– the more data, the better
Process more data in less time and with less hardware? It can be done. The Big Data Benchmark recently conducted leaves nothing to be desired in terms of clarity: When it comes to Big Data, Talend’s software offers some significant speed advantages over the competition.
For the benchmark, the expertise of MCG Global Services, a market leader in information management, was used to compare the performance of Talend Big Data Integration with the Informatica Big Data Edition. Three cases were the focus of this benchmark.
Case 1: the impact of website visits on sales
How do website visits affect sales? How many pages do customers visit before they make a purchase?
In this case, Talend had assembled a dataset in 10 minutes, while Informatica needed 1 hour and 27 minutes to do so. Talend was therefore 9 times faster.
Case 2: the impact of coupons on sales
In what ways do coupon promotion campaigns affect sales? Do customers who have seen a coupon advertisement visit our website and buy more or additional products that they would not have bought otherwise, without a coupon?
Talend took 15 minutes to assemble a dataset, compared to 2 hours and 23 minutes by Informatica. Talend was 10 times faster here.
Case 3: The influence of product recommendations
How do product recommendations affect sales? Do customers tend to buy additional products based on these recommendations?
To compile a dataset in this case, Talend needed 22 minutes, while Information took a whopping 4 hours and 9 minutes. Talend was thus 11 times faster.
The more data, the better Talend performs
Talend’s advantages over Informatica increase as data volumes grow. How? By taking advantage of Apache Spark’s capabilities where operations on data occur in memory, Talend can integrate datasets much faster. In fact, Spark uses fast Remote Procedure Calls for efficient handling and execution of jobs. The Spark framework uses a thread pool for task execution, rather than a pool of Java Virtual Machine processes, which allows for much faster processing and execution of jobs. Big Data solutions such as Talend that generate native code for the Spark framework can therefore execute tasks in milliseconds, unlike solutions that do not support Spark such as the Informatica Big Data Edition that require seconds or sometimes even minutes for the same task.
