Forum

Author Topic: Hardware requirements  (Read 704 times)

ndr

  • Newbie
  • *
  • Posts: 2
    • View Profile
Hardware requirements
« on: July 09, 2021, 03:30:19 PM »
Hi everybody,

I need a new machine to work mostly with Metashape as well as other softwares such as GIS and CAD. 

Laptop is what I need now, and I am stuck on the choice between two.

Both cost the same exact price and have 512 Gb SSD and same graphic.



1) has a 8 cores, 16 threads, base clock 2,9 -- max clock 4,2; 17nm manifacturing process, 16 GB RAM not expandable but higher freq.

2) has a 6 cores, 12 threads, base clock 2,6 - max clock freq 5; 7nm manifacturing process  16 GB RAM expandable to 32 GB
but lower freq.

the first CPU seems faster, but in that case I cannot get more than 16 RAM.

Will I be good with just 16 GB?

Imagining to work on heavy projects, would you choose the solution with the faster CPU but limited RAM, or the solution with the slower CPU but  expandable RAM?

Thanks to you all in advance!



andyroo

  • Sr. Member
  • ****
  • Posts: 397
    • View Profile
Re: Hardware requirements
« Reply #1 on: July 09, 2021, 10:09:10 PM »
Not sure what your general workflow and project size is, but my guess is that you'd see faster results with the AMD depending on workflow (maybe even 20%+ improvement) but the RAM would limit the # of images you can process at once. Making mesh is more demanding in my experience than making dense cloud/dem, so kinda depends on your workflow and estimated project size. Would be cool to get numbers running the Puget Systems extended benchmark for each - maybe your vendor could do that and give you results?

Whichever one you get, compare hyperthreading/SMT on vs off - you're right at the point where I'm not sure you'd see improvement or not, but I've seen significant differences on some systems.

ndr

  • Newbie
  • *
  • Posts: 2
    • View Profile
Re: Hardware requirements
« Reply #2 on: July 10, 2021, 03:51:32 AM »
Not sure what your general workflow and project size is, but my guess is that you'd see faster results with the AMD depending on workflow (maybe even 20%+ improvement) but the RAM would limit the # of images you can process at once. Making mesh is more demanding in my experience than making dense cloud/dem, so kinda depends on your workflow and estimated project size. Would be cool to get numbers running the Puget Systems extended benchmark for each - maybe your vendor could do that and give you results?

Whichever one you get, compare hyperthreading/SMT on vs off - you're right at the point where I'm not sure you'd see improvement or not, but I've seen significant differences on some systems.

Hi Andyroo, thank you for you answer.

In fact I am more concerned about being limited on the number of images by a not expandable 16 GB RAM, than about the speed of CPU, in this case.

Let's say I have a close-range dataset with 800 images, nadir and oblique, and I have to produce dense cloud, mesh, DEM and  orthomosaic.

Do you think the 6 core CPU I described can make it? or they are too few that will badly affect the quality or even not get the job done?

Cause if the 6 cores gets the job done even if with a slower pace, than I'll choose it and get the chance to bring the RAM up to 32 GB, as I see it is probably the real limiting factor in this case. Am I wrong in this choice? How do you see it?