EDACafe Industry Predictions for 2023 – Arteris

By Frank Schirrmeister, Vice President, Arteris

Frank Schirrmeister

What Will Be The Driving Trends For 2023?

Across the board, we will see an increasing demand for higher performance and power efficiency. As the demand for computing power continues to grow from data centers to network and device edges, there is a need for more efficient and powerful semiconductor devices. This trend will further drive the development of new materials, designs, and manufacturing processes to deliver better performance and lower power consumption. Semiconductor IP will face increased scrutiny to meet stringent low-power demands, further increasing the pressure to optimize the power contribution of NoCs by reducing wires, registers, and NoC base components like switches.

We will see a further bifurcation between more complex and integrated device architectures at the high-end computing domain and the lower complexity for sensor endpoints that require much more power-aware devices and advanced mixed-signal developments. More integrated and sophisticated device architectures may require integrating multiple components, such as sensors, processors, and memory, onto a single chip. From a network-on-chip (NoC) perspective, at the high end, we will see hierarchies involving 10s of interconnect areas with a mix of cache-coherent and non-cache-coherent NoCs, with the number of initiators and targets per NoC further growing. But even for smaller designs at the sensor and device edges, NoCs will increasingly become the standard over custom manual interconnects and bus structures.

As device densities continue to increase, interest in three-dimensional (3D) chip stacking will only accelerate. The physical layer has made considerable progress with the emergence of UCIe and today’s predominant streaming protocol usage. In 2023, the attention will shift to the protocols, defining application-specific control and transport layers. We will experience some fragmentation between AXI, CHI, CXL, and others.

Physically aware IP design that is already necessary for analog mixed-signal IP, will edge closer to a must-have for NoC development. Definition of the requirements of the NoC through the choice of an NoC architecture, selection of the routing algorithm, design of the physical layout of the NoC, and simulation and verification can no longer be done in isolation but need to be tightly connected and co-optimized

The automotive area will continue to be dominated by requirements around electrification, autonomy, connectivity, personalization, and sustainability. As self-driving technology advances, designers are also considering how to design autonomous vehicles, focusing on creating comfortable and convenient spaces for passengers with seamless integration of connected devices, providing an improved experience for drivers and passengers. And with growing consumer interest in vehicles that reflect their style and preferences, designers are focusing on offering more customization options, such as personalized interiors and exteriors. There is also a trend towards designing more sustainable vehicles, both in terms of their environmental impact and lifecycle.

For enterprise computing in data centers, energy efficiency, security & compliance, cloud & edge computing, and sustainability will drive a diverse set of requirements. Data centers consume large amounts of energy, and there is a trend toward designing and operating data centers that are more energy efficient. Expect updates in advanced cooling systems, efficient servers and networking equipment, and renewable energy sources. With silicon at its core, NoCs and optimized processor architectures offer much low-power optimization potential. When designing and building data centers, developers must also consider security and compliance issues, ensure that data centers are secure against physical and cyber threats, and meet regulatory requirements for data protection and privacy. Amongst others, the growth of the internet of things (IoT) is driving the development of edge computing, which involves bringing computing power closer to the source of data generation. Data center developers must consider how to design and build data centers that can support edge computing applications and must understand the balance of computing between devices, edge nodes, and the cloud data center.

5G, AI/ML, wearables, and virtual & augmented reality drive the requirements for consumer devices. The rollout of 5G networks continues in the coming years, and the integration of artificial intelligence (AI) and machine learning (ML) into consumer devices will continue to grow. This trend includes the development of smart home devices, such as voice assistants and smart thermostats, and integrating AI/ML into other consumer devices, such as smartphones and laptops. The market for wearable devices, such as smartwatches and fitness trackers, is expected to grow in the coming years. Developers will need to consider how to design wearable devices that are attractive, functional, and comfortable.

AI will have a significant impact on our future. Advances in artificial intelligence and machine learning, from data centers through edges to endpoints, will further drive the development of specialized semiconductor devices that can support these technologies. Its unique requirements in scale and replicability will cause specific variations of NoCs, often leading to the co-optimization of computing and interconnect. AI/ML will also further improve design productivity and bears the potential to overhaul classic design methodologies and even configurability of IP significantly.

2023 promises to be exciting year, filled with advanced technologies, new methodologies, creative thinking, and ever-expanding consumer demands for the next big thing.

About Author:

Frank Schirrmeister is vice president of solutions and business development at Arteris. He leads activities in the automotive, data center, 5G/6G communications, mobile, aerospace and data center industry verticals and the technology horizontals artificial intelligence, machine learning and safety. Before Arteris, Frank held various senior leadership positions at Cadence Design Systems, Synopsys and Imperas, focusing on product marketing and management, solutions, strategic ecosystem partner initiatives and customer engagement.

Category: Predictions

This entry was posted on Tuesday, January 24th, 2023 at 12:58 pm. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.