Page 1 of 1

#21 Re: Optimized Apps for Projects

Posted: Fri Feb 08, 2019 8:56 am
by davidbam
Bryan wrote: Sat Jan 26, 2019 4:36 am You only need the 1st one ... AVX2 for the 2990WX or your Intel machines that support AVX2.
Do you happen to know if these apps will work on a FreeBSD machine? I have a 44 thread Xeon e5 v4 which definitely has AVX2

#22 Re: Optimized Apps for Projects

Posted: Fri Feb 08, 2019 9:39 am
by davidbam
Looks like it does - there is an ABI in place for Linux compatibility for other projects. I will report once I get some validated WU

#23 Re: Optimized Apps for Projects

Posted: Sun Feb 17, 2019 4:04 pm
by davidbam
E5-2690 v2 Xeons don't have AVX2, is AVX the next-best optimisation to use please? Flags at end of lscpu output

Code: Select all

$ lscpu
Architecture:        x86_64
CPU op-mode(s):      32-bit, 64-bit
Byte Order:          Little Endian
CPU(s):              40
On-line CPU(s) list: 0-39
Thread(s) per core:  2
Core(s) per socket:  10
Socket(s):           2
NUMA node(s):        2
Vendor ID:           GenuineIntel
CPU family:          6
Model:               62
Model name:          Intel(R) Xeon(R) CPU E5-2690 v2 @ 3.00GHz
Stepping:            4
CPU MHz:             3292.020
CPU max MHz:         3600.0000
CPU min MHz:         1200.0000
BogoMIPS:            5985.49
Virtualisation:      VT-x
L1d cache:           32K
L1i cache:           32K
L2 cache:            256K
L3 cache:            25600K
NUMA node0 CPU(s):   0-9,20-29
NUMA node1 CPU(s):   10-19,30-39
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 arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm cpuid_fault epb pti ssbd ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid fsgsbase smep erms xsaveopt dtherm ida arat pln pts flush_l1d

#24 Re: Optimized Apps for Projects

Posted: Sun Feb 17, 2019 7:44 pm
by Bryan
Theoretically it should be. However, if the 2690 downclocks when running AVX then the SSE3 might actually be faster.

#25 Re: Optimized Apps for Projects

Posted: Sun Feb 17, 2019 7:46 pm
by davidbam
2264 secs on average for first 20 WU with AVX. None have yet validated

#26 Re: Optimized Apps for Projects

Posted: Sun Feb 17, 2019 8:08 pm
by davidbam
Fairly hot for a Xeon but can't say I've ever checked it when running WCG. The temperature difference between the two CPUs is quite normal in a Dell as the exhaust from first CPU fan is intake for the second

Code: Select all

Package id 0:  +73.0°C  (high = +81.0°C, crit = +95.0°C)
Core 0:        +70.0°C  (high = +81.0°C, crit = +95.0°C)
Core 1:        +69.0°C  (high = +81.0°C, crit = +95.0°C)
Core 2:        +71.0°C  (high = +81.0°C, crit = +95.0°C)
Core 3:        +71.0°C  (high = +81.0°C, crit = +95.0°C)
Core 4:        +73.0°C  (high = +81.0°C, crit = +95.0°C)
Core 8:        +69.0°C  (high = +81.0°C, crit = +95.0°C)
Core 9:        +70.0°C  (high = +81.0°C, crit = +95.0°C)
Core 10:       +70.0°C  (high = +81.0°C, crit = +95.0°C)
Core 11:       +70.0°C  (high = +81.0°C, crit = +95.0°C)
Core 12:       +69.0°C  (high = +81.0°C, crit = +95.0°C)

Package id 1:  +80.0°C  (high = +81.0°C, crit = +95.0°C)
Core 0:        +80.0°C  (high = +81.0°C, crit = +95.0°C)
Core 1:        +78.0°C  (high = +81.0°C, crit = +95.0°C)
Core 2:        +78.0°C  (high = +81.0°C, crit = +95.0°C)
Core 3:        +80.0°C  (high = +81.0°C, crit = +95.0°C)
Core 4:        +76.0°C  (high = +81.0°C, crit = +95.0°C)
Core 8:        +75.0°C  (high = +81.0°C, crit = +95.0°C)
Core 9:        +80.0°C  (high = +81.0°C, crit = +95.0°C)
Core 10:       +79.0°C  (high = +81.0°C, crit = +95.0°C)
Core 11:       +81.0°C  (high = +81.0°C, crit = +95.0°C)
Core 12:       +78.0°C  (high = +81.0°C, crit = +95.0°C)

#27 Re: Optimized Apps for Projects

Posted: Sun Feb 17, 2019 10:55 pm
by Bryan
The V3 and V4s have AVX2 and they down clock when running AVX2 due to the temperature rise. The V3s have a bug that causes all cores to down clock if any of them are on AVX2. The V4s will only down clock the cores that are running AVX2.