Cloud-native computing, a rapidly growing trend in the data center, is set to become the second most important computing trend by 2024. As performance cores have become larger and faster, there is now a shift towards using smaller and more efficient cores. This new approach allows for greater density and cost-effectiveness in server deployments.
The server industry has realized that while faster per core performance is beneficial for certain applications, there are many workloads that do not require such high-performance cores. These applications, such as line-of-business functions and smaller websites, need to be online 24/7 but do not require massive amounts of compute power. Moving away from hypervisors with per-core licensing or per-socket license constraints, the industry is now focused on scaling up systems with smaller, more power-efficient cores that can be densely packed.
Lowering the performance per core by 25% while decreasing power consumption by 40% or more allows for the efficient servicing of these applications. The trend towards lower clock speed cores with maximum frequencies in the 2-3GHz range further contributes to power efficiency. This shift towards cloud-native computing is driven by the need for increased density and reduced power consumption, especially in data centers where space and cooling constraints are significant factors.
Companies are transitioning their applications from older Xeon servers to modern cloud-native cores, which offer approximately the same performance per core while allowing for 4-5 times the density per system. As new generations of CPUs are released, these density figures continue to increase. The AMD EPYC “Bergamo” CPU, for example, features up to 128 cores/ 256 threads, making it the densest publicly available x86 server CPU on the market.
Cloud-native computing is reshaping the data center landscape by enabling higher efficiency and cost-effectiveness. With a focus on smaller, more efficient cores, businesses can achieve greater density and reduced power consumption, ultimately leading to improved performance and resource utilization. As the demand for cloud-native computing continues to rise, it is clear that this trend is here to stay, revolutionizing the way we approach data center infrastructure.
Frequently Asked Questions (FAQ)
1. What is cloud-native computing?
Cloud-native computing refers to a trend in the data center industry that emphasizes the use of smaller and more efficient cores for server deployments. This approach allows for greater density and cost-effectiveness in managing workloads.
2. Why is cloud-native computing becoming important?
Cloud-native computing is set to become the second most important computing trend by 2024 due to its potential to increase density, reduce power consumption, and optimize resource utilization in data centers.
3. Why do some workloads not require high-performance cores?
Certain applications, such as line-of-business functions and smaller websites, do not require massive amounts of compute power. Therefore, using smaller and more power-efficient cores can efficiently service these workloads while saving energy and resources.
4. How does cloud-native computing contribute to power efficiency?
By lowering the performance per core by 25% and decreasing power consumption by 40% or more, cloud-native computing enables more efficient servicing of applications. The trend towards lower clock speed cores also contributes to power efficiency.
5. Why are companies transitioning to cloud-native cores?
Companies are transitioning from older Xeon servers to modern cloud-native cores because they offer similar performance per core while allowing for significantly higher density per system. This results in improved resource utilization and cost-effectiveness.
Key Terms:
– Cloud-native computing: A trend in the data center industry that focuses on using smaller and more efficient cores for server deployments.
– Core: A central processing unit (CPU) is divided into multiple cores, each capable of executing tasks independently.
– Hypervisor: A software layer that enables the management and virtualization of computer resources.
– Density: Refers to the number of cores or threads that can be packed into a system or data center.
– Power consumption: The amount of electrical power used by a device or system.
Related Links:
– Data Center Knowledge
– AMD Official Website
– Intel Official Website
The source of the article is from the blog agogs.sk