R with OpenBLAS - r

Question

I have the code in R that works
well using the OpenBLAS, on a more modest computer with the following configurations:
[pedro#pedro-avell ~]$ lscpu
Arquitetura: x86_64
Modo(s) operacional da CPU: 32-bit, 64-bit
Ordem dos bytes: Little Endian
Tamanhos de endereço: 39 bits physical, 48 bits virtual
CPU(s): 8
Lista de CPU(s) on-line: 0-7
Thread(s) per núcleo: 2
Núcleo(s) por soquete: 4
Soquete(s): 1
Nó(s) de NUMA: 1
ID de fornecedor: GenuineIntel
Família da CPU: 6
Modelo: 60
Nome do modelo: Intel(R) Core(TM) i7-4710MQ CPU # 2.50GHz
Step: 3
CPU MHz: 3435.870
CPU MHz máx.: 3500,0000
CPU MHz mín.: 800,0000
BogoMIPS: 4989.80
Virtualização: VT-x
cache de L1d: 32K
cache de L1i: 32K
cache de L2: 256K
cache de L3: 6144K
CPU(s) de nó0 NUMA: 0-7
Opções: 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 arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm cpuid_fault epb invpcid_single pti ssbd ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid xsaveopt dtherm ida arat pln pts flush_l1d
In the test with a computer with configurations above, I observed by benchmark a 60% improvement in code performance time when considering parallelization. For that, I had to consider export OPENBLAS_NUM_THREADS=1. Therefore,
parallel::mclapply(1:8, FUN = function(x) func_metodos(), mc.cores = 4)), wherein func_metodos() is a function that I implemented in R and want to repeat 8 times.
However, when considering the same simulation on the computer with configurations below, I do not know how to tell the operating system to take into account the existence of 2 sockets, each with 10 cores and 2 threads per core. Should I consider export OPENBLAS_NUM_THREADS = 1 or export OPENBLAS_NUM_THREADS = 2? I considered both and ran the code in R:
parallel::mclapply(1:8, FUN = function(x) func_metodos(), mc.cores = 40))
[marcelo#ursal ~]$ lscpu
Arquitetura: x86_64
Modo(s) operacional da CPU: 32-bit, 64-bit
Ordem dos bytes: Little Endian
Tamanhos de endereço: 46 bits physical, 48 bits virtual
CPU(s): 40
Lista de CPU(s) on-line: 0-39
Thread(s) per núcleo: 2
Núcleo(s) por soquete: 10
Soquete(s): 2
Nó(s) de NUMA: 2
ID de fornecedor: GenuineIntel
Família da CPU: 6
Modelo: 85
Nome do modelo: Intel(R) Xeon(R) Silver 4114 CPU # 2.20GHz
Step: 4
CPU MHz: 800.021
CPU MHz máx.: 3000,0000
CPU MHz mín.: 800,0000
BogoMIPS: 4401.33
Virtualização: VT-x
cache de L1d: 32K
cache de L1i: 32K
cache de L2: 1024K
cache de L3: 14080K
CPU(s) de nó0 NUMA: 0-9,20-29
CPU(s) de nó1 NUMA: 10-19,30-39
Note: When I consider export OPENBLAS_NUM_THREADS = 1 or export OPENBLAS_NUM_THREADS = 2, it appears that the threads are distributed to only 20 of the 40 available cores.
Thanks for any suggestions to solve this issue so that I can use the 40 available cores.