Question Alder Lake - Official Thread

[igor_kavinski]

Especially given that you're using a laptop, I doubt you'll see any benefit from disabling cores or hyperthreading. Will probably hurt, if anything.

My reasoning is that PowerBI may be using the E-cores which may slow it down.

Disabling hyperthreading will allow more bandwidth to be available to the remaining cores and may help with the loading time.

 

[igor_kavinski]

I am now in search of a new desktop, wondering what specs i have to look for in order to make use of a good CPU.

On Intel side, your option is Core i9-13900K with DDR5-7200 CL34 RAM.

AMD has Ryzen 7950X with DDR5-6000 CL30 RAM.

The Intel one may load your file a lot quicker due to the faster RAM. But the AMD one may be faster at processing data and doing so with less electricity usage. Go with whichever one seems economical to you.

To be honest, I wouldn't expect these two to miraculously load your file in seconds. But if they do, that would be surprising. The slow loading time could be more of a software issue (for example, Microsoft Excel is pretty slow with huge files). The RAM on desktop PCs is much faster so you may definitely feel an improvement. Just don't expect a miracle.

 

[coercitiv]

My reasoning is that PowerBI may be using the E-cores which may slow it down.

Disabling hyperthreading will allow more bandwidth to be available to the remaining cores and may help with the loading time.

His screenshot shows 24% overall CPU utilization and it looks like he's importing data from a 3GB .csv file. I would argue there's a high chance that data import is single-threaded or bottle-necked by the way it's parsing the file. The hardware is clearly underutilized.

so what is your recommendation ?

The first thing I would suggest is to change the Task Manager view from Overall Utilization to Logical Processors, this should give you a better idea of what is happening. This is done easily by right-clicking on the graph and choosing the other display option just like in the screenshot bellow:

Spoiler: Task Manager: Performance Display

Use this detailed graph to monitor the 3GB .csv file import again, and watch how the load is distributed. If there's only one or two threads with high utilization then the import process is likely single-threaded and you are limited by software first and hardware second. You should know the "Logical Processors" view shows the P core threads first, E cores threads last: so the first 12 threads are for the P cores (2 per core), the last 8 threads are from the E cores (1 per core).

Keep in mind this is mostly a hardware forum, where we tend to address the issue from the PoV of the hardware resource instead of software optimization. There may very well be people like you who have already run into issues with importing large datasets from .csv into Power Bi. Bellow you can find two such examples, suggesting there are options inside Power Bi that may help with performance in particular cases:
https://community.powerbi.com/t5/Desktop/Large-csv-data-file-of-about-30gb/m-p/2885861#M992742
https://www.fourmoo.com/2016/10/25/power-bi-quick-tips-working-with-multiple-large-csv-files/

 

[Starjack]

His screen shot also shows that he's using Windows 10 for a 12th Gen Processor. I thought is because it was specified that Intel 12 Gen processors might take more advantage using Windows 11 than 10 to get better utilization of the cores, epsecially due to something call Thread Director. Could always correct me if you want.

 

[igor_kavinski]

Core i5-1235U HP laptop benchmarks: https://forums.anandtech.com/threads/hp-probook-450-15-6-inch-g9-benchmarks-core-i5-1235u.2609535/post-40919196