A team of engineers from the Rochester Institute of Technology and the University of Minnesota has announced a significant advancement in the field of biological computing. They have conceptualized a "microfluidic integrated circuit" capable of executing artificial neural network computations on data encoded within DNA strands.
Amlan Ganguly, one of the paper's authors, highlighted the challenges of the current data storage paradigm, stating, "In our era of exponentially growing data, finding sustainable storage solutions is imperative.
The energy and space demands of traditional data centres are not viable long-term solutions." Ganguly's insights were shared in a press release accompanying the study published in PLOS One.
Ganguly further emphasized the untapped potential of DNA as a medium for data storage. DNA's capacity for information density is theoretically far superior to conventional storage devices, potentially offering a three to six orders of magnitude improvement in compactness. Moreover, DNA's inherent stability and longevity could lead to more reliable storage methods when properly utilized.
The parallels between DNA and traditional silicon-based computing are striking. DNA has the capability to sequence and synthesize data, akin to the reading and writing functions of silicon chips. The research team, including Ganguly, is exploring ways to manipulate materials at the molecular scale.
Their goal is to reduce the reliance on electronic storage components and transition towards biological storage and processing systems that are more compatible with living organisms.
This innovative approach holds promise for applications in forensics and biomedicine, and it could herald a new era of durable and efficient storage systems that enhance data retrieval processes.
Ganguly explained the operational concept: "We envision using DNA molecule concentrations to represent numerical values, with computational operations being carried out through the manipulation of these molecules. This method bridges the gap between data storage and computation, utilizing DNA as the medium for processing."
The envisioned microfluidic system aims to deliver the output of an artificial neural network. It is designed to operate as both a sensor and a computational processor, featuring minuscule channels equipped with nanotechnology sensors. These sensors are tasked with isolating, identifying, and capturing molecules within fluid samples.
.png&w=3840&q=75)
