WO2016127701A1 - 一种基于无理数存储测试数据的解压方法 - Google Patents

一种基于无理数存储测试数据的解压方法 Download PDF

Info

Publication number
WO2016127701A1
WO2016127701A1 PCT/CN2015/097127 CN2015097127W WO2016127701A1 WO 2016127701 A1 WO2016127701 A1 WO 2016127701A1 CN 2015097127 W CN2015097127 W CN 2015097127W WO 2016127701 A1 WO2016127701 A1 WO 2016127701A1
Authority
WO
WIPO (PCT)
Prior art keywords
test
num
ate
irrational
decompression method
Prior art date
Application number
PCT/CN2015/097127
Other languages
English (en)
French (fr)
Inventor
詹文法
朱世娟
程一飞
吴海峰
Original Assignee
安庆师范学院
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 安庆师范学院 filed Critical 安庆师范学院
Publication of WO2016127701A1 publication Critical patent/WO2016127701A1/zh

Links

Classifications

    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03MCODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M7/00Conversion of a code where information is represented by a given sequence or number of digits to a code where the same, similar or subset of information is represented by a different sequence or number of digits
    • H03M7/30Compression; Expansion; Suppression of unnecessary data, e.g. redundancy reduction

Definitions

  • the invention relates to the field of testability design of integrated circuits, in particular to a method for decompressing test data based on irrational number storage of integrated circuits.
  • Integrated circuit products need to be tested to ensure the yield of their products.
  • VLSI ultra-large-scale integration
  • DFT full scan design
  • the full scan based test scheme improves the controllability and observability of the circuit, completely reducing the complexity of test generation.
  • this type of solution greatly prolongs the test application time, and there is a problem that the test cost is too high. Excessive testing costs have become a major issue in today's IC testing.
  • Test costs are related to many factors, and the increasing volume of test data is one of the important factors associated with test costs.
  • the test data grows exponentially year by year, and the huge data leads to the following problems: (1) The bandwidth between the hard disk and the Automatic Test Equipment (ATE) is limited, so that the test data is transmitted from the hard disk to the ATE for a longer time than the test data. The time that ATE is transmitted to the Circuit Under Test (CUT) will result in wasted time waiting for the test data to load from the hard disk to the ATE. (2) ATE has limited storage capacity, making it necessary to crop or load test data in batches. If the test data is cropped, the quality of the test will be degraded; if the test data is loaded in stages, the test time will increase.
  • ATE Automatic Test Equipment
  • a compression scheme based on nonlinear coding divides the original test data into a plurality of symbols (strings), each of which is replaced with one codeword to constitute compressed test data (T E ).
  • the T E is stored in the test device.
  • the compressed data is first loaded into the decoder by preprocessing, and all the code words are decompressed into corresponding strings by the decoder.
  • the decompressed data is then applied to the CUT, the response is captured and a response analysis is performed.
  • a compression scheme based on a linear decompressor This type of scheme uses linear operations to extend the data stored in the ATE to the test vectors required by the CUT. Compression based on linear decompressor is the research hotspot and focus of current test excitation compression technology. This type of technology can achieve higher compression ratios for test sets with a high X ratio. The compression process generally does not depend on the circuit under test and the test set, so it is especially suitable for test data compression of IP cores, most commercial tests. Compression tools use this type of technology.
  • the code-based test data compression technology is not well compatible with multiple scan chain structures. It is necessary to provide an independent decompression circuit for each scan chain to achieve the highest decompression efficiency.
  • the test usually has a Random Resistant Fault (RRF)
  • the two methods (2) and (3) have the disadvantages of low fault coverage and long test sequence.
  • RRF Random Resistant Fault
  • the test efficiency can be further improved by weighting or by adopting a mixed mode, as the circuit scale is expanded, the RRF is increased, and the required hardware overhead is significantly increased.
  • Chinese invention patent application 201210414485.X proposes a test data compression method for quickly finding irrational numbers, and discloses a dynamic coding compression technique, which does not directly store the run length by using code words, but forms a regular representation of the run length. (where m, l, k are all integers) irrational numbers, only four integers such as m, l, k and the original test data length p are stored for storage. The storage of the entire test set is converted into a single or several unreasonable integer storage.
  • a method of binary search for irrational numbers is proposed, which converts the calculation of irrational numbers into the search for irrational numbers, which reduces the complexity of the algorithm.
  • This method of converting test data into an irrational number to store data can theoretically compress test data indefinitely, and can fundamentally solve the storage problem of test data.
  • the difficulty is how to expand the irrational number into a decimal.
  • the computer can't complete the square root operation of big data. Even if it can be expanded into decimals, the time is very long, and the whole process of ATE conversion from irrational to decimal is needed. This process itself is also a waste of test cost.
  • the object of the present invention is to overcome the deficiencies of the prior art and to provide a method for decompressing stored test data represented by an irrational integer representation of improving test efficiency.
  • the present invention is implemented by the following technical solution, a decompression method for storing test data based on irrational numbers, which is applied to an automatic test device for testing a chip under test; the decompression method includes the following steps:
  • Step a transforming the original test set into a single or a plurality of integer representations corresponding to the irrational numbers, forming a regular representation of the length of the run length as Irrational numbers, where m, l, k are all integers, storing only m, l, k and the original test data length p during storage and stored in the control computer of the automatic test equipment;
  • Step d order Judging (t'l) the magnitude of k and m;
  • step h the test result is compared with the theoretical value. If the results are consistent, the tested chip passes the test; if not, the tested chip does not pass the test.
  • the decompression method further includes the step e.
  • the decompression method further includes the step f.
  • the invention is repeatedly iterated by the dichotomy method, and the test can be run while decompressing, which avoids the complicated square root operation and avoids the waiting time of the square root operation.
  • Step a transforming the original test set into an integer representation corresponding to a single or several irrational numbers, forming a regular representation of the length of the run as An irrational number, where m, l, k are all integers, and only m, l, k and the original test data length p are stored and stored in the control computer of the automatic test equipment.
  • b N-1 (b n +1); converts the run length of the integer part b 0 and the first n-1 fractional part b n-1 into a test vector, and then inputs the chip under test by the ATE channel, and records the transmission to the ATE channel.
  • the number of runs is num.
  • Step F t 2 and Comparative t 'runs the same number of data bits that is already decompressed, from the beginning to the num t 2 and t' after the same number of data bits is converted into the run length of test vectors via a transmission channel to the ATE The chip under test, and update num to the same number of data bits as t 2 and t'.
  • t 2 t' and repeat step d.
  • step g the run length from num to the end of p is transmitted to the chip under test through the ATE channel.
  • step h the test result is compared with the theoretical value. If the results are consistent, the tested chip passes the test; if not, the tested chip does not pass the test.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Test And Diagnosis Of Digital Computers (AREA)

Abstract

一种基于无理数存储测试数据的解压方法,其通过二分法多次迭代能边解压边运行测试,既避免复杂的开方运算也避免开方运算的等待时间。解压方法包括以下步骤:a,将原始测试集变换成至少一个无理数对应的整数: AA;b,由控制计算机估计出步骤a中单个无理数所对应的单精度或双精度小数:b 0.b 1b 2…b n-1b n,记t 1=b 0.b 1b 2…b n-1(b n-1),t 2=b 0.b 1b 2…b n-1(b n+1),将整数部分b 0和前n-1位小数部分b n-1对应游程长度转换成测试向量后依次输入被测芯片,并记录传输到ATE通道的游程数量num;c,记 BB;d,令 CC,判断(t'l) k与m的大小;g,在(t'l) k=m时,将从num开始到p结束的游程长度传输到被测芯片;h,将测试结果与理论值比较,如果结果一致,则被测芯片通过测试;否则不通过。

Description

一种基于无理数存储测试数据的解压方法 技术领域
本发明涉及集成电路可测试性设计技术领域,尤其设计集成电路的一种基于无理数存储测试数据的解压方法。
背景技术
集成电路产品需要进行测试以保证其产品的良率。近几十年来,随着超大规模集成(VLSI)技术的迅速发展,芯片中晶体管的密度成指数倍增加,IC测试成为半导体工业中最大的挑战之一。作为常见的可测试性设计(DFT)之一,全扫描设计广泛应用于IC测试中。基于全扫描的测试方案提高了电路的可控制性和可观察性,彻底地降低了测试生成的复杂性。然而,该类方案大大延长了测试应用时间,存在测试成本过高的问题。过高的测试费用已成为当今IC测试面临的主要问题。
测试成本与许多因素有关,其中日益增加的庞大的测试数据量是与测试成本相关的重要因素之一。测试数据逐年呈指数规律增长,庞大的数据导致了以下问题:(1)硬盘和自动测试设备(Automatic Test Equipment,ATE)之间带宽有限,使得测试数据从硬盘传输到ATE的时间大于测试数据从ATE传输到被测电路(Circuit Under Test,CUT)的时间,会导致浪费在等待测试数据从硬盘到ATE之间的加载时间加长。(2)ATE的存储容量有限,使得必须裁剪或分批加载测试数据。如果裁剪测试数据就会导致测试质量降低;如果分次加载测试数据,就会增加测试时间。(3)ATE与CUT之间的带宽有限,使得不能降低测试数据从ATE的存储器到CUT的加载时间。虽然更换高档次的ATE可以在一定程度上缓解上述问题,但这势必会增加测试成本(ATE价格在50-120美元/台)。上述问题是由于测试数据量增加带来的,显然,如果在测试质量不变的情况下减 少测试数据量,同样也能解决上述问题。因此迫切需要研究测试数据量减少技术。
关于测试数据量减少技术的研究,主要集中在三个方面。
(1).基于非线性编码的压缩方案。非线性编码将原始测试数据分割成多个符号(字符串),每个字符串用一个码字替代从而构成了压缩的测试数据(TE)。TE存储在测试设备中,在测试时,首先通过预处理将压缩后的数据载入解码器,所有的码字经解码器解压成相应的字符串。然后将解压后的数据施加到CUT,捕获响应并进行响应分析。
(2).基于广播的压缩方案。这类方案将相同的值广播到多条扫描链中。由于它的简单性和高效性,这种方法成为许多测试压缩结构的基础。
(3).基于线性解压器的压缩方案。这类方案利用线性操作将存储在ATE中的数据扩展成CUT需要的测试向量。基于线性解压器的压缩是目前测试激励压缩技术的研究热点和重点。这类技术对于X比例很高的测试集能够获得更高的压缩率,压缩过程一般不依赖于被测电路和测试集,因而特别适合IP芯核的测试数据压縮,绝大多数的商业测试压缩工具都采用这类技术。
由于在CUT与ATE之间数据传输存在着信号难以同步的缺点,不解决好同步问题,将会严重影响测试效率,改进通讯方式,又将会增加通讯协议的复杂性。另外,基于编码的测试数据压缩技术对多扫描链结构并不能很好的相容,需要对每一条扫描链都提供一个独立的解压电路才能使解压效率最高。
正是由于这些原因,对基于编码的测试数据压缩技术的研究仅停留在学术界,到目前为止还没有实用的相关EDA工具出现。
由于测试时通常存在着抗随机故障(Random Resistant Fault,RRF),故(2)和(3)两种方法存在故障覆盖率不高、测试序列较长的弊端。虽然可以通过加权或采用混合模式等方法来进一步提高测试效率,但随着电路规模的扩大,RRF的增多,所需要的硬件开销将显著增加。
中国发明专利申请201210414485.X提出一种快速查找无理数的测试数据 压缩方法,公开了一种动态编码压缩技术,并不直接用代码字来存储游程长度,而是将游程长度出现的规律表示成形如
Figure PCTCN2015097127-appb-000001
(其中m,l,k全部是整数)无理数,存储时只用存储m,l,k和原始测试数据长度p等四个整数。将对整个测试集的存储转换成了单个或若干个无理数对应的整数存储。另外,提出了一种二分查找无理数的方法,将对无理数的计算转换成对无理数的查找,减少了算法的复杂度。
这种将测试数据转换成无理数的方法来存储数据,从理论上可以无限压缩测试数据,可以从根本上解决测试数据的存储问题。然而该技术中存在如何将无理数还原成原始测试集的问题,难点是如何将无理数展开成小数。使用传统的方法,计算机无法完成大数据的开方运算,即使可以能够展开成小数,时间也非常长,需要ATE从无理数到小数转化的整个过程,这个过程本身也是测试成本的浪费。
发明内容
本发明的目的在于克服现有技术的不足,提供了一种提高测试效率的无理数整数表示的存储测试数据的解压方法。
本发明采用以下的技术方案实现,一种基于无理数存储测试数据的解压方法,其应用于自动测试设备中用于测试被测芯片;所述解压方法包括以下步骤:
步骤a,将原始测试集变换成单个或若干个无理数对应的整数表示,将游程长度出现的规律表示成形如
Figure PCTCN2015097127-appb-000002
无理数,其中m,l,k全部是整数,存储时只存储m,l,k和原始测试数据长度p并存储在所述自动测试设备的控制计算机中;
步骤b,由所述控制计算机估计出步骤a中单个无理数所对应的单精度或双精度小数:b0.b1b2…bn-1bn,其中b0、b1、b2、…、bn-1、bn为对应的游程长度,记t1=b0.b1b2…bn-1(bn-1),t2=b0.b1b2…bn-1(bn+1);将整数部分b0和前n-1位小数部 分bn-1对应游程长度转换成测试向量后依次由ATE通道输入被测芯片,并记录传输到ATE通道的游程数量num;
步骤c,记
Figure PCTCN2015097127-appb-000003
则:m=(tl)k,得到(t1l)k≤(tl)k=m≤(t2l)k
步骤d,令
Figure PCTCN2015097127-appb-000004
判断(t′l)k与m的大小情况;
步骤g,在(t′l)k=m时,将从num开始到p结束的游程长度通过ATE通道传输到被测芯片;
步骤h,将测试结果与理论值比较,如果结果一致,则被测芯片通过测试;如果不一致,则被测芯片不通过测试。
作为上述方案的进一步,所述解压方法还包括步骤e,在(t′l)k<m时,比较t1与t′相同的数据位数即为已经解压的游程数量,将从num开始到t1与t′相同的数据位数的游程长度通过ATE通道转换成测试向量后传输到被测芯片,并更新num为t1与t′相同的数据位数,令t1=t′,重复步骤d。
作为上述方案的进一步,所述解压方法还包括步骤f,在(t′l)k>m时,比较t2与t′相同的数据位数即为已经解压的游程数量,将从num开始到t2与t′相同的数据位数的游程长度通过ATE通道转换成测试向量后传输到被测芯片,并更新num为t2与t′相同的数据位数,令t2=t′,重复步骤d。
本发明通过二分法多次迭代,可以边解压边运行测试,既避免了复杂的开方运算,也避免了开方运算的等待时间。
具体实施方式
为了使本发明的目的、技术方案及优点更加清楚明白,以下结合实施例,对本发明进行进一步详细说明。应当理解,此处所描述的具体实施例仅仅用以解释本发明,并不用于限定本发明。
本发明的基于无理数存储测试数据的解压方法具体步骤如下。
步骤a,将原始测试集变换成单个或若干个无理数对应的整数表示,将游 程长度出现的规律表示成形如
Figure PCTCN2015097127-appb-000005
无理数,其中m,l,k全部是整数,存储时只存储m,l,k和原始测试数据长度p并存储在所述自动测试设备的控制计算机中。
由原始测试集,运行中国发明专利申请201210414485.X提出一种快速查找无理数的测试数据压缩方法,将原始测试集变换成单个或若干个无理数对应的整数表示。假设原始的测试集为001 00XX001 XX000XXXX1 XX1 000001 XXX00X0X1 001,可以转化为无理数
Figure PCTCN2015097127-appb-000006
即对应的m=29,l=2,k=2,p=41。
步骤b,由所述控制计算机估计出步骤a中单个无理数所对应的单精度或双精度小数:b0.b1b2…bn-1bn,其中b0、b1、b2、…、bn-1、bn为对应的游程长度,记t1=b0.b1b2…bn-1(bn-1),t2=b0.b1b2…bn-1(bn+1);将整数部分b0和前n-1位小数部分bn-1对应游程长度转换成测试向量后依次由ATE通道输入被测芯片,并记录传输到ATE通道的游程数量num。
假设由控制计算机算出
Figure PCTCN2015097127-appb-000007
控制计算机将2、6、9对应的游程长度转换成测试向量后依次由ATE通道输入被测芯片,传输的游程数量num=3。另有t1=2.691,t2=2.693。
步骤c,记
Figure PCTCN2015097127-appb-000008
则:m=(tl)k,得到(t1l)k≤(tl)k=m≤(t2l)k。从t1=b0.b1b2…bn-1(bn-1)、t2=b0.b1b2…bn-1(bn+1)很容易证明t1≤t≤t2,即有(t1l)k≤(tl)k=m≤(t2l)k
步骤d,令
Figure PCTCN2015097127-appb-000009
判断(t′l)k与m的大小情况。若(t′l)k<m,跳至步骤e;若(t′l)k>m,跳至步骤f;若(t′l)k=m,跳至步骤g。
步骤e,比较t1与t′相同的数据位数即为已经解压的游程数量,将从num开始到t1与t′相同的数据位数的游程长度通过ATE通道转换成测试向量后传输到被测芯片,并更新num为t1与t′相同的数据位数。令t1=t′,重复步骤d。
步骤f,比较t2与t′相同的数据位数即为已经解压的游程数量,将从num开始到t2与t′相同的数据位数的游程长度通过ATE通道转换成测试向量后传输到被测芯片,并更新num为t2与t′相同的数据位数。令t2=t′,重复步骤d。
步骤g,将从num开始到p结束的游程长度通过ATE通道传输到被测芯片。
步骤h,将测试结果与理论值比较,如果结果一致,则被测芯片通过测试;如果不一致,则被测芯片不通过测试。
上述例子中,首先有
Figure PCTCN2015097127-appb-000010
计算(t′l)k=(2.692×2)2=28.98746<m=29。此时t1和t′相同的位数为4位,分别为2、6、9、2,前三位2、6、9已经由ATE通道传输给被测芯片了。此次由ATE通道传输第四位数据2转换成测试向量001再传输到被测芯片,令t1=t′=2.692,重复步骤d有
Figure PCTCN2015097127-appb-000011
计算(t′l)k=(2.6925×2)2=28.99823<m=29。
此时t1和t′相同的位数为4位,该四位已经全部由ATE通道传输给被测芯片了,此次不再传输数据到被测芯片,令t1=t′=2.6925,重复步骤d有
Figure PCTCN2015097127-appb-000012
计算(t′l)k=(2.69275×2)2=29.00361>m=29。此时t2和t′相同的位数仍然为4位,该四位已经全部由ATE通道传输给被测芯片了,此次仍然不再传输数据到被测芯片,令t2=t′=2.69275,重复步骤d有
Figure PCTCN2015097127-appb-000013
计算(t′l)k=(2.69275×2)2=29.00092>m=29。
此过程一直进行到t′的前七位分别2、6、9、2、6、2、7、5,即有t′≈2.692582,结束测试过程。
以上所述仅为本发明的较佳实施例而已,并不用以限制本发明,凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等,均应包含在本发明的保护范围之内。

Claims (3)

  1. 一种基于无理数存储测试数据的解压方法,其应用于自动测试设备中用于测试被测芯片;其特征在于:所述解压方法包括以下步骤:
    步骤a,将原始测试集变换成单个或若干个无理数对应的整数表示,将游程长度出现的规律表示成形如
    Figure PCTCN2015097127-appb-100001
    无理数,其中m,l,k全部是整数,存储时只存储m,l,k和原始测试数据长度p并存储在所述自动测试设备的控制计算机中;
    步骤b,由所述控制计算机估计出步骤a中单个无理数所对应的单精度或双精度小数:b0.b1b2…bn-1bn,其中b0、b1、b2、…、bn-1、bn为对应的游程长度,记t1=b0.b1b2…bn-1(bn-1),t2=b0.b1b2…bn-1(bn+1);将整数部分b0和前n-1位小数部分bn-1对应游程长度转换成测试向量后依次由ATE通道输入被测芯片,并记录传输到ATE通道的游程数量num;
    步骤c,记
    Figure PCTCN2015097127-appb-100002
    则:m=(tl)k,得到(t1l)k≤(tl)k=m≤(t2l)k
    步骤d,令
    Figure PCTCN2015097127-appb-100003
    判断(t′l)k与m的大小情况;
    步骤g,在(t′l)k=m时,将从num开始到p结束的游程长度通过ATE通道传输到被测芯片;
    步骤h,将测试结果与理论值比较,如果结果一致,则被测芯片通过测试;如果不一致,则被测芯片不通过测试。
  2. 如权利要求1所述的基于无理数存储测试数据的解压方法,其特征在于:所述解压方法还包括步骤e,在(t′l)k<m时,比较t1与t′相同的数据位数即为已经解压的游程数量,将从num开始到t1与t′相同的数据位数的游程长度通过ATE通道转换成测试向量后传输到被测芯片,并更新num为t1与t′相同的数据位数,令t1=t′,重复步骤d。
  3. 如权利要求1所述的基于无理数存储测试数据的解压方法,其特征在于:所述解压方法还包括步骤f,在(t′l)k>m时,比较t2与t′相同的数据位数即为已经解压的游程数量,将从num开始到t2与t′相同的数据位数的游程长度通过ATE通道转换成测试向量后传输到被测芯片,并更新num为t2与t′相同的数据位数,令t2=t′,重复步骤d。
PCT/CN2015/097127 2015-02-12 2015-12-11 一种基于无理数存储测试数据的解压方法 WO2016127701A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510076243.8A CN104811208B (zh) 2015-02-12 2015-02-12 一种基于无理数存储测试数据的解压方法
CN201510076243.8 2015-02-12

Publications (1)

Publication Number Publication Date
WO2016127701A1 true WO2016127701A1 (zh) 2016-08-18

Family

ID=53695749

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/097127 WO2016127701A1 (zh) 2015-02-12 2015-12-11 一种基于无理数存储测试数据的解压方法

Country Status (2)

Country Link
CN (1) CN104811208B (zh)
WO (1) WO2016127701A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110798223A (zh) * 2019-11-28 2020-02-14 安庆师范大学 一种最小游程切换点标记编码压缩方法及装置

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104811208B (zh) * 2015-02-12 2016-10-12 安庆师范学院 一种基于无理数存储测试数据的解压方法
CN107026651B (zh) * 2017-04-11 2020-07-14 安庆师范大学 二分区间的数据压缩方法及集成电路的测试数据存储方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070011514A1 (en) * 2003-04-29 2007-01-11 Koniklijke Philips Electronics Data compression
CN101968528A (zh) * 2010-08-19 2011-02-09 詹文法 集成电路测试中的测试数据压缩方法
CN102932000A (zh) * 2012-10-25 2013-02-13 詹文法 快速查找无理数的测试数据压缩方法
CN104811208A (zh) * 2015-02-12 2015-07-29 安庆师范学院 一种基于无理数存储测试数据的解压方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070011514A1 (en) * 2003-04-29 2007-01-11 Koniklijke Philips Electronics Data compression
CN101968528A (zh) * 2010-08-19 2011-02-09 詹文法 集成电路测试中的测试数据压缩方法
CN102932000A (zh) * 2012-10-25 2013-02-13 詹文法 快速查找无理数的测试数据压缩方法
CN104811208A (zh) * 2015-02-12 2015-07-29 安庆师范学院 一种基于无理数存储测试数据的解压方法

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110798223A (zh) * 2019-11-28 2020-02-14 安庆师范大学 一种最小游程切换点标记编码压缩方法及装置

Also Published As

Publication number Publication date
CN104811208A (zh) 2015-07-29
CN104811208B (zh) 2016-10-12

Similar Documents

Publication Publication Date Title
WO2016127701A1 (zh) 一种基于无理数存储测试数据的解压方法
CN102522120A (zh) 一种无需存储字典的字典编码压缩方法
US20040117710A1 (en) Weight compression/decompression system
WO2021052329A1 (zh) 一种测试向量的无损压缩和解压缩方法
US7299236B2 (en) Test data compression and decompression method using zero-detected run-length code in system-on-chip
CN102932000B (zh) 快速查找无理数的测试数据压缩方法
CN102904579B (zh) 基于逐次逼近法的编码压缩方法
US11762019B2 (en) Method of testing electronic circuits and corresponding circuit
KR102656567B1 (ko) 다양한 유형의 신경망 모델을 변환 및 활용할 수 있도록 하는 방법 및 장치
WO2019137231A1 (zh) 一种译码方法及装置
TWI490855B (zh) 解壓縮電路與相關的解壓縮方法
US9762285B1 (en) Compression using mu-law approximation
Vohra et al. Optimal selective count compatible runlength encoding for SoC test data compression
CN110798223B (zh) 一种最小游程切换点标记编码压缩方法及装置
CN110021349A (zh) 基因数据的编码方法
TWI498891B (zh) 解壓縮電路與相關的壓縮方法與解壓縮方法
Chen et al. A dictionary-based test data compression method using tri-state coding
CN110865299B (zh) 一种基于前向相容的折叠集分层压缩方法及装置
US7529988B1 (en) Storage of descriptive information in user defined fields of failure bitmaps in integrated circuit technology development
Alkhalayleh et al. A new lossless method of image compression by decomposing the tree of Huffman technique
Biswas et al. Hybrid test vector compression in system-on-chip test—An overview and methodology
Karimi et al. Data compression for system-on-chip testing using ATE
US11967975B1 (en) Method and apparatus for recursive data compression using seed bits
CN112073069B (zh) 一种适用于集成电路测试的测试向量无损压缩方法
US20150236684A1 (en) Multiplexer circuit, computer-readable recording medium having stored therein program for designing multiplexer circuit, and apparatus for designing multiplexer circuit

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15881849

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15881849

Country of ref document: EP

Kind code of ref document: A1