of SCFs exceeds the number of redundant columns.
    • 内置自检系统(124)和二维内存冗余分配方法。 所述内置自检系统适于根据需要将两个冗余列(116)和一个冗余行(120)分配给嵌入式存储器(104),以修复行(108)和列(108)中的单个单元故障(SCF) 记忆。 自检系统包括左优先编码器(136),右优先编码器(140)和大于二检测器(144)。 左优先级编码器对最接近相应字最高有效位的第一SCF的位置进行编码。 右优先编码器编码最接近对应字的最低有效位的第一SCF的位置。 大于2的检测器确定一个单词是否包含两个以上的SCF。 如果大于2的检测器检测到一个字含有两个以上的SCF,则内置的自检系统将相应的行识别为必须修复的行,因为SCF的数量超过冗余列的数量。
    • 5. 发明授权
    • Programmable memory built-in self-test combining microcode and finite state machine self-test
    • 可编程存储器内置自检组合微码和有限状态机自检
    • US06651201B1
    • 2003-11-18
    • US09626715
    • 2000-07-26
    • R. Dean AdamsThomas J. EckenrodeSteven L. GregorKamran Zarrineh
    • R. Dean AdamsThomas J. EckenrodeSteven L. GregorKamran Zarrineh
    • G01R3128
    • G01R31/3187G01R31/2891G11C29/16
    • A finite state machine (FSM) is used to generate, in real time, potentially long sequences of signals which control generation of signals for application to a memory structure during a self-test procedure which is provided in hardware on the same chip with the memory structure. The FSM-based instruction generator requires much less area than is required for storage of a corresponding number of microcode instructions and allows the built-in self-test (BIST) controller to have a modular architecture permitting re-use of hardware designs for the BIST arrangement with consequent reduction of elimination of design costs of the BIST arrangement to accommodate new memory designs. The sequential nature of the operation of a finite state machine as it progresses through a desired sequence of states is particularly well-suited to controlling capture of signals where access to high. speed data transfer circuits cannot otherwise be accommodated.
    • 有限状态机(FSM)用于实时地产生潜在的长序列信号,该序列控制信号的产生,以便在与存储器相同的芯片上的硬件中提供的自检过程期间应用于存储器结构 结构体。 基于FSM的指令生成器需要比存储相应数量的微代码指令所需的面积少得多的内存自检(BIST)控制器具有模块化架构,允许重新使用硬件设计用于BIST 从而减少了BIST安排的设计成本的消除以适应新的存储器设计。 有限状态机在其进行期望的状态序列的操作的顺序性质特别适合于控制对高访问信号的捕获。 速度数据传输电路不能适应。
    • 8. 发明授权
    • Programable multi-port memory BIST with compact microcode
    • 可编程多端口存储器BIST具有紧凑的微码
    • US07168005B2
    • 2007-01-23
    • US10354535
    • 2003-01-30
    • R. Dean AdamsThomas J. EckenrodeSteven L. GregorKamran Zarrineh
    • R. Dean AdamsThomas J. EckenrodeSteven L. GregorKamran Zarrineh
    • G06F11/00
    • G06F11/2242G11C8/16G11C29/16
    • A microcode programmable built-in-self-test (BIST) circuit and method for testing a multiported memory via multiple ports, either simultaneously or sequentially, as directed by a microcode instruction word. The microcode instruction word contains a plurality of executable subinstructions and one bit of information that controls whether the test operations prescribed in the plurality of subinstructions shall be executed in parallel or in series. The executable subinstructions are dispatched by a primary controller to subcontrollers which perform test operations at each port according to the subinstructions. The microcode programable BIST architecture flexibly facilitates the testing of multiple devices, multiported devices, including multiported memory structures and complex dependent multiported memory structures. The BIST supports in-situ testing of the functionality of the memory at wafer, module, and burn-in, as well as system-level testing.
    • 微代码可编程内置自检(BIST)电路和方法,用于通过多个端口同时或顺序地测试多端口存储器,如微代码指令字所指示的。 微代码指令字包含多个可执行子指令和一位信息,用于控制多个子指令中规定的测试操作是并行还是串行执行。 可执行子指令由主控制器分派到根据子提示在每个端口执行测试操作的子控制器。 微码可编程BIST架构灵活地促进了多个设备,多端口设备的测试,包括多端口存储器结构和复杂的多端口存储器结构。 BIST支持在晶片,模块和老化模式下对存储器的功能进行现场测试,以及系统级测试。
    • 9. 发明授权
    • Storage protection keys in two level cache system
    • 二级缓存系统中的存储保护密钥
    • US5450563A
    • 1995-09-12
    • US968767
    • 1992-10-30
    • Steven L. Gregor
    • Steven L. Gregor
    • G06F12/08G06F12/14G06F12/00
    • G06F12/0875G06F12/1475G06F12/0897
    • The cache system comprises a level one (L1) data cache, a level one (L1) key cache for storing a plurality of access keys for respective pages or blocks of data referenced by the central processor. A level three (L3) storage stores the data requested by the central processor and an access key array including the plurality of access keys. A level two (L2) data cache is coupled between the L3 storage and the L1 data cache and stores a copy of data fetched from the L3 storage for the L1 data cache pursuant to a read request and data written by the central processor. The level two (L2) key cache is coupled between the L3 storage access key array and the L1 key cache and stores the plurality of access keys for respective pages or blocks of data in the L2 data cache. The general strategy of the cache system is to copy the access key corresponding to each central processor request into the L1 key cache the first time that any page of data is inpaged pursuant to a read or write request, so that the access key is readily available to the central processor for the current and subsequent requests. If the central processor request is to store data to the L2 data cache, the corresponding access key but not the data is copied into the L1 level, and the data is updated only in the L2 data cache, assuming that the data did not reside in the L1 data cache due to a previous read request.
    • 高速缓存系统包括一级(L1)数据高速缓存,一级(L1)密钥高速缓存,用于存储由中央处理器引用的相应页面或数据块的多个访问密钥。 三级(L3)存储器存储由中央处理器请求的数据和包括多个访问密钥的访问密钥阵列。 第二级(L2)数据高速缓存耦合在L3存储器和L1数据高速缓冲存储器之间,并且根据读取请求和由中央处理器写入的数据存储从用于L1数据高速缓存的L3存储器取出的数据的副本。 第二级(L2)密钥缓存耦合在L3存储访问密钥阵列和L1密钥高速缓存之间,并将多个访问密钥存储在L2数据高速缓存中的各个页面或数据块。 缓存系统的一般策略是在第一次根据读或写请求对任何数据页进行页面复制时,将与每个中央处理器请求相对应的访问密钥复制到L1密钥高速缓存中,使得访问密钥容易获得 到当前和后续请求的中央处理器。 如果中央处理器请求是将数据存储到L2数据高速缓存中,则相应的访问密钥而不是数据被复制到L1级,并且仅在L2数据高速缓存中更新数据,假设数据不在 由于先前的读取请求导致的L1数据高速缓存。