新材料與P溝道和N溝道晶體管的混合將有望使Intel公司制造出僅有目前芯片功耗10%的處理器。 Intel最近透露了一種P溝道晶體管進展的細節,這種晶體管制作于硅基片上,采用混合半導體材料,也稱作III-V材料,因為它由元素周期表上第三列和第五列的元素組成,而硅位于第四列。該研究的成果造就了迄今為止性能最優的P溝道晶體管。 具Intel透露,這兩種結果結合在一起可以形成CMOS邏輯電路的基礎。CMOS邏輯電路同時采用N溝道和P溝道。這種技術有潛力用于制造將來的微處理 器,其工作溫度會低很多。Intel這個全球最大的芯片廠商聲稱,新晶體管的工作電壓可降低50%,其功耗只有當今晶體管的10%。 如果這種技術以后能夠實現批量生產,那將產生兩種后果:一個事處理器尺寸非常小就能實現目前處理器的性能,或者同樣尺寸的處理器會非常復雜,功能會強大得多。顯然,該技術不僅會顯著降低微處理器的功耗,還會惠及圖形處理等高集成設備。 Intel目前正在繼續研究這項技術,據稱尚需數年方可實現。 ---------------------- 原文: The mix of new materials as well as P-channel and N-channel transistors can potentially enable Intel Corp. to build microprocessors that consume just 10% of today’s chips in future. Intel recently disclosed advancement details on a P-channel transistor, built on a silicon substrate, that makes use of compound semiconductors, also known as III-V materials because they are made of elements that straddle silicon in the periodic table, silicon being in column IV. This research resulted in the highest performing P-channel transistors reported to date. A year earlier, Intel described III-V N-channel transistors, also built on a silicon substrate. According to Intel, when combined, these two results could form. the building blocks for CMOS logic circuits, which use both N-channel and P-channel transistors. Potentially suitable for future microprocessors, they run far cooler: the world’s largest chipmaker claims that t about 50% the voltage, consuming only 10% the power of today's transistors. Potentially, if the innovation proves to be viable for mass production of microprocessors, this enables either very small and power efficient chips which are powerful by today’s standards or chips that are dramatically more complex compared to contemporary, but consume the same amount of power. Obviously, the new materials can also enable dramatic improvements in energy efficiency not only for microprocessors, but also for graphics processors, core-logic sets and other highly-integrated devices. Currently Intel is already researching the aforementioned materials and it will take years before the innovation could materialize. |