i7-1260P huge performance degradation when AVX instructions are mixed with legacy SSE

[Rafaello7]

i7-1260P CPU, when AVX instructions using ymm registers are mixed with legacy SSE instructions, the code execution slows down dramatically. The same code executed on another processor (i3-10110U) exhibits only a minimal slowdown. Is it normal, or is it a bug in the i7-1260P CPU?

I have posted sample code on another forum.

[Jocelyn_Intel]

Hello, @Rafaello7  

Thank you for posting on the Intel® communities.  

Please be aware that the proper team that supports these inquiries; the Developer Software Forums team, has answered your questions in the original thread. We will proceed to close this case avoid duplicated threads and keep order. 

This thread will no longer be monitored by Intel. Thank you for your understanding. 

Best regards,  

Jocelyn M.   

Intel Customer Support Technician.