Ask Your Question

is there parallel computation or not

asked 2020-07-10 09:36:02 -0500

updated 2020-07-10 12:45:01 -0500

supra56 gravatar image

I would like to choose right CPU for application with cv2. The weakest part of the code in the sense of time consuming is definition сv2.findTransformECC. Is there parallel computation in the cv2.findTransformECC or it's just one thread process? I am thinking between AMD Ryzen 3 PRO 3200GE and Intel Core i5 6200U

I think AMD Ryzen 3 PRO 3200GE has advantages if cv2.findTransformECC is single thread process and Intel Core i5 6200U in case multithread process in cv2.findTransformECC

Is there any thoughts on this questions? is cv2.findTransformECC single thread calculation and AMD Ryzen 3 PRO 3200GE CPU has advantages as more cpu Clockspeed rate but less threads then Intel Core i5 6200U?

edit retag flag offensive close merge delete

1 answer

Sort by » oldest newest most voted

answered 2020-07-10 10:39:26 -0500

berak gravatar image

yes and no ;(

the function itself does not contain any parallel_for_ loops or ocl kernels, but it calls a lot of other functions like filter2D or warpAffine, which are heavily optimized

edit flag offensive delete link more

Question Tools

1 follower


Asked: 2020-07-10 09:36:02 -0500

Seen: 71 times

Last updated: Jul 10 '20