On Tue, Feb 2, 2016 at 2:17 AM, Bill Hart <[email protected]> wrote: > I have a paid SMC account, but only the cheapo option. Not sure if that's > relevant. >
Your cheapo SMC subscription provides network access upgrades for *20* projects. -- William > > On Tuesday, 2 February 2016 11:07:43 UTC+1, Dima Pasechnik wrote: >> >> >> >> On Tuesday, 2 February 2016 09:12:39 UTC, Bill Hart wrote: >>> >>> I don't know if a VM would cause issued for the timing or not. Given that >>> the timings need to be cycle accurate, it could be an issue. It's not just >>> tuning we are doing, but superoptimisation. >>> >>> At present I'm not able to wget or git clone anything from my account on >>> there anyway. >> >> >> you cannot do this from a free account, due to e.g. DDOS attacks launched >> from free SMC accounts in the past. >> We can get you ODK-paid accounts if you like. >> >> >> >>> >>> So I will probably not try to use SMC for that. It's not really its >>> intended purpose. >>> >>> Bill. >>> >>> On Monday, 1 February 2016 16:37:13 UTC+1, William wrote: >>>> >>>> On Mon, Feb 1, 2016 at 7:32 AM, Dima Pasechnik <[email protected]> wrote: >>>> > Are Intel(R) Xeon(R) CPU @ 2.30GHz interesting for you? >>>> > (That's what one gets on SMC. We have some SMC licenses here bought >>>> > for ODK) >>>> > No idea if one could get something better there, but you could ask >>>> > William. >>>> >>>> It's a Virtual Machine, so may or may not be useful for their tuning. >>>> The main plus is that it is a Haswell architecture, which is pretty >>>> new: >>>> >>>> $ cat /proc/cpuinfo >>>> processor : 0 >>>> vendor_id : GenuineIntel >>>> cpu family : 6 >>>> model : 63 >>>> model name : Intel(R) Xeon(R) CPU @ 2.30GHz >>>> stepping : 0 >>>> microcode : 0x1 >>>> cpu MHz : 2299.968 >>>> cache size : 46080 KB >>>> physical id : 0 >>>> siblings : 1 >>>> core id : 0 >>>> cpu cores : 1 >>>> apicid : 0 >>>> initial apicid : 0 >>>> fpu : yes >>>> fpu_exception : yes >>>> cpuid level : 13 >>>> wp : yes >>>> flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge >>>> mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb >>>> rdtscp lm constant_t >>>> sc nopl xtopology eagerfpu pni pclmulqdq ssse3 fma cx16 sse4_1 sse4_2 >>>> x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm >>>> fsgsbase bmi1 avx2 sme >>>> p bmi2 xsaveopt >>>> bugs : >>>> bogomips : 4599.93 >>>> clflush size : 64 >>>> cache_alignment : 64 >>>> address sizes : 46 bits physical, 48 bits virtual >>>> power management: >>>> >>>> > Dima >>>> > >>>> > >>>> > On Monday, 1 February 2016 11:43:05 UTC, Bill Hart wrote: >>>> >> >>>> >> Hi all, >>>> >> >>>> >> today Alex Best started in Kaiserslautern on this OpenDreamKit >>>> >> project. >>>> >> He's working on support in MPIR for the latest Intel and AMD >>>> >> processors. >>>> >> However, we have very little access to anything later than about >>>> >> 2010. >>>> >> >>>> >> Does anyone have more recent Intel and AMD machines (linux servers) >>>> >> that >>>> >> they could provide log in access to? We'd be doing timings of all the >>>> >> basic >>>> >> MPIR functions and eventually superoptimising the assembly language >>>> >> functions. >>>> >> >>>> >> In general, one can expect up to a 2x speedup for each different >>>> >> microarchitecture (code from previous arches is rarely anywhere near >>>> >> optimal >>>> >> for more recent ones). E.g. we just tried some code which is roughly >>>> >> optimal >>>> >> for bobcat on a piledriver and it was two times slower than the old >>>> >> K10 code >>>> >> we have! >>>> >> >>>> >> If you can help, please contact me off list. >>>> >> >>>> >> Bill. >>>> >> >>>> >> P.S: We are aware of the GCC farm, but almost everything there we >>>> >> have >>>> >> access to elsewhere. >>>> > >>>> > -- >>>> > You received this message because you are subscribed to the Google >>>> > Groups >>>> > "sage-devel" group. >>>> > To unsubscribe from this group and stop receiving emails from it, send >>>> > an >>>> > email to [email protected]. >>>> > To post to this group, send email to [email protected]. >>>> > Visit this group at https://groups.google.com/group/sage-devel. >>>> > For more options, visit https://groups.google.com/d/optout. >>>> >>>> >>>> >>>> -- >>>> William (http://wstein.org) > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To post to this group, send email to [email protected]. > Visit this group at https://groups.google.com/group/sage-devel. > For more options, visit https://groups.google.com/d/optout. -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
