AVX2 doesn't scale

[GoGetEM]

I am writing a program that uses AVX2.  I wrote the code and am now testing throughput.  With a single thread I get about 7 Mbps of processing.  As I add more threads, the throughput gets up to about 14 Mbps if I use 12 cores:

I am curious if anyone knows how to make AVX2 scale well when multi-threading.   I assume that each core has its own AVX2 silicon dedicated to it, is this correct?

Thanks,

K-

processor : 11
vendor_id : GenuineIntel
cpu family : 6
model : 158
model name : Intel(R) Xeon(R) E-2276M CPU @ 2.80GHz
stepping : 13
microcode : 0xd6
cpu MHz : 2690.582
cache size : 12288 KB
physical id : 0
siblings : 12
core id : 5
cpu cores : 6
apicid : 11
initial apicid : 11
fpu : yes
fpu_exception : yes
cpuid level : 22
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb invpcid_single ssbd ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid mpx rdseed adx smap clflushopt intel_pt xsaveopt xsavec xgetbv1 xsaves dtherm ida arat pln pts hwp hwp_notify hwp_act_window hwp_epp md_clear flush_l1d arch_capabilities
bugs : spectre_v1 spectre_v2 spec_store_bypass swapgs taa itlb_multihit srbds
bogomips : 5599.85
clflush size : 64
cache_alignment : 64
address sizes : 39 bits physical, 48 bits virtual

[Alberto_Sykes]

GoGetEM, Thank you for posting in the Intel® Communities Support.

In order for us to provide the most accurate assistance on this scenario, we just wanted to confirm a few details:

Are you a developer?

Are you working on a project?

What is the project about?

Any questions, please let me know.

Regards,

Albert R.

Intel Customer Support Technician

A Contingent Worker at Intel

[GoGetEM]

I am a developer.

Yes I am working on a project.

We are doing an error correction algorithm using AVX2.

K-

[Alberto_Sykes]

GoGetEM, Thank you very much for confirming that information.

In this case, in order to gather the information that you are looking for, please visit, sign-in and submit your inquiry in our Intel® Developer Zone site, in there you will receive further peer to peer assistance on this topic:

https://software.intel.com/content/www/us/en/develop/home.html

Regards,

Albert R.

Intel Customer Support Technician

A Contingent Worker at Intel 

[GoGetEM]

Alberto,

It is not clear from the links where and what I should submit.

Please be specific.

K-

[GoGetEM]

Your Support request number is 04883850

[Alberto_Sykes]

GoGetEM, Thank you very much for your response.

Just to let you know, I just move your thread to the proper department, which is "Software Products", just in case here is the link:

https://community.intel.com/t5/Software-Products/ct-p/software-products

They will be able to further assist you with this topic as soon as possible.

Regards,

Albert R.

Intel Customer Support Technician

A Contingent Worker at Intel