This article presents a neuromorphic circuit for memristive tonotopic mapping using volatile resistive switching memory (RRAM) devices. The goal is to replicate the auditory system's ability to process temporal signals in a spatiotemporal manner, enabling efficient and scalable neuromorphic computing. The research demonstrates the key features of signal processing in the cochlea, such as logarithmic integration and tonotopic mapping of signals, and shows that the tonotopic classification is suitable for speech recognition. The study is inspired by the neurobiological processes in the human auditory system, where sound is processed internally through an internal representation of physical features rather than relying on the spatial arrangement of sensory neurons. The cochlea achieves this by mapping different frequency components along logarithmically spaced positions of the cochlear channel. The challenge is to emulate this spatiotemporal signal processing using simple and scalable hardware. The research uses volatile RRAM devices with a 1T1R structure, which can switch between high and low resistive states based on unipolar voltages. The devices exhibit stochastic switching behavior, with the switching probability depending on the voltage amplitude and spike frequency. The study shows that the switching probability can be modeled and used to implement a tonotopic mapping of audio frequencies. The RRAM circuit for frequency sensing is designed to provide tonotopic sensing of auditory signals. The circuit uses multiple RRAM devices with separate top electrodes and a common bottom electrode to collect the summation current. The gate voltage is common for all devices, ensuring that the current is approximately the same for each device in the ON state. The spike trains applied to different top electrodes have the same frequency, while the voltage amplitude decreases from one electrode to the next. The results show that the number of devices in the ON state increases with the input frequency, demonstrating a logarithmic dependence of frequency sensitivity. This is consistent with the tonotopic mapping in the cochlea. The study also demonstrates the ability to classify different phonemes based on their frequency and pressure characteristics, showing the potential for speech recognition using the tonotopic map. The research highlights the advantages of memristive devices for neuromorphic computing, including high scalability, low power consumption, and the ability to directly compute within the memory. The study also discusses the importance of device-level computation for reducing the need for external circuitry and improving the efficiency and capabilities of neuromorphic systems. The results support the use of memristive devices for hardware processing of temporal signals with logarithmically spaced features, expanding the set of available neuromorphic primitives necessary to achieve the energy efficiency and high integration density promised by neuromorphic computing.This article presents a neuromorphic circuit for memristive tonotopic mapping using volatile resistive switching memory (RRAM) devices. The goal is to replicate the auditory system's ability to process temporal signals in a spatiotemporal manner, enabling efficient and scalable neuromorphic computing. The research demonstrates the key features of signal processing in the cochlea, such as logarithmic integration and tonotopic mapping of signals, and shows that the tonotopic classification is suitable for speech recognition. The study is inspired by the neurobiological processes in the human auditory system, where sound is processed internally through an internal representation of physical features rather than relying on the spatial arrangement of sensory neurons. The cochlea achieves this by mapping different frequency components along logarithmically spaced positions of the cochlear channel. The challenge is to emulate this spatiotemporal signal processing using simple and scalable hardware. The research uses volatile RRAM devices with a 1T1R structure, which can switch between high and low resistive states based on unipolar voltages. The devices exhibit stochastic switching behavior, with the switching probability depending on the voltage amplitude and spike frequency. The study shows that the switching probability can be modeled and used to implement a tonotopic mapping of audio frequencies. The RRAM circuit for frequency sensing is designed to provide tonotopic sensing of auditory signals. The circuit uses multiple RRAM devices with separate top electrodes and a common bottom electrode to collect the summation current. The gate voltage is common for all devices, ensuring that the current is approximately the same for each device in the ON state. The spike trains applied to different top electrodes have the same frequency, while the voltage amplitude decreases from one electrode to the next. The results show that the number of devices in the ON state increases with the input frequency, demonstrating a logarithmic dependence of frequency sensitivity. This is consistent with the tonotopic mapping in the cochlea. The study also demonstrates the ability to classify different phonemes based on their frequency and pressure characteristics, showing the potential for speech recognition using the tonotopic map. The research highlights the advantages of memristive devices for neuromorphic computing, including high scalability, low power consumption, and the ability to directly compute within the memory. The study also discusses the importance of device-level computation for reducing the need for external circuitry and improving the efficiency and capabilities of neuromorphic systems. The results support the use of memristive devices for hardware processing of temporal signals with logarithmically spaced features, expanding the set of available neuromorphic primitives necessary to achieve the energy efficiency and high integration density promised by neuromorphic computing.