Applications tend to cluster around locations where data density is highest. So data gravity would seem to work against edge computing and in favor of hyperscale computing. But the rise of multi-cloud is a countervailing force, as data stores might therefore be less concentrated, especially to support internet of things use cases.
In many ways, that is in line with the traditional drivers of compute density and network latency, the ability to substitute local computing for wide area network costs and performance issues (latency or bandwidth), as well as the ability to substitute WAN transport for compute cost.
In one sense, edge computing might help information technology professionals normalize data gravity, as edge processing shifts the value of the venue towards the edge.
No comments:
Post a Comment