Trifacta’s in-memory engine, Photon executes single node data transformations up to 6X faster, complementing Trifacta’s support for distributed engines Apache Spark and Google Cloud Dataflow
SAN FRANCISCO May 16, 2017 Trifacta, the global leader in data wrangling, today announced Enterprise Strategy Group (ESG) published its latest ESG Lab Review, evaluating the speed and efficiency of Trifacta’s Photon Compute Engine for data wrangling. The report validates Trifacta’s in-memory engine, Photon is the fastest, most efficient data processing engine for wrangling data sets that don’t require parallel processing, completing transformations on single node environments up to 6X faster, while using up to 98 percent less memory compared to Apache Spark. The audit confirms Photon enables users to execute data wrangling jobs within desktop or single-server settings with unparalleled speed and efficiency.
The Photon engine is packaged in every edition of Trifacta, powering the data processing directly within the application as well as serving as the data processing engine for single node data wrangling execution. ESG Lab focused its performance evaluation of Photon only on single node execution performance and not on in-application processing execution.
“For many organizations and for business analysts in particular demand for an efficient, purpose-built data processing engine for smaller scale data wrangling needs has never been more apparent,” said Nik Rouda, senior analyst at ESG. “With Photon, Trifacta fills a void in the market it provides a fast, highly efficient data wrangling engine for diverse data sets that don’t require parallel processing execution. Photon complements Trifacta’s highly intuitive user experience and workflow with a purpose-built data processing framework to power its application performance and the execution of wrangling jobs that don’t require parallel processing.”
In addition to Photon, Trifacta supports a variety of multi-purpose engines, including Apache Spark and Google Dataflow for large-scale parallel processing, but created Photon to specifically handle performance of wrangling execution directly in the application and in non-parallel computing platforms such as a desktop or single server. The ESG Lab Review validates Photon outperforms general-purpose computing frameworks such as Spark for data sets that only require the computing resources of a desktop or single server. By providing support for a variety of data processing engines including Photon, Spark and Google Dataflow, each optimized for different data volumes and computing environments, Trifacta users can always leverage the most efficient compute engine for the wrangling task at hand.
Highlights from the ESG Lab Review include:
- Numerical Transformation Performance: For the execution of simple and complex numerical transformation on 1GB, 2GB, 5GB, and 10GB data sets including addition, division, sorting and grouping, Photon performed as much as 2.6X faster than Spark.
- Textual Transformation Performance: For the execution of simple and complex textual transformation tasks on 1GB, 2GB, 5GB, and 10GB data sets that included merging, extracting, creating, sorting, and joining, Photon performed as much as 6.6X faster than Spark.
- Memory Utilization: For the execution of simple and complex textual transformation tasks on 1GB, 2GB, 5GB, and 10GB data sets that included merging, extracting, creating, sorting, and joining, Photon memory requirements were as much as 85X less than Spark.
“We developed Photon to enhance two aspects of our product: application performance and scale, and highly-optimized single server execution of wrangling jobs. Providing a fluid user experience and optimizing processing power for any scale of data are core tenets of Trifacta’s architecture and our differentiation compared to other data preparation products,” said Sachin Chawla, VP of engineering at Trifacta. “The stellar results found in ESG’s evaluation are a testament to the engineering talent at Trifacta and the vision of our founding team.”
Trifacta Wrangler is used by tens of thousands of users at more than 7,300 companies in 143 countries around the globe. The world’s leading brands, including Google, PepsiCo, eBay, Munich Re, Royal Bank of Scotland, Kaiser Permanente, and LinkedIn are unlocking the potential of their data and accelerating time to insight using Trifacta’s market-leading data wrangling solution.
For more information about the ESG Lab Review, visit here: http://www.trifacta.com/gated-form/esg-evalution-trifacta-wrangler-enterprise/
- Read about Photon Compute Engine in Trifacta’s blog
- Learn more about Trifacta
- Download Trifacta Wrangler for free
- Follow us on Twitter
- Become a fan on Facebook
- Connect on LinkedIn
Trifacta, the global leader in data wrangling software, significantly enhances the value of an enterprise’s big data by enabling users to easily transform and enrich raw, complex data into clean and structured formats for analysis. Leveraging decades of innovative work in human-computer interaction, scalable data management and machine learning, Trifacta’s unique technology creates a partnership between user and machine, with each side learning from the other and becoming smarter with experience. Trifacta is backed by Accel Partners, Cathay Innovation, Greylock Partners and Ignition Partners.
Nolan Necoechea for Trifacta