Zuwan wannan guntu ya canza yanayin ci gaban guntu!
A ƙarshen 1970s, na'urori masu sarrafawa na 8-bit har yanzu sune fasahar ci gaba a lokacin, kuma matakan CMOS sun kasance cikin rashin ƙarfi a cikin filin semiconductor. Injiniyoyi a AT&T Bell Labs sun ɗauki mataki mai ƙarfin gwiwa zuwa gaba, tare da haɗa matakai na masana'antu na 3.5-micron CMOS tare da sabbin kayan aikin sarrafawa na 32-bit a ƙoƙarin ƙetare masu fafatawa a aikin guntu, wanda ya zarce IBM da Intel.
Kodayake ƙirƙirar da suka kirkira, microprocessor Bellmac-32, sun kasa cimma nasarar kasuwanci na samfuran farko kamar Intel 4004 (wanda aka saki a 1971), tasirinsa yana da girma. A yau, kwakwalwan kwamfuta a kusan dukkanin wayoyi, kwamfyutocin tafi-da-gidanka, da allunan sun dogara da ƙa'idodin ƙa'idodin ƙarfe-oxide semiconductor (CMOS) wanda Bellmac-32 ya fara aiki.
1980s suna gabatowa, kuma AT&T yana ƙoƙarin canza kansa. Shekaru da dama, katafaren kamfanin sadarwa da ake yi wa lakabi da "Uwar Bell" ya mamaye harkokin sadarwar murya a Amurka, kuma reshensa na Western Electric ya samar da kusan dukkanin wayoyin da aka saba amfani da su a gidaje da ofisoshin Amurka. Gwamnatin tarayya ta Amurka ta bukaci wargajewar kasuwancin AT&T bisa dalilai na kin amincewa, amma AT&T ta ga wata dama ta shiga filin na'ura mai kwakwalwa.
Tare da kamfanonin kwamfuta da aka riga an kafa su a kasuwa, AT&T ya sami wahalar kamawa; dabararsa ita ce ta yi tsalle, kuma Bellmac-32 ita ce jirgin ruwanta.
An karrama dangin guntu na Bellmac-32 tare da lambar yabo ta IEEE Milestone. Za a gudanar da bukukuwan kaddamarwa a wannan shekara a harabar Nokia Bell Labs da ke Murray Hill, New Jersey, da kuma gidan tarihin tarihin kwamfuta a Mountain View, California.
CHIP na musamman
Maimakon bin ka'idojin masana'antu na kwakwalwan kwamfuta 8-bit, shugabannin AT&T sun kalubalanci injiniyoyin Bell Labs don haɓaka samfurin juyin juya hali: microprocessor na farko na kasuwanci wanda zai iya canja wurin rago 32 na bayanai a cikin zagayen agogo guda. Wannan yana buƙatar ba kawai sabon guntu ba amma har da sabon gine-gine-wanda zai iya sarrafa sauyawar sadarwa da kuma zama ƙashin bayan tsarin kwamfuta na gaba.
"Ba kawai muna gina guntu mai sauri ba," in ji Michael Condry, wanda ke jagorantar rukunin gine-gine a ginin Bell Labs' Holmdel, New Jersey. "Muna ƙoƙarin tsara guntu wanda zai iya tallafawa duka murya da ƙididdigewa."
A lokacin, ana ganin fasahar CMOS a matsayin mai ban sha'awa amma mai haɗari ga ƙirar NMOS da PMOS. NMOS kwakwalwan kwamfuta sun dogara gaba ɗaya akan nau'in transistor na nau'in N, waɗanda suke da sauri amma masu fama da yunwa, yayin da kwakwalwan kwamfuta na PMOS sun dogara da motsi na ramukan da aka caje, wanda ya kasance a hankali. CMOS ya yi amfani da ƙirar ƙira wanda ya ƙaru da sauri yayin adana ƙarfi. Fa'idodin CMOS sun kasance masu tursasawa cewa masana'antar nan da nan ta gane cewa ko da tana buƙatar ninki biyu na transistor (NMOS da PMOS na kowace ƙofar), yana da daraja.
Tare da saurin haɓaka fasahar semiconductor wanda Dokar Moore ta siffanta, farashin ninka yawan adadin transistor ya zama mai iya sarrafawa kuma a ƙarshe ba ya yin komai. Koyaya, lokacin da Bell Labs ya shiga wannan babban caca mai haɗari, manyan fasahar masana'antar CMOS ba ta da tabbas kuma farashin ya yi yawa.
Wannan bai tsorata Bell Labs ba. Kamfanin ya zana kan ƙwararrun cibiyoyin karatunsa a Holmdel, Murray Hill, da Naperville, Illinois, kuma ya haɗu da "ƙungiyar mafarki" na injiniyoyin semiconductor. Ƙungiyar ta haɗa da Condrey, Steve Conn, tauraro mai tasowa a ƙirar guntu, Victor Huang, wani mai ƙirar microprocessor, da ma'aikata da dama daga AT&T Bell Labs. Sun fara ƙware sabon tsarin CMOS a cikin 1978 kuma sun gina microprocessor 32-bit daga karce.
Fara da zane-zanen gine-gine
Condrey tsohon IEEE Fellow ne kuma daga baya ya yi aiki a matsayin Babban Jami'in Fasaha na Intel. Tawagar gine-ginen da ya jagoranta ta himmatu wajen gina tsarin da ya goyi bayan tsarin aiki na Unix da harshen C. A lokacin, duka Unix da C sun kasance a cikin ƙuruciyarsu, amma an ƙaddara su mamaye. Domin keta iyakar ƙwaƙwalwar ajiyar kilobytes (KB) mai mahimmanci a lokacin, sun gabatar da tsarin koyarwa mai rikitarwa wanda ke buƙatar ƙananan matakan aiwatarwa kuma zai iya kammala ayyuka a cikin zagaye ɗaya.
Injiniyoyi kuma sun ƙirƙira kwakwalwan kwamfuta waɗanda ke goyan bayan bas ɗin layi ɗaya na VersaModule Eurocard (VME), wanda ke ba da damar rarraba kwamfuta da ba da damar nodes da yawa don sarrafa bayanai a layi daya. Har ila yau, kwakwalwan kwamfuta masu jituwa na VME suna ba su damar amfani da su don sarrafa ainihin lokaci.
Ƙungiyar ta rubuta nau'in nata na Unix kuma ta ba ta damar ainihin lokaci don tabbatar da dacewa tare da aikin sarrafa masana'antu da makamantansu. Injiniyoyin Bell Labs suma sun ƙirƙiro dabaru na domino, wanda ya ƙara saurin sarrafawa ta hanyar rage jinkiri a cikin rikitattun ƙofofin dabaru.
An haɓaka ƙarin gwaje-gwaje da fasahohin tabbatarwa kuma an gabatar da su tare da ƙirar Bellmac-32, ƙayyadaddun tabbaci na guntu da yawa da aikin gwaji wanda Jen-Hsun Huang ke jagoranta wanda ya sami lahani na sifili ko kusa-sifili a cikin hadadden ƙirar guntu. Wannan shine farkon gwajin da'irar hadedde (VLSI) a duniya. Injiniyoyin Bell Labs sun ɓullo da wani tsari mai tsari, akai-akai suna bincika aikin abokan aikinsu, kuma daga ƙarshe sun sami haɗin gwiwa maras kyau a tsakanin iyalai da yawa, wanda ya ƙare a cikin cikakken tsarin microcomputer.
Sashe na gaba ya zo mafi ƙalubale: ainihin masana'antar guntu.
"A lokacin, zane-zane, gwaji, da fasaha na masana'antu masu yawan gaske ba su da yawa," in ji Kang, wanda daga baya ya zama shugaban Cibiyar Kimiyya da Fasaha ta Koriya ta Koriya (KAIST) kuma abokin IEEE. Ya lura cewa rashin kayan aikin CAD don tabbatar da cikakken guntu ya tilasta ƙungiyar buga manyan zane-zane na Calcomp. Waɗannan ƙididdiga suna nuna yadda ya kamata a shirya transistor, wayoyi, da haɗin haɗin kai a cikin guntu don ba da fitarwar da ake so. Tawagar ta tattara su a kasa da tef, inda suka samar da wani katon murabba'in da ya zana fiye da mita 6 a gefe. Kang da abokan aikinsa sun zana kowane da'irar hannu da hannu cikin fensir kala-kala, suna neman karyewar haɗin kai da haɗaɗɗen haɗin gwiwa ko kuma waɗanda ba su dace ba.
Da zarar an kammala ƙirar jiki, ƙungiyar ta fuskanci wani ƙalubale: masana'anta. An samar da kwakwalwan kwamfuta a masana'antar Western Electric a Allentown, Pennsylvania, amma Kang ya tuna cewa yawan amfanin ƙasa (kashi na kwakwalwan kwamfuta a kan wafer wanda ya cika aiki da ƙa'idodi masu inganci) ya yi ƙasa sosai.
Don magance wannan, Kang da abokan aikinsa suna tuƙi zuwa shuka daga New Jersey kowace rana, suna naɗe hannayensu kuma sun yi duk abin da ya dace, gami da sharaɗɗen benaye da daidaita kayan gwaji, don gina ƙawance tare da gamsar da kowa cewa samfuran da suka taɓa ƙoƙarin samarwa za a iya yin su a can.
"Tsarin gina ƙungiyar ya tafi cikin kwanciyar hankali," in ji Kang. "Bayan 'yan watanni, Western Electric ya sami damar samar da kwakwalwan kwamfuta masu inganci a adadi wanda ya zarce bukatar."
An fito da sigar farko ta Bellmac-32 a cikin 1980, amma ta kasa cika tsammanin. Mitar aikin sa shine 2 MHz kawai, ba 4 MHz ba. Injiniyoyin sun gano cewa na’urar gwajin zamani ta Takeda Riken da suke amfani da ita a lokacin ba ta da kura-kurai, inda tasirin layin sadarwa tsakanin na’urar binciken da shugaban gwajin ya haifar da rashin daidaito. Sun yi aiki tare da ƙungiyar Takeda Riken don haɓaka teburin gyara don gyara kurakuran auna.
Chips na ƙarni na biyu na Bellmac yana da saurin agogon da ya wuce 6.2 MHz, wani lokacin har zuwa 9 MHz. An yi la'akari da wannan cikin sauri a lokacin. The 16-bit Intel 8088 processor da IBM ya saki a farkon PC a 1981 yana da gudun agogon kawai 4.77 MHz.
Me yasa Bellmac-32 bai yi ba't zama na al'ada
Duk da alkawarinsa, fasahar Bellmac-32 ba ta sami karɓuwa ta kasuwanci ba. A cewar Condrey, AT&T ya fara duban NCR mai kera kayan aiki a ƙarshen 1980s kuma daga baya ya juya zuwa saye, wanda ke nufin kamfanin ya zaɓi tallafawa layin samfuran guntu daban-daban. A lokacin, tasirin Bellmac-32 ya fara girma.
"Kafin Bellmac-32, NMOS ya mamaye kasuwa," in ji Condry. "Amma CMOS ya canza wuri mai faɗi saboda ya tabbatar da cewa ya zama hanya mafi inganci don aiwatar da shi a cikin fab."
Bayan lokaci, wannan fahimtar ta sake fasalin masana'antar semiconductor. CMOS zai zama ginshiƙi na microprocessors na zamani, yana ƙarfafa juyi na dijital a cikin na'urori kamar kwamfutocin tebur da wayoyi.
Gwajin ƙarfin hali na Bell Labs-ta yin amfani da tsarin masana'anta da ba a gwada shi ba da faɗin tsararrun gine-ginen guntu-ya kasance ci gaba a tarihin fasaha.
Kamar yadda Farfesa Kang ya ce: "Mu ne kan gaba a abin da zai yiwu. Ba wai kawai muna bin hanyar da ake da ita ba ne, muna ta sabon salo." Farfesa Huang, wanda daga baya ya zama mataimakin darektan Cibiyar Microelectronics ta Singapore kuma shi ma dan IEEE ne, ya kara da cewa: "Wannan ya hada da ba kawai guntu gine-gine da ƙira ba, har ma da babban tabbaci na guntu - ta yin amfani da CAD amma ba tare da kayan aikin kwaikwayo na dijital na yau ba ko ma da allunan burodi (daidaitaccen hanyar duba tsarin da'ira na tsarin lantarki ta amfani da kwakwalwan kwamfuta kafin a haɗa sassan da'irar tare).
Condry, Kang da Huang sun waiwayi wannan lokacin cikin nishadi kuma suna nuna sha'awa ga fasaha da sadaukarwar ma'aikatan AT&T da yawa waɗanda ƙoƙarinsu ya sa dangin Bellmac-32 ya yiwu.
Lokacin aikawa: Mayu-19-2025