菇電腦來了？菌絲體做運算元件

隨著全球對運算能力的需求急劇攀升，科學家正尋找可提供高效但低能耗處理能力的新材料。近期《Times of India》報導指出，研究團隊正利用香菇（Lentinula edodes）菌絲體作為生物電子元件，試圖建構可模仿神經網路的生物計算系統，並將其視為未來低功耗人工智慧與數據處理的新路徑。

As global computing demands surge, scientists are searching for new materials capable of delivering energy-efficient processing without reliance on rare-earth minerals or complex semiconductor fabrication. A recent article in The Times

研究指出，在類神經運算（neuromorphic computing）與記憶阻元（memristor）元件的發展中，目前主要材料仍仰賴稀土金屬與能源密集型製程，而環境成本高昂。

of India reports that researchers are exploring shiitake fungi (Lentinula edodes) as bioelectronic components mimicking neural behavior, pointing toward sustainable, low-cost paths for artificial intelligence and data processing.

為此，科學家轉向生物系統，並發現菌絲體具備類似神經元的電信號傳輸與記憶行為。

In the realm of neuromorphic computing - designed to emulate the human brain’s neural architecture - traditional semiconductor-based memristors demand resource-intensive manufacturing. Researchers turned to unconventional biological

實驗中，香菇菌絲透過電極介面展現可重複的電阻變化，與神經突觸的“學習”機制相似。

systems and found that fungal mycelial networks display neuron-like electrical signaling and memory behavior. In one PLOS One study, shiitake-based memristors achieved up to 90 % accuracy at frequencies reaching 5.85 kHz.

其記憶效能可達約 90% 的準確度，頻率最高為 5.85 kHz。

Beyond their computing potential, fungi also offer remarkable resilience. Shiitake mycelia maintained memristive behavior after dehydration or exposure to radiation, making them viable for harsh-environment electronics and extraterrestrial applications where traditional materials falter.

此外，菌絲體的生物特性尤適用於極端環境。香菇菌絲在脫水後仍保有電訊號記憶功能，且對輻射及惡劣條件具較強韌性。報導指出，此類生物電子系統有望應用於太空電子裝置或惡劣場域，降低系統對稀有材料與高能耗製程的依賴。

Compared with conventional silicon devices, fungal electronics promise biodegradability, lower power consumption, and reduced environmental overhead. The research team cultivated conductive fungal networks on organic substrates such

菇類計算裝置與傳統半導體相比，具備生物可降解、低能耗、低環境負荷等優點。研究團隊在實驗中以有機培養基育菌──如法羅穀物與小麥胚芽──成功生成導電菌絲網絡，並用於模擬記憶與學習機制。團隊認為，當這類裝置優化並量產後，或可成為分散式、生態友善的智能裝置基底。

as farro seed and wheat germ, demonstrating the feasibility of growing computing components with minimal infrastructure. They argue that the transition from silicon to mycelium is not science fiction but a visible evolutionary path.

雖然“菌電腦”尚處於驗證階段，但其潛力不容忽視。研究者寫道：「從硅基半導體轉向菌絲體並非科幻，而是一種可見的技術演進。」他們表示，下一步將著眼於縮小元件尺寸、提升性能穩定性並驗證實際應用場景。隨著數據處理需求持續增長，這樣的新材料可能為運算技術開創新篇章。

Although fungal computing remains in early experimental stages, its implications are significant. The next milestones will focus on refining cultivation techniques, enhancing stability, miniaturizing components, and validating real-world applications. As data processing demands and energy concerns rise, these biologically inspired materials could reshape the future of intelligent machines.