comparing timescaledb vs InfluxDB vs redis

I’m looking into building some low latency real time application (<100ms) based on time series data. I’m comparing timescaledb vs InfluxDB vs Redis. I’m comparing them using

From a lot of blog posts, I found that TimescaleDB had better performance than InfluxDB, but I found InfluxDB having in memory capabilities (while TimescaleDB does not). Then I wonder if by utilizing such in memory capabilities, would influx db perform better on timescaledb during live production.

Some industry expert mentioned Redis for such application. But it seems it is not really a database.

In summary my question is,

1, is TimescaleDB really not having in memory capabilities?

2, if I trigger in memory capabilities in InfluxDB, would the production level performance be probably better than TimescaleDB?

3, Is Redis a timeseries database, offering similar capabilities of TimescaleDB or InfluxDB?