Synopsys releases new white paper to promote cloud revolution

Translated from - Semiwiki

Since 2019, we have been moving our engineering and finance systems to the cloud. Our efforts have been successful, but not easy. We have faced many infrastructure challenges in this transition. This long list of challenges related to building a large amount of computing infrastructure is necessary to drive the cloud revolution. Synopsys recently published a white paper on these challenges that is very informative. If you work in cloud computing technology, you should understand the obstacles faced and how Synopsys is enabling the cloud computing revolution.

If you talk to people about moving to the cloud, you’ll get one of two responses:

My work is already in the cloud.

I think the cloud is not ready yet, but it is the future.

The Gartner survey chart shows that this trend is gradually maturing. Synopsys' white paper reveals many technical challenges related to the large-scale cloud construction around us. The author of this white paper is Scott Durrant, cloud segment manager of Synopsys. Before joining Synopsys, Scott worked at Intel for 24 years and also had experience in the enterprise software market such as McAfee. Scott understands the technical foundation of cloud computing.

He starts with some interesting trends about cloud migration, such as growth rates, widespread deployment of AI, and expansion of edge computing. IDC makes a prediction about the size of the global data layer in the next few years that will either excite you or scare you. Scott then spends considerable time looking at the six main functional areas of cloud computing - the underlying technologies, challenges, and market trends. You'll learn a lot. Here's a brief summary of each area.

Compute Server

In addition to computing power, communication bandwidth and energy efficiency. This article explores the various memory technologies and interfaces, as well as the requirements for standard and high-speed SerDes channels such as Compute Express Link (CXL). HBM2E is a promising technology.

Network Infrastructure

The primary focus here is on network switching. We discuss the move to 400G speeds and the various architectures to achieve that. The market share perspective here is different.

Storage System

We use artificial intelligence to optimize these systems. The impact of non-volatile memory technology, as well as cache coherence and related standards are examined and cannot be ignored. This market has a strong player, but not as strong as the network infrastructure market.

Visual Computing

This is a new category. It refers to the hardware and software required to do real-time video processing and analysis. For example, online collaboration, movie streaming, virtual reality, safety and assisted driving. These applications require some very high-end processing power.

Edge Infrastructure

The purpose of edge is to reduce latency. The amount of data collected by IoT devices is exploding. You can see estimates of the number of connected devices that will be deployed in the white paper. The challenge is to process all this data with latency appropriate for the application. This has basically led to the emergence of edge computing layers so that the right processing can be done at the right distance from the application. All of this challenges the way we used to think about data centers.

AI

Last, but not least, is the discussion on AI accelerators. These devices form the backbone of the entire infrastructure. Some applications require performance first and power is a secondary requirement, while others require low power first with the required performance level. The white paper discusses the technologies in this area and the relevant standards.

Throughout the discussion there are examples of where Synopsys IP fits into the various architectures presented, and there is no doubt that Synopsys has done a great job in this regard.