Game-Changing Breakthrough in Pharma: Speedata's Chip Revolutionizes Drug Discovery, Accelerating Complex Pharmaceutical Workload from 90 Hours to 19 Minutes in Simulation
Speedata, a company specializing in Analytics Processing Unit (APU) technology, has announced a groundbreaking achievement in the pharmaceutical industry. They conducted a simulation of their APU on a compound similarity analysis workload, yielding remarkable results. While the analysis typically required 90 hours when processed on a CPU, Speedata's APU completed it in just 19 minutes, marking a significant improvement of 280 times in speed.
Compound similarity analysis is a crucial process in drug discovery, but it often faces challenges due to bottlenecks in input/output, compute, and memory. By accelerating compound similarity functions for drug-target interaction predictions, Speedata's chip empowers researchers to analyze larger datasets, ultimately expediting the research process and hastening the development of life-saving drugs.
Mark Ramsey, former Chief Data Officer of GSK and Managing Partner of Ramsey International, commended Speedata's APU, stating that it opens the door for breakthroughs in pharmaceutical and health research. With its exceptional performance, drug discovery data analysis can now be completed in minutes, presenting limitless possibilities for life-saving implications. Ramsey believes that this is just the beginning of hardware acceleration use cases in the pharmaceutical space for Speedata.
In the simulation, nine million compounds were used across 100 servers for the compound similarity analysis workload. Comparing Speedata's APU to a standard CPU on the same infrastructure, the APU achieved a remarkable speedup of 280 times, completing the analysis in just 19 minutes. Additionally, Speedata conducted the same analysis on a single server equipped with four APUs, accomplishing the workload in 8 hours. This demonstrates that the analysis, which previously required a datacenter with 100 servers running for four days, can now be performed on a single machine overnight.
Speedata's APU addresses the main bottlenecks of data analytics, significantly enhancing speed and performance while reducing costs and increasing efficiency. The APU's architecture seamlessly integrates with existing legacy software, allowing for a smooth transition of workloads without the need for extensive code or framework modifications.
Jonathan Friedmann, Co-founder & CEO of Speedata, emphasizes the impact of these results, particularly in expediting time-to-market in the pharmaceutical industry. Pharmaceutical workloads represent just one example of the many data analytics workloads that Speedata's processor can accelerate across critical industries. The company looks forward to helping companies, data centers, and cloud providers improve their data analytics capabilities, envisioning a future of significant advancements in big data.
