Mobileye Reduced 45% On Their AWS Costs Leveraging Granulate

45%

Reduced AWS Costs

44.5%

Faster Job Completion Time

Mobileye runs core workloads on AWS for greater speed, agility, and compute power. Mobileye turned to Granulate's real-time continuous optimization solution to achieve better performance for PySpark workloads and significantly reduce costs on their EC2 workloads. Mobileye documented a 44.5% reduction in average job completion time. The immediate impact of the shorter job completion times was 45% cost reduction, due to fewer instance hours required to run the same workload.

About

Mobileye is the global leader in the development of computer vision and machine learning, data analysis, localization and mapping for Advanced Driver Assistance Systems and autonomous driving. Mobileye technology keeps passengers safer on the road, reduces the risk of traffic accidents, saves lives, and has the potential to revolutionize the driving experience by enabling autonomous driving.

Because our new code gets seamlessly accelerated every day, we can develop more quickly, and this means our time to market is much faster. To create a country map using Intel Workload Optimized by Granulate, it’s now taking us half the time at half the cost

Tal Babaioff, vice president of mapping and localization and co‒general manager of REM at Mobileye, an Intel company

The Challenge

Mobileye runs big-data workloads on AWS, using PySpark to support its RoadBookTM of localized drivable paths and visual landmarks, which provides centimeter-accurate mapping for autonomous driving. Amazon EC2 R4.8/16/24XL and R5.8/16/24XL on Spot instances, specifically Intel-powered instances enables Mobileye to quickly innovate on top of a highly scalable, fault-tolerant infrastructure, but these services have also become a significant cost driver for the organization.

Results

Cost Reduction – 45%
Reduced Job Completion Time – 44.5%

Granulate agents were deployed on a Mobileye Elastic Compute Cloud (EC2) workload as container side-cars and began learning the workload, dataflow, and resource usage patterns within minutes. Following a few days of learning, the agents were ready to be activated to start optimizing the application’s performance.

Mobileye monitored the performance and results throughout the Granulate activation. Mobileye documented a 44.5% reduction in average job completion time. The immediate impact of the shorter job completion times was a 45% cost reduction, due to fewer instance hours required to run the same workload. Mobileye purchased Granulate through the AWS Marketplace using a Private Offer.