Exploiting thermal accumulation on sub-ns timescales for tuning potentiation of RRAM
In the latest issue of IEEE Electron Device Letters (Early Access, 2025), a research team by Daniel Schön, Faisal Munir, Pascal Stasner, Stefan Wiefels and Stephan Menzel published the article "Exploiting thermal accumulation on sub-ns timescales for tuning potentiation of RRAM".

Gradually switching resistive random access memory (RRAM) enables brain-like behaviour that can be exploited in novel neuromorphic computing architectures. In this letter, we demonstrate experimentally that it is possible to modulate the switching speed of a pulse train in the sub-nanosecond range by tuning the frequency. The temperature in the RRAM device caused by Joule heating is used as a second state variable to enable short-term plasticity. We show that this effect occurs in our device up to a delay time of about 250 ps and then disappears, indicating an ultra-short thermal time constant in the sub-nanosecond range. It is also demonstrated that the relative change in conductance that can be achieved is also dependent on the voltage applied.