How do the cores in an i.MX7Dual compare to an i.MX6ULL?

[chris_f]

I'm particularly interested in double precision floating point performance. I've benchmarked a 528MHz i.MX6ULL and need more DP FLOPS. Can I expect a 1GHz i.MX7Dual to give me ~2x the DP FLOPS for each core?

I understand it may seem a strange question but I benchmarked a 1GHz TI Cortex-A8 part this week and that was significantly slower than the 528MHz i.MX6ULL for double precision MACs. Double the speed for integers but it seems it has a less capable Neon. ARM's documentation on the subject is quite dense.

So, before I buy yet another EVK ... Are each of the Cortex-A7 MPCores in an i.MX7Dual the same as the one in the i.MX6ULL?

[igorpadykov]

Hi chris

one can look at considerations about Cortex A7 performance

arm - Peak FLOPs per cycle for ARM11 and Cortex-A7 cores in Raspberry Pi 1 and 2 - Stack Overflow 

also some data can be shared creating service request.

Best regards
igor
-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------