DAX performance results are the same as with dynamodb

I created DAX cluster with 3 nodes x dax.r5.xlarge and cache TTL both query , record 300000 ms.

I also created dynamodb table with about 200 records and run simple python script based on https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/DAX.client.run-application-python.04-query-test.html which queries table 100 times. The results with and without DAX are almost the same, in same cases without DAX the results are event better. Can you explain what I'm doing wrong? We also tested the same behavior with many tables which size is more than 50Gb and the results are almost the same. There is no expected improvements of using DAX. Any suggestions?

DynamoDB

python query-test.py

Querying the table 100 times, using the Boto3 client. .................................................................................................... Total time: 16.4085 sec. Average time: 0.164084955000435.

DAX

python query-test.py daxs://dax-dev.yyy.dax-clusters.us-east-1.amazonaws.com

Querying the table 100 times, using the DAX client. .................................................................................................... Total time: 16.3092 sec. Average time: 0.16309245042037218.

DynamoDB

python query-test.py

Querying the table 100 times, using the Boto3 client. .................................................................................................... Total time: 14.5192 sec. Average time: 0.1451919324998744.

DAX

python query-test.py daxs://dax-dev.yyy.dax-clusters.us-east-1.amazonaws.com

Querying the table 100 times, using the DAX client. .................................................................................................... Total time: 15.6273 sec. Average time: 0.1562731408304535.