South Korea’s fertility rate slumped to 0.78 last year, from 0.81 in 2021. The slide has worsened in recent months, falling to 0.70 in the April-to-June quarter. Since 2013, the country of 52 million has reported the lowest fertility rate among wealthy members of the OECD—where the average fertility rate stands at 1.58. No other OECD member has a fertility rate below 1. South Korea’s population began declining in 2020, with the number of deaths overtaking total births. Its military conscripts are expected to shrink by nearly half over the next two decades. The military has started to deploy more unmanned combat aircraft and increase the number of women serving. The country’s total student enrollment has shrunk for 18 years straight.
- Date Posted:
- August 30, 2023
Models of artificial intelligence (AI) that have billions of parameters can achieve high accuracy across a range of tasks but they exacerbate the poor energy efficiency of conventional general-purpose processors, such as graphics processing units or central processing units. Analog in-memory computing (analog-AI) can provide better energy efficiency by performing matrix–vector multiplications in parallel on ‘memory tiles’. However, analog-AI has yet to demonstrate software-equivalent (SWeq) accuracy on models that require many such tiles and efficient communication of neural-network activations between the tiles. We demonstrate fully end-to-end SWeq accuracy for a small keyword-spotting network and near-SWeq accuracy on the much larger MLPerf8 recurrent neural-network transducer (RNNT.)