What something costs can be evaluated only in relation to its value, and that applies to edge computing--including multi-access edge computing--as well. For smaller and mid-sized organizations, the cost of MEC might be higher than relying on remote computing at a hyperscale data center, but typically will be far less costly than an on-premises, owned solution, according to Ericsson.
As always, assumptions matter. Ericsson used figures for a connectivity provider in Sweden. Aggregation sites have power capacity installed up to 10kW, hosting an average of eight server units, each with four cores.
With 8,000 access sites and one aggregation site per 10 access sites, there is a virtual processor (vCPUs) capacity of 25,600 (800 sites x 8 servers per site x 4 cores per server) for enterprise applications at CSP-owned edge sites.
Servers are depreciated over three years while investments in power and cooling systems are depreciated over 14 years.
Up to 1.6MW (800 sites x 8 servers per site x 250W per server) of incremental power is required. With an assumed power efficiency factor of 2, 3.2MW power is needed on average to power all the aggregation sites.
The cost of compute resources at each aggregation site is estimated at USD 20,000. This cost is very similar to USD per critical watt for building a large-scale data center, according to Ericsson.
Opex is the sum of electricity cost as well as operations and maintenance. Ericsson used a range of USD 0.10–0.15/kWh for power costs.
All that matters only in relation to the business value of applications that either require or significantly benefit from MEC. That would include ultra-low latency use cases, but also instances where privacy concerns are high or the cost of WAN bandwidth is significant.
No comments:
Post a Comment