部分外企筆試真題總結(jié)

PAGEPAGE10/ 題目 1題目 2題目 2題目D(MS 題目 3題目 3 題目 題目 題目 題目 題目 Goldman 技術(shù) 2、F(n)=1n>8n<12F(n)=2n<2F(n)=3n=6F(n)=4使用*sign(n)F(n)sign(n)=0n=0sign(n)=‐1n<0sign(n)=1n>0?91~100全部開(kāi)關(guān)朝上（開(kāi)）1的倍數(shù)反方向撥一次開(kāi)關(guān)；2的倍數(shù)反方向又撥一次開(kāi)關(guān)；3的倍數(shù)反方向又撥一次開(kāi)關(guān)……問(wèn)：最后為關(guān)熄狀態(tài)的燈101b（a2（b）3（c）4（d）5（e）6（a）2，3（b）3，4（c）4，5（d）5，6（e）6，7；4、答案是a的問(wèn)題的個(gè)數(shù)是：（a）0（b）1（c）2（d）3（e）4（a）10（b）9（c）8（d）7（e）66、答案是a（a）（bc（cd（）e（e）（a）4（b）3（c）2（d）1（e）0（a）2（b）3（c）4（d）5（e）6（b）（c）（d）（e）5（a）a（b）b（c）c（d）d（e）eD(MSIQ測(cè)驗(yàn)，5EQ測(cè)試，64道題，第三部分，寫(xiě)作/翻譯30iq題）2d5d3堂，有可能是a、b、c三個(gè)同學(xué)盜竊已知：1、a2、b3、c4、a、b、c每個(gè)人都上了老師d6、老師d的課中有一堂，三個(gè)人中到了兩個(gè)，事后證明不是那兩個(gè)人偷得3abcdggggg問(wèn)，abcdefg各市什么數(shù)字（不重復(fù)CD30km/ha應(yīng)當(dāng)怎么做才能夠使這次往返旅行的平60km/h5圖形）EQ題會(huì)問(wèn)一些個(gè)人性格方面的問(wèn)題，比如你是否易怒~~~~64道應(yīng)聘SST的會(huì)要求用給定的場(chǎng)景寫(xiě)一篇英文的商務(wù)文章TST300wordsEAP的翻譯，還有一個(gè)朋友碰到了.netframework的翻譯，不是很難，解決方案就是**更換電池或給50元購(gòu)買(mǎi)該廠(chǎng)家新手機(jī)的折換券。請(qǐng)給所有已購(gòu)買(mǎi)的用戶(hù)寫(xiě)信告321.2萬(wàn)元錯(cuò)賣(mài)給李先生，王小姐的經(jīng)理15ABWC，PSI這個(gè)部分有近二十道多項(xiàng)選擇題和若干填空題,內(nèi)容繁多零散,大部分是關(guān)于數(shù)據(jù)結(jié)構(gòu)和c/c++語(yǔ)1。regularexpression不能描述的是?ints(intintcount=0;intx=v;while(x)return4,C語(yǔ)言中，.h文件和.c7,1，2，3，4，5，689regularlanguagecontextlanguage之間的關(guān)系是什么a)regularlanguagemorepowerfulthancontextfree(后面兩個(gè)個(gè)選項(xiàng)記不清楚了,反正大意就是比較兩個(gè)誰(shuí)好.最后一個(gè)選項(xiàng)是兩者不能進(jìn)行這樣的比較)10testing11typeerror一個(gè)線(xiàn)性車(chē)庫(kù),55輛車(chē),一車(chē)最多坐兩人6個(gè)人.Tom,Jack,GawenLaurie,Mark.Mark.TomGawenPaul坐一輛車(chē).Laurie跟別人共享車(chē).Gawen在第三排或第四排5道題目,4個(gè)條件不能確定車(chē)序.有的題目是問(wèn)存在的可能,有的題目是附加條件后1TranslateMIPSassemblycodeintoafunctioninyourcodeshouldbenoanyGOTOs/pointersMIPScode:

liv0,0liaddt1,a0,a0lbt2,0(t1)beqt2,zero,l3bnet2,a1,l2addv0,v0,1addt0,t0,1 jr(callerregister:t0~t9,a0~a3,v0~v1;callee然后列出了指令表,li是賦值,lb是將字段后第一個(gè)寄存器內(nèi)容個(gè)字節(jié)的內(nèi)容復(fù)制到第二個(gè)寄存器,beqben是不等值轉(zhuǎn)移,j是無(wú)條件轉(zhuǎn)移,jr轉(zhuǎn)移至寄存器標(biāo)示的內(nèi)容.arr為指向數(shù)組的指針len為數(shù)組長(zhǎng)度count表示數(shù)組元素?cái)?shù)目.intinsert(int*arr,size_tlen,size_tcount,intval)intsearch(int*arr,size_tlen,size_tcount,intval)intremove(int*arr,size_tlen,size_tcount,intval)strCmp(constchar*string1,constchar*string2)以及其返回值表明的意義,(>0,==0,<0),設(shè)計(jì)測(cè)試case10Case，關(guān)于電訊盈科近期大力發(fā)展的兩個(gè)業(yè)務(wù)，還有一些評(píng)價(jià)。500字以?xún)?nèi)寫(xiě)一個(gè)投資建議，IBD一樣FC全英文，3小時(shí)第一部分：讀圖題(12題)，第二和第五部分：閱讀理解(15題)：第三部分：計(jì)算題(轉(zhuǎn)換思維)，承壓能力(解題強(qiáng)度較大，65分鐘內(nèi)不可有絲毫松懈)；迅速進(jìn)入狀態(tài)(考前無(wú)讀題時(shí)?Thebigeconomicdifferencebetweennuclearandfossil‐fuelledpowerstationsisthatnuclearreactorsaremoreexpensivetobuildanddecommission,butcheapertosun.Sodisputesovertherelativeefficiencyofthetwosystemsrevolvenotjustaroundpricesofcoalanduraniumtodayandtomorrow,butalsoaroundthewayinwhichfutureincomeshouldbecomparedwithcurrentincome.?Themaindifferencebetweennuclearandfossil‐fuelledpowerstationsisaneconomicone.ThepriceofcoalisnotrelevanttodiscussionsabouttherelativeefficiencyofnuclearIfnuclearreactorswerecheapertobuildanddecommissionthanfossil‐fuelledpowerstations,theywoulddefinitelyhavetheeconomicadvantage.?Atanygivenmomentwearebeingbombardedbyphysicalandpsychologicalcompetingforourattention.Althoughoureyesarecapableofhandlingmorethan5millionbitsofdatapersecond,ourbrainarecapableofinterpretingonlyabout500bitspersecond.Withsimilardisparitiesbetweeneachoftheothersensesandthebrain,itiseasytoseethatwemustselectthevisual,auditory,ortactilestimulithatwewishtocomputeatanyspecificPhysicalstimuliusuallywininthecompetitionforourattention.Thecapacityofthehumanbrainissufficienttointerpretnearlyallthestimulithesensescanregisterunderoptimumconditions.Eyesareabletocopewithagreaterinputofinformationthanears.CCANNOTBACCANNOTBCCANNOTSAYPartIINUMERCALTESTWhichcountryhadthehighestnumberofpeopleaged60oroveratthestartofWhatpercentageofthetotal15mmbuttonproductionwasclassedassub‐standardinAA10.5%BB13%CC15%DD17.5%EE20%AB23.5%ACAD27.5%AE28%BCHowmanylivebirthsoccurredin1985inSpainandItalytogether(tothenearestWhatwastheneteffectontheUKpopulationofthelivebirthanddeathratesinDecreaseofIncreaseofIncreaseofCannotIncreaseofByhowmuchdidthetotalsalesvalueofNovember?sbuttonproductionvaryfromNoWhatwasthelossinpotentialsalesrevenueattributabletotheproductionofsub‐standard(asopposedtostandard)buttonsoverthe6monthperiod?1：PopulationStructure1985populationatstartoflivebitrhsper1000population(jan‐dec)deathsper1000population(jan‐dec)%ofpopulationatstartofyearaged:%ofpopulationatstartofyearaged:60orUK56.613.311.819France55.213.910.021Italy19W.Germany61.09.611.515Spain38.612.17.7232:productionof15mmtotal(standardandsub‐standard)production(inthousands)standardproduction(inthousands)july70aug60sept85oct100nov95dec100saleprice:standard:$5.7per100sub‐stand:$2.85per100DW.ABCBIncreaseofENoCfloat:a,b,c問(wèn)值),xy2某個(gè)程序在一個(gè)嵌入式系統(tǒng)(200MCPU,50MSDRAM)中已經(jīng)最化了，換到另一個(gè)系統(tǒng)(300MCPU,50M的SDRAM)中運(yùn)行，還需要優(yōu)化嗎？什么情況下，sin(x+y)+y~12354aquicksortb.bublesortc.mergebinaryhash1：概率題。x,yf(x,y)intig(0,1)*dx*intig(0,x)*k*dy，k為常數(shù)，k=?E(xy)=?注：intig(a,b)為abP(AUB)*p(AB)<=P(AUB)*p(AB)>=P(AUB)*p(AB)<=P(A)+P(AUB)*p(AB)>=P(A)+3:200kHz,10dB，求信道波特率intmain()inta,b,c,abc=0;intabc=printf(?%d,%d?,abc,c);return0;7DFT2^N長(zhǎng)度的離散傅立葉變換的方法有哪些？如何用fft求N*M點(diǎn)的離散卷機(jī)？8：給出firiir1cmain2。15N3GPIO,intel的四種體系結(jié)構(gòu)10題vcc，指針，HyporThreadingDual‐core1000itemitem形式為（x,y)overlappedNkbk為多少最能節(jié)省存儲(chǔ)空間（N/(k‐b)mod1=1/2）amount(貨物數(shù)*單價(jià)）JeffandDiamondlikeplayinggameofcoins,OnedaytheydesignedanewsetofTotally10Onecantakeaway1,2or4coinsatonetimebyWhotakesthelastGiventheserulesWhetherthewinningstatusispre‐determinedor22，3，581，4，7才能贏，我先抽得話(huà)不可能達(dá)到這幾個(gè)數(shù)，很顯然，只能讓對(duì)方先TheUIspecialistDafnafoundaproblemthatsomeoftheItemsonthemarketingdocumentformoverlappedwitheachother.Becausethisformwasimplementedbydifferentdevelopersandtheydidn?tcaretheparticularappearanceofoneitem.ProductmanagerTidavdecidedtowriteonesmallcheckingtooltogeneratetheoverlappeditemsonallforms.Hecalledinhisguystodiscussaboutit.Supposetheinputistheintegercoordinates(x,y)odtheitems(allrectangles)ononeform.Constructanefficientmethodtofindouttheoverlappeditems.Hint:Themostdirectwaytodosoiscomparingeachitemswiththeothers,Given1000forms.Eachwith100‐1000itemsonaverage.TheO(n2)algorithmiscostly.SomegurusuggestedthatoneO(n)methodcouldhelponlyif6.5kilobytesextrastorageisavailable.Oneelitearguedthathecouldcutdownthenumberto1%,It?snowyourturntodescribetheidea.Writeoutthepseudocodes,vertifyhisalgorithmandproposemoreadvancedoptimizationifpossible.3inafilesystem,dataneednotbesequentiallylocatedinphysicalblocks,Weuseanumberoftablesstoringnodesimformation.Supposenowweuseafixednodesizeofvariable‐lengthn,ittakes[n/(k‐b)]nodestostorethisitem.(Herebisaconstant,signifyingthatbwordsofeach