The part of execution in which an operand or instruction is read from main stora ge and written into a control unit or arithmetic unit register 執(zhí)行過程中的一個階段所需的時間,在此期間,計算機從主存儲器中取出指令或操作數(shù),并將其存入控制器或運算器的寄存器中。
In modern vlsi technology , hundreds of thousands of arithmetic units fit on a 1cm 2 chip . the challenge is supplying them with instructions and data . stream architecture is able to solve the problem well 在摩爾定律作用下,在單芯片上可集成的晶體管數(shù)快速增長,單芯片擁有成百上千的運算單元不再是問題,關(guān)鍵是如何給如此多的alu提供足夠的指令和數(shù)據(jù)。
3 - d graphics on mobile phones is quite similar to 3 - d graphics on pc in years past . there is no hardware acceleration , and processor speeds are quite low , and there also is the lack of floating point arithmetic unit in mobile phones 因此論文從通用的部分開始論述,然后明了移動平臺的特征,并試圖解釋三維引擎的一般原理和設(shè)計一個具有粗適性的基于游戲的三維圖形引擎。
2 montoye r k , hokenek e , runyon s l . design of the ibm risc system 6000 floating - point execution unit . ibm journal of research and development , 1990 , 34 : 59 - 71 . 3 oberman s . floating - point arithmetic unit including an efficient close data path 我們采用90納米cmos標(biāo)準(zhǔn)單元工藝以及synopsys自動布局布線流程進行實驗,實驗結(jié)果表明該算法在高性能雙通路結(jié)構(gòu)的浮點加減運算中引入后,可以使得近路徑的運算延遲整體降低10 . 2 % ,且算法本身沒有造成新的關(guān)鍵路徑。
So , we must design multimedia application - oriented computer architecture to fit the data processing demand of video compressing programs , we analyzed the parallelism of two representative video compressing programs - opendivx and tml9 , and drew a conclusion that it is effective to run video compressing application programs on the processor which uses parallel arithmetic units 相對于視頻壓縮應(yīng)用而言,普通計算機的處理能力大大落后于處理需求。因此,對于多媒體應(yīng)用,必須采用并行的方法來解決,但是不能簡單地使用普通并行機,必須針對這部分應(yīng)用的特點,采用并行的思想來設(shè)計面向多媒體應(yīng)用的計算機體系結(jié)構(gòu)。
With turning the scale of asic ( appl ication specified integrated circuits ) to s0c ( system on chip ) , which conunon1y is composed of mcu , specified function ip cores , memory , periphery interface etc , the ip reuse techno1ogy is very important in s0c design flow , which can realize the constructions of different levels components . the approach of configurable system , method and design f1ow for udsm ( u1tra deep sub micron ) asic , logic system design using hdl 1anguage , coding style , static and dynamic verification strategy are a1so presented in chapter 2 . in chapter 3 we study the vlsi - - dsp architecture design , dense computation and high speed high performance digital signal processing unit structure , which includes high speed mac components and distributed arithmetic unit 在工程設(shè)計方法及結(jié)構(gòu)化設(shè)計和高層次綜合的研究中,介紹了在深亞微米工藝條件使用的方法和asic設(shè)計流程,討論了高層次綜合的核心如何從描述推出電路構(gòu)成的設(shè)計思路,針對不同目標(biāo)的設(shè)計技巧討論了采用hdl語言進行邏輯系統(tǒng)設(shè)計的方法,給出了用vhdl語言進行代碼設(shè)計時的規(guī)范和風(fēng)格,在面向soc的驗證策略討論了動態(tài)和靜態(tài)的驗證技術(shù),提出了進行單獨模塊驗證、芯片的全功能驗證和系統(tǒng)軟硬件協(xié)同驗證的整體策略。
After that , it gives the measures of designing dsp ' s assembler as a part of the dsp ' s software development environment together with the c - compiler . moreover , this paper explores the method of design the floating - point arithmetic unit . referring to the ieee754 - 1985 standard for binary floating - point arithmetic , the algorithm and the behavior description of floating - point adder and multiplier is given , and the simulation and verification is shown at the end of this paper 此外,本文還對處理器的浮點運算單元設(shè)計做了初步的研究,以ansi ieee - 754浮點數(shù)二進制標(biāo)準(zhǔn)為參考,借鑒了經(jīng)典的定點加法器和乘法器的設(shè)計,嘗試性的給出了浮點加法單元和乘法單元的實現(xiàn)模型和行為級上的硬件描述,并對其進行仿真和驗證。