\n<\/p>\n
Cornell \u00dcniversitesi ara\u015ft\u0131rmac\u0131lar\u0131, hem ultra h\u0131zl\u0131 veri sinyallerini hem de kablosuz ileti\u015fim sinyallerini ayn\u0131 anda i\u015fleyebilen, d\u00fc\u015f\u00fck enerji t\u00fcketimli bir mikro\u00e7ip<\/a> geli\u015ftirdi. \u201cmikrodalga<\/a> beyin<\/a>\u201d ad\u0131 verilen bu teknoloji<\/a>, mikrodalga fizi\u011fini kullanarak \u00e7al\u0131\u015fan ilk ger\u00e7ek mikrodalga sinir a\u011f\u0131 olma \u00f6zelli\u011fini ta\u015f\u0131yor.<\/p>\n Nature Electronics dergisinde yay\u0131mlanan \u00e7al\u0131\u015fmaya g\u00f6re, silikon \u00fczerine entegre edilen \u00e7ip, radyo sinyali \u00e7\u00f6zme, radar hedef takibi ve dijital veri i\u015fleme gibi g\u00f6revleri saniyede onlarca gigahertz h\u0131z\u0131nda ger\u00e7ekle\u015ftirebiliyor. \u00dcstelik bunu, 200 miliwatt\u2019tan daha az enerji t\u00fcketerek ba\u015far\u0131yor.<\/p>\n ABD Savunma \u0130leri Ara\u015ft\u0131rma Projeleri Ajans\u0131 (DARPA) ve Ulusal Bilim Vakf\u0131\u2019n\u0131n destekledi\u011fi Cornell NanoScale Science and Technology Facility b\u00fcnyesinde y\u00fcr\u00fct\u00fclen projeye g\u00f6re, geleneksel dijital sinir a\u011flar\u0131n\u0131n aksine bu sistem, saat tabanl\u0131 ad\u0131m ad\u0131m i\u015flemler yerine mikrodalga band\u0131nda analog ve do\u011frusal olmayan davran\u0131\u015flarla \u00e7al\u0131\u015f\u0131yor. Bu sayede veri i\u015fleme ad\u0131mlar\u0131n\u0131 k\u0131saltarak hem d\u00fc\u015f\u00fck g\u00fc\u00e7 t\u00fcketimi hem de y\u00fcksek h\u0131z sa\u011fl\u0131yor.<\/p>\n \u00c7ip, basit mant\u0131k i\u015flemlerinden karma\u015f\u0131k veri s\u0131n\u0131fland\u0131rmalar\u0131na kadar bir\u00e7ok g\u00f6revi yerine getirebiliyor. Kablosuz sinyal t\u00fcrlerini s\u0131n\u0131fland\u0131rmada y\u00fczde 88\u2019in \u00fczerinde do\u011fruluk oran\u0131na ula\u015farak dijital sistemlerle k\u0131yaslanabilir performans g\u00f6sterdi. Ayr\u0131ca donan\u0131m g\u00fcvenli\u011fi, kablosuz ileti\u015fimde anomali tespiti ve d\u00fc\u015f\u00fck g\u00fc\u00e7 t\u00fcketimi sa\u011flan\u0131rsa ak\u0131ll\u0131 saat ya da telefon gibi u\u00e7 cihazlarda yerel yapay zeka modelleri \u00e7al\u0131\u015ft\u0131rma gibi uygulamalara da uygun.<\/p>\n Ara\u015ft\u0131rmac\u0131lar, teknolojinin \u00f6l\u00e7eklenebilir oldu\u011funa ve mevcut mikrodalga ile dijital i\u015fleme platformlar\u0131na entegre edilebilece\u011fine inan\u0131yor.<\/p>\n<\/div>\nANALOG DAVRANI\u015eLARLA \u00c7ALI\u015eIYOR<\/strong><\/h3>\n
Y\u00dcZDE 88’\u0130N \u00dcZER\u0130NDE DO\u011eRULUK ORANI<\/strong><\/h3>\n