Ampere Is Creating Cloud Native Processors

The current cloud computing market is large and growing rapidly. With the legacy PC and enterprise microprocessor offerings failing to meet the needs of the new cloud market, Ampere believes a new cloud-native computing architecture is needed. The Ampere team has taken a complete platform approach, and they’re doing so with Cadence tools.

Legacy architectures have struggled to deliver consistently high performance at the right power efficiency, scalability, and flexibility. Ampere, on the other hand, has a robust roadmap for its cloud-native processors. The team is developing not just the processor but the entire system, from the socket to the BIOS. Crucially, they’re working with developers and partners to create a software ecosystem that spans various operating systems, compilers, applications, and network and storage devices. 

Ampere’s cloud-native designs have very large design requirements, tens of billions of transistors, high performance, and area efficiency. Ampere uses the Arm ISA and IP and is collaborating closely with Cadence to build best-in-class custom CPUs for data centers. For implementation tools, they’re using Cadence’s Genus Synthesis Solution and Innovus Implementation System alongside Tempus Timing Solution and Joules RTL Power Solution to get the best results for multi-corner performance and power optimization.

The Ampere CPU team also uses the Cadence custom design tool suite for in-house IP development and characterization. They also use key IP from the Cadence IP design portfolio and rely on the Palladium Emulation Platform for performance validation, verification, and pre-silicon firmware development. Learn more about how Ampere is creating cloud-native processors with Cadence and Arm.

“Designed with Cadence” is a series of videos that showcases creative products and technologies that are accelerating industry innovation using Cadence tools and solutions. For more Designed with Cadence videos, check out the Cadence website and YouTube channel.