Case Study: Bigabid cuts compute costs by 60% with Granulate

Bigabid Cuts Compute Costs by 60%

Bigabid, a leader in the demand-side platform (DSP) market, needed to double its business without increasing cloud costs. Running a multi-tiered C++ application with an RTB module on AWS spot instances, the company was processing more than 80,000 requests per minute and had to keep each request within 120 ms while also reducing revenue lost to bid timeouts. To address this challenge, Bigabid turned to Granulate’s real-time, continuous optimization solution.

Granulate installed agents that observed and automatically optimized Bigabid’s application, improving memory allocation and release, reducing fragmentation, and increasing concurrency without any code changes. The result was a 60% reduction in compute costs, a 70% reduction in p95 latency, and a 90% reduction in bid timeouts, capturing more potential revenue. Bigabid also reported that the solution required only about 5 hours of staff time.

Bigabid

Amit Attias

Chief Transformation Officer