Increasing demand for computing power and growing energy efficiency requirements have greatly influenced the development of data center technologies in recent years. AMD and 451 Research, a leading technology analysis and market research firm, have played a critical role in shaping and analyzing these developments. Of interest here are advances AMD has made in the data center space and 451 Research’s findings on the importance of these developments to the market.
For example, a recent report published by AMD in collaboration with 451 Research indicates that processor core optimization is one of the most effective ways to contain data center carbon footprints and energy costs. Organizations worldwide are increasingly recognizing the importance of monitoring and controlling their environmental impact. The study presents various approaches on how organizations can sustainably increase the efficiency of their IT infrastructures.
This is because companies around the world are more concerned than ever about taking responsibility for managing and reducing their environmental impact. Otherwise, they could be sanctioned by investors, governments could take regulatory action, and customers could seek out greener, more efficient providers. Although more companies are setting climate targets, it remains an open question whether their good intentions are sufficient to generate the ambition and measurable results needed to achieve much-needed significant greenhouse gas emission reductions. Bringing about improvement requires companies to optimize the efficiency of their IT infrastructure.
The July 2022 User Survey highlights the growing importance of this realization, with approximately half of IT decision makers surveyed indicating that IT activities account for either the majority (25%) or all (19%) of their organization’s environmental impact. Our study shows that many companies already have firm reduction targets for environmental impact, and consider energy efficiency and emissions reduction to be important factors influencing their technology purchasing decisions. Many also express a willingness to pay a higher price for greener products and expect their suppliers to provide appropriate tools to monitor environmental impacts.
But the benefits are not limited to reassuring investors and demonstrating good corporate practices. In the specific context of IT, efficiencies achieved through environmental measures translate directly into significant cost savings. With rising energy prices worldwide, the importance of total cost savings becomes even more apparent. The study shows that selecting an appropriate processor that requires fewer physical servers can help companies reduce overall energy consumption and increase operational efficiency by reducing data center space and power requirements.
For example, AMD has set a goal to increase the energy efficiency of AI training and high-performance computing applications based on accelerated compute nodes by 30 times by 2025. If this goal is met, billions of kilowatt-hours of electricity can be saved by 2025 by reducing the energy consumption of these systems for a single computation by 97% over five years. In the survey, modernizing core IT infrastructure (indicated by 41% of respondents) and using public cloud services (36%) are the most commonly pursued strategies to meet environmental goals.
Work productivity and collaboration tools (36%) and customer interaction and commerce tools (28%) had the next greatest impact. For those more advanced in their digital transformation processes, data platforms and data science tools (35% compared to the survey average of 27%) and AI and machine learning tools (34% compared to 27%) were seen as the most promising contributors to their environmental efforts.
For example, research and data modeling by 451 Research has shown that organizations that move their IT to the cloud are likely to reduce their energy consumption, leaving a smaller carbon footprint. However, organizations still using on-premises data centers need to plan the transition carefully and ensure that operations management covers both on-premises and cloud-based resources in the hybrid IT environment to minimize cloud operating costs.
For efficiency, there are three main levers that determine the potential savings:
- Equipment in a cloud data center is typically updated very regularly with the latest technology, maximizing efficiency both in terms of power per watt and power delivered to customers.
- Cloud infrastructures are shared and optimized for maximum utilization – and utilization is one of the biggest factors in total cost of ownership.
- The latest cloud data centers are specifically designed for efficiency in three areas: efficient infrastructure (power and cooling); efficient servers, storage and networking; and use of clean power sources, renewable energy technologies and renewable energy development projects.
AMD has made significant progress in recent years in developing powerful and efficient processors for data centers. This includes the successful introduction of EPYC processors and the continued improvement of the Radeon Instinct GPU line. AMD’s EPYC processors provide data centers with the very high scalability and flexibility needed to handle Big Data analytics and AI workloads. With a high number of cores and memory bandwidth, these processors enable data centers to customize performance and capacity based on their needs.
AMD’s Zen architecture enables greater energy efficiency, reducing operating costs for data centers. By leveraging 7nm technology and advanced cooling solutions, AMD has further optimized performance per watt, resulting in better overall efficiency. AMD has also raised the bar for the industry with its Secure Encrypted Virtualization and Secure Memory Encryption technologies. These technologies provide comprehensive protection for data and applications in data centers, ensuring that confidential information is protected from unauthorized access.
Those interested in reading the full report can find it at 451 Research (source)
30 Antworten
Kommentar
Lade neue Kommentare
Mitglied
1
Veteran
Veteran
Urgestein
Urgestein
Urgestein
Mitglied
Veteran
Urgestein
Veteran
Urgestein
Urgestein
Urgestein
Urgestein
Alle Kommentare lesen unter igor´sLAB Community →