返回项目页面
对比文件列表
1998-10-13_US5822619A_发明授权_US5822619A System and method for restructuring data strings_+++A_B_C_D_E_F_G_H_I_J_K_L_M_N_O_P+++.docx
2026-04-10 22:45
1999-08-03_US5933650A_发明授权_US5933650A Alignment and ordering of vector elements for single instruction multiple data processing_+++C_F_G_L_M_N_a_b_d_e_h_i_j_k_o_p+++.docx
2026-04-10 22:45
2001-01-09_US6173366B_发明授权_US06173366B2 Load and store instructions which perform unpacking and packing of data bits in separate vector and integer cache storage_+++A_B_C_D_F_G_H_J_N_P_a_e_i_k_l_m_o+++.docx
2026-04-10 22:45
2005-06-15_GB2409064A_发明专利_GB2409064A Performing in parallel a data processing operation on data elements_+++A_B_C_F_G_H_J_K_L_M_N_O_a_d_e_i_p+++.docx
2026-04-10 22:45
2007-09-27_US2007226469A_发明申请_US20070226469A1 Permutable address processor and method_+++a_b_c_d_e_f_g_h_j_k_l_m_n_o_p+++.docx
2026-04-10 22:45
2008-10-23_WO2008126041A_发明申请_WO2008126041A1 METHOD OF STORING DATA, METHOD OF LOADING DATA AND SIGNAL PROCESSOR_+++A_B_C_D_E_G_H_I_J_K_L_N_P_f_m_o+++.docx
2026-04-10 22:45
2009-02-05_US2009037694A_发明申请_US20090037694A1 Load Misaligned Vector with Permute and Mask Insert_+++H_N_c_g_j_k_m_p+++.docx
2026-04-10 22:45
2014-01-09_US2014013076A_发明申请_US20140013076A1 EFFICIENT HARDWARE INSTRUCTIONS FOR SINGLE INSTRUCTION MULTIPLE DATA PROCESSORS_+++N_a_b_c_j_k_l_m_p+++.docx
2026-04-10 22:45
2014-05-13_US8726281B_发明授权_US08726281B2 Method and system for improving performance and reducing energy consumption by converting a first program code into a second program code and implementing SIMD_+++a_b_c_d_g_h_j_k_m_n_p+++.docx
2026-04-10 22:45
2015-08-06_WO2015116474A_发明申请_WO2015116474A1 FORMAT CONVERSION DURING LOAD_STORE OPERATIONS IN A VECTOR PROCESSOR_+++A_B_C_E_G_H_I_J_K_L_M_N_P_d_f_o+++.docx
2026-04-10 22:45
2015-12-02_EP2950202A_发明公开_EP2950202A1 PROCESSOR AND DATA GATHERING METHOD_+++A_B_C_D_E_F_G_H_I_J_K_L_M_N_O_P+++.docx
2026-04-10 22:45
2017-02-02_US2017031865A_发明申请_US20170031865A1 APPARATUS AND METHOD FOR TRANSFERRING A PLURALITY OF DATA STRUCTURES BETWEEN MEMORY AND A PLURALITY OF VECTOR REGISTERS_+++H_I_J_K_L_M_N_a_b_c_d_e_g_o+++.docx
2026-04-10 22:45
2017-06-22_US2017177359A_发明申请_US20170177359A1 Instructions and Logic for Lane-Based Strided Scatter Operations_+++A_B_C_D_E_F_G_H_I_J_K_L_M_N_O_P+++.docx
2026-04-10 22:45
2017-07-06_US2017192784A_发明申请_US20170192784A1 Systems, Apparatuses, and Methods for Lane-Based Strided Gather_+++A_C_G_H_J_N_O+++.docx
2026-04-10 22:45
US2010106944A1_Description_20260324_1701_+++C_D_E_G_H_J_K_M_N_P_a_b+++.docx
2026-04-10 22:45
US2010332794A1_Description_20260324_1701_+++A_B_C_E_G_H_N_P_d_i_j_o+++.docx
2026-04-10 22:45
2014-01-09_US2014013076A_发明申请_US20140013076A1 EFFICIENT HARDWARE INSTRUCTIONS FOR SINGLE INSTRUCTION MULTIPLE DATA PROCESSORS_+++N_a_b_c_j_k_l_m_p+++.docx

# 对比文件技术特征公开性分析

## 对比文件信息

- **对比文件名称**:2014-01-09_US2014013076A_发明申请_US20140013076A1 EFFICIENT HARDWARE INSTRUCTIONS FOR SINGLE INSTRUCTION MULTIPLE DATA PROCESSORS

- **目标专利名称**:用于针对可变位宽度数据格式的打包和解包网络以及方法(CN111788553B)

- **模型名称**:本次分析基于专利对比文件技术特征公开性评估标准

## 特征比对表格

技术特征对比文件引用公开性论述
**技术特征A:《隐含公开》**[0027] "A SIMD operation implemented in hardware and may take one or more machine cycles to execute..."<br>[0027] "One approach for implementing a SIMD operation is to use one or more registers that are each capable of storing multiple distinct data values. Such registers are referred to herein as SIMD registers."<br>[0037] "SIMD register 110 comprises subregister 112, subregister 114, and subregister 116"<br>[0039] "Subregister 112 is loaded to contain the packed bits about to be unpacked"对比文件公开了SIMD寄存器包含多个子寄存器的结构,每个子寄存器相当于一个"位位置",子寄存器具有预定的位槽宽度(如16位)。对比文件[0027]明确说明SIMD寄存器可存储多个不同的数据值,这隐含公开了第一多个位位置中每个位位置均包括具有预定数目位槽的寄存器的技术特征。
**技术特征B:《隐含公开》**[0031] "Specifically, executing this instruction stores values originally formatted as a series of contiguous bits in a SIMD register, into separate contiguous subregisters. Furthermore, each value may be shifted and aligned within the subregister that the value is stored in."<br>[0041] "In step 2: A shift is performed to align each unpacked value to its respective target subregister."对比文件公开了每个值被存储在子寄存器的相应位置中,并通过移位和对齐操作确保数据正确存放。这隐含公开了每个位槽均被配置为临时存储数据中的相应位的技术特征。
**技术特征C:《隐含公开》**[0032] "bunpk2 s2=s1,r1, assuming the value stored in r1 is three, and upon execution of the example, 3-bit values stored contiguously in register s1 may be unpacked into the 2-byte subregisters in s2."<br>[0167] "Some of the embodiments illustrated above each subregister may be four, eight, or sixteen bits wide. Other embodiments may include other widths, including, but in no way limited to four bits, eight bits, thirty-two bits, sixty-four bits."对比文件公开了子寄存器可具有不同的位宽度(4位、8位、16位、32位、64位等),且bunpk指令可处理不同位宽度的值。这隐含公开了预定数目的位可对应于若干不同数据位宽度中的一个数据位宽度的技术特征。
**技术特征D:《未直接公开》**[0030] "A SIMD instruction, fixed-width bit unpack ("bunpk"), may be used to unpack fixed-width bit values into fixed-width bytes."<br>[0031] "Specifically, executing this instruction stores values originally formatted as a series of contiguous bits in a SIMD register, into separate contiguous subregisters."对比文件公开的是解包(unpack)操作,将连续的位值展开到子寄存器中。而目标专利技术特征D要求的是**打包(pack)**操作,将数据从第一多个位位置打包到第二多个位位置。对比文件虽然公开了位操作相关的SIMD指令,但未公开将位从第一多个位位置打包到第二多个位位置的固定位打包网络。
**技术特征E:《未直接公开》**[0040] "In step 1: A shuffle is performed to copy the packed values into corresponding subregisters to be unpacked. For example, the entire contents of subregister 112, which holds the first three values, are copied and stored in subregister 122, subregister 124, and subregister 126"对比文件公开的是解包操作中的数据复制和移位过程,而非打包操作中"将相同特定位槽位置中的所有位路由到位束"的技术特征。对比文件虽然描述了数据在不同寄存器间的传输,但未涉及形成位束的打包路由机制。
**技术特征F:《未直接公开》**无相关引用对比文件未公开"固定的点对点连接"这一技术特征。对比文件描述的是SIMD指令的操作过程,未涉及硬件网络连接的固定点对点结构。
**技术特征G:《未直接公开》**[0030]-[0035] 描述的是解包指令bunpk处理固定宽度位值的能力,而非打包网络容纳不同数据位宽度的能力。对比文件公开的是解包指令对不同位宽的处理能力,而非"固定的位打包网络被配置为容纳若干不同数据位宽度中的每一个"的技术特征。
**技术特征H:《未直接公开》**[0030] "A SIMD instruction, fixed-width bit unpack ("bunpk"), may be used to unpack fixed-width bit values into fixed-width bytes."<br>[0060] "A SIMD instruction, variable-length byte unpack ("vlbunpk"), may be used to unpack variable-length byte values into fixed-width bytes."对比文件公开的是解包操作(unpack),而非打包操作(pack)。目标专利要求将数据从第一多个位位置打包到第二多个位位置,而对比文件描述的是相反方向的操作。
**技术特征I:《未直接公开》**无相关引用对比文件未明确比较解包操作中源位置数量与目标位置数量的关系。解包操作中数据被展开到更多子寄存器中,与目标专利技术特征I所述的"第二多个位位置比第一多个位位置少"相反。
**技术特征J:《隐含公开》**[0083]-[0089] "burst [r1]=r2,r3,r4. In the embodiment above, r1 is a pointer into memory indicating the byte at which the resulting decompressed bit-vector should be stored"<br>[0102] "Memory 390 may be on-chip cache located on a processor. Alternatively memory 390 may be off-chip cache, or main memory. Memory 390 may store full or partial decompressed bit-vectors"对比文件通过burst指令公开了第二多个位位置可包括在存储器设备中,存储器用于存储解压缩后的位向量。这隐含公开了第二多个位位置包含在存储器设备内且被配置为存储位束的技术特征。
**技术特征K:《隐含公开》**[0101] "Register 330 corresponds to the result register and contains the partially decompressed bit-vector at the end of step 1. Specifically, register 330 contains the first four bytes of the decompressed bit-vector"<br>[0099]-[0101] 描述了burst指令中使用多个存储位置存储结果对比文件隐含公开了第二多个位位置可包括多个存储器单元的概念。
**技术特征L:《隐含公开》**[0101] "Specifically, register 330 contains the first four bytes of the decompressed bit-vector"<br>[0093] "In this example, the result register is four bytes wide"对比文件隐含公开了位束的存储大小与寄存器或存储器字长相关的技术特征。
**技术特征M:《隐含公开》**[0024] "The instructions may operate over memory, registers, or immediate operands. Furthermore, instructions with memory operands may operate on data with different degrees of granularity, such as a cache-line or page-sized chunk."<br>[0083]-[0089] burst指令的存储器形式对比文件公开了SIMD指令可在存储器数据路径上操作,隐含公开了打包/解包网络可作为计算数据路径网络或存储器数据路径的一部分。
**技术特征N:《直接公开》**[0006] "Single instruction multiple data ("SIMD") processors perform the same operation on multiple data items simultaneously. SIMD processors exploit data level parallelism by executing a single instruction against data in multiple registers or subregisters."<br>[0025] "A SIMD instruction is an instruction that, when processed, may cause an operation to be performed multiple times, simultaneously and/or in parallel, on multiple distinct data values."对比文件明确公开了SIMD处理器通过在多个寄存器或子寄存器上执行单条指令来实现数据级并行处理。这直接公开了第一多个位位置中的位由具有若干通道的SIMD单元生成的技术特征。SIMD处理器的多个通道对应于目标专利中的第一多个位位置。
**技术特征O:《未直接公开》**无相关引用对比文件未公开位束大小与SIMD单元通道数倍数关系的技术特征。
**技术特征P:《隐含公开》**[0030]-[0043] 详细描述了固定宽度位解包指令bunpk的执行过程,包括混洗、移位和对齐操作。<br>[0060]-[0071] 描述了可变长度字节解包指令vlbunpk的执行过程。对比文件公开了SIMD指令可用于执行解包操作,将打包的位值解包到子寄存器中。解包操作本质上是打包操作的反向操作,隐含公开了装置可操作以用于解包操作的技术特征。

## 综合分析结论

**直接公开的技术特征**:技术特征N

**隐含公开的技术特征**:技术特征A、技术特征B、技术特征C、技术特征J、技术特征K、技术特征L、技术特征M、技术特征P

**未直接公开或隐含公开的技术特征**:技术特征D、技术特征E、技术特征F、技术特征G、技术特征H、技术特征I、技术特征O

**特别说明**:对比文件主要涉及SIMD处理器的指令设计,公开了大量解包(unpack)相关的指令和操作,但目标专利的核心发明点在于**固定的位打包网络(fixed bit packing network)**将数据从第一多个位位置打包到第二多个位位置。对比文件未公开打包方向的操作、未公开固定的点对点连接结构、未公开位束形成机制,以及未公开"第二多个位位置比第一多个位位置少"的压缩存储技术特征。

<<<a>>> <<<b>>> <<<c>>> <<<N>>> <<<j>>> <<<k>>> <<<l>>> <<<m>>> <<<p>>>

权利要求与技术特征
Powered by Django

网站备案号:渝ICP备2023012882号


重庆市非显而易见网络科技有限责任公司 A Anti NPE NPE