The Intelligent Edge, also known as edge computing, is becoming increasingly essential for operational efficiencies and this change from a centralised model of computing is happening at speed.
Shifting the intelligence from the cloud to the edge enables operational and management decisions on real-time events to be made at the local level.
According to IDC, 73% of enterprises view the edge as a strategic area of investment and at least 90% of enterprise applications will embed AI by 2025.
Realising that key benefit involves various edge-related challenges. Initially, the development was constrained by the computing resources of the edge hardware. However, advances in silicon chip technology enabled the creation of small, low-power chipsets that could be embedded in edge hardware to deliver the required performance.
It is a key issue at the intelligent edge where a lot of memory as well as computational power is now needed. Memory resources need to match computing resources. Both are needed to perform perception tasks locally, with high accuracy, low latency, and energy efficiency.
In this expert analyst report, Bob Emmerson and Robin Duke-Woolley of Beecham Research explore:
- The new challenges for the Edge
- The benefits of intelligent gateways and edge servers
- How new workloads will leverage AI accelerators
- Why memory is not an infinite resource
- The essential role that memory plays in edge solutions
- The edge essentials that anticipate tomorrow’s data challenges