Synopsys Cloud
Cloud native EDA tools & pre-optimized hardware platforms
Request a Free Trial →
As we enter the era of pervasive intelligence, the demand for AI chips is mushrooming and the race to get those chips — and the systems they enable — to market is on. To meet the demand, semiconductor engineering teams are being asked to accelerate chip design cycles and reduce the risks that come with exponentially increasing complexity and scale.
Because traditional pre-silicon RTL simulation alone is too slow and computationally intensive to handle the scale of modern AI chip designs and software, most engineering teams have turned to hardware-assisted chip verification to speed up, expand, and accelerate hardware-software validation efforts. But not all companies have the time or resources to establish and maintain on-premises chip design emulation capabilities in their own data center, if they even maintain their own data center.
That’s why we introduced ZeBu® Cloud, a secure (ISO 27001 and SOC 2 Type 1 certified), Synopsys-hosted solution that delivers flexible, turnkey emulation to accelerate early RTL verification and regression, software bring-up, performance validation, power analysis, and system validation for modern chip designs. Based on the popularity of this cloud-based approach, we’re now expanding the offering to address the need for more capacity, accessibility, and regional data compliance.
Shown: ZeBu Server 5 systems in Synopsys’ U.S. ZeBu Cloud Data Center provide a high-performance, flexible, energy-efficient, and reliable pre-silicon emulation and software validation solution for AI chip developers.
More Options, More Flexibility with ZeBu Cloud
Like its on-premises counterpart, ZeBu Cloud provides industry-leading performance to verify designs at the pre-silicon stage, helping identify downstream issues well before they reach a point when correcting them becomes more problematic. Its hybrid emulation capabilities allow designers to combine virtual models with simulation techniques. And it facilitates more accurate, pre-silicon software development by testing the interactions between hardware and software in an emulated environment. In doing so, ZeBu Cloud compresses the verification and software validation process — accelerating design cycles, reducing risk, and enabling precision at scale.
To meet the increasing demand for these capabilities, we are tripling the capacity of the ZeBu Cloud offering in our U.S. data center. We have also certified our data center in the Netherlands for ZeBu Cloud to support EMEA customers needing localized data storage and regional data compliance.
Supporting a Variety of AI Chip Design Needs
ZeBu Cloud provides a quick-start, convenient access model for a variety of needs and diverse range of customers, from agile startups to large, established enterprises.
Larger, established companies often have on-premises emulation systems in place. However, they may still face peak usage periods when their internal resources are stretched thin or capacity is maxed out. Alternatively, they may want to partition the testing of certain subsystems across on-premises and cloud environments. In such scenarios, ZeBu Cloud offers a flexible solution for handling overflow demand and enables the parsing of subsystem testing, ensuring projects stay on track even during periods of high activity.
Establishing an on-premises emulation system is a non-trivial, capital-intensive, and time-consuming endeavor, and many want to avoid such commitments — especially in the early stages of chip design and testing. For companies without onsite emulation capabilities, the cloud provides a faster path to validation testing without large, upfront capital investments in infrastructure and other resources.
Some companies, like the innovative AI startup SiMa.Ai, start with ZeBu Cloud and then transition to an on-premises setup — or use both — as their needs evolve.
“Synopsys ZeBu Cloud was instrumental in enabling our team to quickly initiate pre-silicon emulation for optimizing our first MLSoC edge AI platform,” said Srivi Dhruvanarayan, vice president of Hardware at SiMa.ai. “Leveraging the ZeBu Empower pre-silicon power profiling capability, we improved frames per second, per watt by 2.5X in just eight months. We also observed an impressive 95-97% accuracy between the pre-silicon power emulation estimates and actual silicon performance, giving us confidence in the design. Based on these outstanding results and their significant impact on our business, we have also invested in on-premises ZeBu systems to test future MLSoC platform generations.”
Why the Cloud?
There are a few key reasons why cloud-based emulation with ZeBu is well positioned to help today’s engineering teams:
Peak capacity handling: Design cycles have ebbs and flows in terms of verification performance requirements, and the cloud model provides an ideal way to adjust capacity on demand.
For companies with on-prem emulation systems, ZeBu Cloud offers a flexible solution for managing peak usage periods, preventing bottlenecks and ensuring project timelines are met. Other companies choose to rely entirely on a cloud-based model that offers more options from a resource standpoint (financial and human).
Financial flexibility: ZeBu Cloud’s SaaS model eliminates the need for significant upfront capital investments. It shifts spending from capital expenditures (CapEx) to operating expenses (OpEx), making it an attractive option for startups in the early stages of funding or companies needing to expand within annual budget constraints. New contract models enable ZeBu Cloud in combination with EDA SaaS to be paid with Cloud Credits for additional financial- and project-relevant decisions.
Global access: ZeBu Cloud is securely and readily accessible from anywhere, fostering collaboration among globally distributed teams and facilitating remote work.
Robust security: ZeBu Cloud adheres to strict security protocols and is compliant with industry standards like ISO 27001 and SOC 2 Type 1 , ensuring the protection of sensitive design data.
Data sovereignty: With data centers in the U.S. and Europe, ZeBu Cloud allows customers to choose where their data resides, addressing concerns about data sovereignty and compliance with regional regulations.
Empowering AI Chip Innovation
Our recent expansion of ZeBu Cloud capacity and accessibility is a testament to the growing demand for AI chips and desire to accelerate design cycles and innovation. By catering to the diverse needs of startups and established companies alike, the cloud-based emulation solution delivers a range of benefits — from reduced development time and risk to financial flexibility, peak capacity management, and regional data compliance.
Continue Reading
Blog
4 min read
/
Mar 20, 2024
Faster, Higher Capacity Emulation and Prototyping for AI Workloads
Blog
10 min read
/
Dec 19, 2022
2022's EDA & Chip Design Advancements
