I'm a simple PBI developer, I see Maco Russo, I click like

This is by far the most innovative technical video I have seen so far. Keep up the good work. Your knowledge is always a great treasure for the Power BI community.

This is phenomenal, thank you so much! This video deserves at least 1 million views! :)

Dropping knowledge bombs from start to finish, I feel smarter than I was 25mins before

Really helpful in understanding the two different engines. Loved the way you used the book analogy and paper and pen before moving on to the code. It's helped me in understanding how to analyse queries when optimising code and reminds me of TCP protocol, for example, in that each process is efficient as it specialises in it's own job, storage engine gets the data, formula engine calculates and visualises. Thank you

For each video you post, I always imagine that you won't be able to produce a better video than the last one you posted... But I'm always wrong! Each video is always better, and it's getting better and better each day! It's unbelievable! You are definitively the DAX Masters! Thanks for sharing! 👏👏👏

Great video, thank you Marco. I like how you drew it out on paper. I now view the SE like a math book... a collection of all of the figures and data required to perform math, however still requires the student (FE) to actually solve for the result.

Excellent video. Thank you. I am a very visual learner and this explanation (along with your book) is incredibly helpful.

Great video and hugely informative. Thank you! Although I spend a lot of time in DAX, this did a great job in breaking down concepts and filling in gaps in what I know. :)

Thanks for the excellent explanation! Looking forward to more of the videos on this topic

Thanks Marco! Great video which I will most likely forward to everybody whoever asks about this instead of trying to explain myself :-)

Very helpfull explanation. Only heart the words but didn't know how it works. I'am greatful for your videos.

Excellent tutorial! Very interesting and useful to know! THX a lot Mr. Russo

very good explanation

Wish more content on this! Thanks

Great explanation, thanks

A nice long video! Yeaa!

Finished watching

Thanks!!! Really good explanation !! 5 stars XD

That is an innovative way of explaining the difference between formula engine and storage engine using the book analogy! 👍 Question: I understand FE needs the SE results materialized In-Memory in an uncompressed form to do it's computation. Is the same applicable to DirectQuery? Does the SQL output gets materialized in memory before formula engine acts on it? Also, if FE don't have to do any computation on SE results, still the materialization happens before sending the results to client ?

Great explanation, thanks a lot for this great content ! Question - if you have more than one visual and a different measure, like Total Cost, would it re-use the SE values (and the FE aggregates at a higher level) or run another query ? .

thanks marco,I love this vedio......

Thank you for taking up this topic. So when materialization happens that's when 1 million row limit is triggered with top 1000001. If there was no materialization then Direct query wouldn't have this limit.

Hi Marco, Is there any way we can optimize export to excel data from the Power BI report?

Thx

Cool

If I am seeing 99% in the Formula Engine with 9 SE Queries but the total time is 87,357ms I know I have a problem. BUT where do I start looking? I assume its not a data issue due to the storage engine was fast (469ms) so I guess a DAX issue...?

Thank you @sqlbi for the amazing tutorial as always about SE and FE along with calculation groups. I have implemented the calculation for a report but the complex nature of the requirement is to have 3 layers of drill down at row level and 3 layers of drill down at column level and all of them have time intelligence e.g., MTD, QTD, YOY against Actuals, Budget and Actuals v Budget information per selected criteria e.g., Net sales, Gross Sales, PBT et cetera. The performance just drags completely and just fails later after hitting threshold. Could there be anything that can be done to improve the performance? The measures uses keep filters, selected measure with time intelligence calculation. Needless to say it is a matrix view. Any tip, suggestion would help.