When it comes to 'Big data,' now everyone has it all.It is widely found in the fields of biotech, finance, agriculture, education and transportation, and every industry is hoping to reshape its business essence.
However, for the semiconductor industry?
To this end, IC Manage Inc., a provider of solutions for design and IP management software to chip companies, recently announced 'Big Data Labs'.
Dean Drako, chief executive of Dean DrakoIC Manage, described the Big Data Labs as a 'platform' for designing IC design customers 'new design analytics tools based on big data.'
In the 'big data' era, semiconductor companies are already designing ICs for data centers, but the question is whether the chip designers themselves use big data or not, and they already have huge amounts of raw data - from different EDA vendors tool.
But have electronic designers found a great way to optimize and accelerate their designs with big data?
In simple terms, the answer is 'not yet'.
Of course, the semiconductor industry has been using data management software for many years. Gary Smith EDA Principal Analyst Laurie Balch explained that IC Manage has always been committed to providing 'a large amount of data to be kept secure and organized so that it can be accessed by others' But for an analysis tool that allows IC designers to apply data to intelligent decisions, Balch said: 'We're in the early stages of getting started.'
IC Manage is not a traditional EDA supplier and Balch explains that instead of making traditional EDA tools such as simulation, composition or layout, the company's expertise is "EDA Enterprise Tools." Balch described IC Manage as a " IC Design Database Market has a place in the company '.
In an era when electronic design is known for creating large amounts of data, she found that chip suppliers are facing the growing challenge of how to manage their data efficiently, and IC Manage may be the first to propose a solution.
Unstructured data
Drako explained that by definition, 'big data' is made up of a lot of unstructured data.
He admits that the electronic design community has seen the enormous challenges of unstructured data - from a variety of tools designed by different EDA companies.
However, most IC designers do not have enough tools to absorb these things, not to mention how they are understood, which is time-consuming and resource-intensive, after all.
Drako also stressed that it is not easy to concatenate separate data sets between tools and suppliers.
In addition, he added: 'Few industries and companies have enough expertise and resources available to rapidly develop workable insights and create management options and implementation details.
This is where IC Manage wants the company to work.
Drako explains that IC Manage overlays unstructured data with organized design data. "By combining unstructured data such as validation records and structured data (electronic design data), we have provided a hybrid database, 'Let the chip company use it to perform high-performance advanced EDA analysis.
The result of IC Manage's hope is to implement a platform that provides visual analytics to help users create interactive reports.
Chip out of the forecast
This is not the first big data design tool IC Manager offers for IC Manage.
A few years ago, the company developed a big data product called 'Envision Design Progress Analytics' that provided a basis for IC Manage's customers to accurately predict the launch of their new chips.
IC Manage is taking a step forward with the launch of Big Data Labs, which is not enough to organize big data and make it available to the design team or anyone across the company.
IC Manage hopes to further develop (and possibly customize) new tools by working closely with customers (chip companies) and partners (EDA tool providers), allowing designers to track the contributions of each designer and revise their history , IP re-use, and various actions. The resulting tools allow designers to see the impact of their decisions on the design process. Drako said the analysis provided will also help them make informed decisions.
Functional verification tools
IC Manage also introduced Vision Verification, the first verification analysis tool based on the Big Data Labs platform, which offers "near-instantaneous visual analytics" by leveraging the platform's ability to connect to multi-vendor environments.
Drako said: 'To understand what is happening, EnVision Verification takes all of the validation data from different EDA vendor environments such as Verilog, Mentor, and Cadence and tracks design activities, regression testing and verification status, and errors. Then find out the change.
Without such big data validation, Drako said: 'Traditionally, if you were part of a team of 300 engineers, you might have to spend a lot of time asking:' Have you changed anything? ',' Which Did you test it? "," Who broke it? "Or" Did we miss something? "
With interactive verification results reporting, Drako said: 'Envision helps accelerate functional verification analysis 10 to 100 times, not only identifying bottlenecks, but also finding the root causes of issues that arise during verification.'
Analyst Balch explains that validation is a "big challenge for electronics designers." Because each one's goal is to design and build first-time right-limited At the cost of chip redesign, "designers must validate it." She points out that verification involves many aspects of testing, and "the results vary depending on the operating conditions, and extreme cases must be understood."
The first tool from IC Manage's newest Big Data Labs is this feature validation tool, along with other logical big data analytics products from the company, including physical verification, timing analysis, and power consumption.
Drako pointed out that functional verification includes many components, including simulations for semiconductors, circuits, digital and analog designs, etc. As a result, Balch speculates that IC Manage will be busy for a long time to further develop its functional verification tools, including customization .
Who will use this tool?
Balch said it's undeniable that data management tools are somewhat 'slow to take off' among chip companies. "Given budget, chip designers are more likely to buy core design tools than big data analytics." They do not think this It's important, and it's also true that this only applies to large design teams. '
As the semiconductor industry continues to be involved in huge mergers and acquisitions, the changes in the broader environment may be faster than previously predicted, for example, if Broadcom successfully acquired Qualcomm, imagine how big a design team will face in the two giants Of the data management nightmare.Mixed companies must monitor the progress of different design teams to ensure that we can share design information and IP.
Compile: Susan Hong