question

gatsby53-6558 avatar image
0 Votes"
gatsby53-6558 asked GiftA-MSFT edited

Difference in processing time between Azure Machine Learning Studio and Azure Machine Learning Studio (classic)

I used to use Azure Machine Learning Studio (classic).
Creating the same workout in Azure Machine Learning Studio takes about 20 times longer than classic.
Virtual machine size is Standard_DS3_v2 (4 core、14 GB RAM、28 GB disk).
Steps that have been executed once will be processed quickly from the next time onward, but steps that have been changed even slightly will take 20 times longer than classic.

How can I process at the same speed as classic?

azure-machine-learning
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

1 Answer

GiftA-MSFT avatar image
0 Votes"
GiftA-MSFT answered GiftA-MSFT edited

Hi, thanks for your feedback. AML classic studio appears to be faster in some cases because it uses a Fixed Compute (and always available). However, AML Classic lacks flexibility and scalability that the new platform offers. With designer, you have greater flexibility but depending on the task (e.g. smaller tasks), the processing time may seem longer than classic due to overhead for preparing each step. For smaller tasks, majority of execution time is spent on overhead. Furthermore, when input data changes, it may take longer. If no changes are made, the pipeline would automatically use the cached result of that module, so it should be faster compared to the first run. The product team are aware of this limitation and working to improve the experience. For compute heavy tasks, we recommend you pick a larger VM to improve processing speeds. Please review this document for ways to Optimize Data Processing. Feel free to submit feedback directly to the product team by using the 'smiley' feedback icon in Azure ML Studio. Other Similar Posts: (1), (2).



--- Kindly Accept Answer if the information helps. Thanks.



5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.