Our current test is running with one block per minute on average, but our consensus algorithm and network structure will let us safely push it faster. We will experiment more to see how fast the blocks can run and how many tx per block. In long term, with the DApps, there will be something like a sharded architecture, so such block speeds may be possible.
