Published: 02 January 2024
Summary
This research offers insights into early-stage frameworks for high-efficiency compute. While many enabling technologies are in R&D, product leaders should track their development to incorporate them into product roadmaps for delivering increased business value to customers.
Included in Full Research
Overview
Key Findings
GenAI applications and other compute-heavy tasks are driving the demand for better processing performance. At the same time, optimization of power consumption and silicon area requires new alternative compute paradigms.
Modifications across the entire stack, including devices, circuits, software, modules and systems, are necessary to derive the best results from energy-efficient compute.
Recommendations
Product leaders developing plans for energy-efficient compute must:
Gain an early understanding of the different emerging technologies, their timelines, viability and manufacturability by partnering with vendors and research groups in academia.
Target quick wins for energy-efficient compute by prioritizing software and/or circuit design techniques that typically involve lower
Clients can log in to view the entire
document.