from 11:05am to 11:40am (Paris Time)
Next-generation optimization accelerators: solving NP-hard problems using integrated coherent Ising machines or memristive crossbar arrays
Dr. Thomas Van Vaerenbergh - Photonics Research Engineer at Hewlett Packard Enterprise
Recent experimental results show how classical accelerators based on analog computing can outperform quantum annealing alternatives in benchmark task that require dense connection matrices.
In Hewlett Packard Labs, we have been studying two alternatives : integrated coherent Ising machines and mem-HNNS (based on memristive crossbar arrays). In this talk, we will discuss our recent progress for both platforms.
For this Ising machines, we will show how the choice of the nonlinearity in the activation function can affect performance and should hence not be overlooked in the accelerator design.
A proper choice of nonlinearity can in some cases weekean the requirement of more advanced control algorithms in the annealer.
For the mem-HNN platform, we have previously shown that our in-memory analog computing approach has at least 10,000x higher Solution/sec/Watt compared to existing digital hardware and quantum counterparts.
An important challenge for commercial viability is that different industrial workloads typically benefit from multiples optimization algorithms.
In this talk, we introduce additional combinatorial optimization algorithms, including quantum-inspired techniques, that can be deployed on our platform. This flexibility in algorithm choices in an important forward step to address the wide variety of enterprise-level use-cases such as airline scheduling, supply chain optimization, real-time bandwidth management, gene sequencing, etc.