Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


AMD users should immediately switch to the latest Linux 5.11 kernel - Page 2
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

AMD users should immediately switch to the latest Linux 5.11 kernel

2»

Comments

  • jsgjsg Member, Resident Benchmarker

    @marvel

    Interesting news (for some) - but please do not word the title in a way that makes many think there is a vulnerability that needs immediate attention when in fact it's just about some performance boost.

    As for the boost, thanks no, I won't play the funny linux lottery. I'm shocked sometimes when I see how linux bazaar development works out and how unreliable its result is. Unless one has a rather basic system upgrading ones distro or kernel risks to basically but a lottery.

    Thanked by 1vedran
  • only on the node? if hypervisor old kernel, any boost for KVM VM?

  • @miroc said:
    Does anyone use CentOS?

    Dafuq question is that?

  • @TimboJones said:

    @miroc said:
    Does anyone use CentOS?

    Dafuq question is that?

    Ignore clowns.

  • @momkin said:

    @Daniel15 said:

    @momkin said:
    I can't believe it , before the upgrade my Ryzen 9 3900 max GHZ was alwayse stuck at 3.9 GHZ , now after the upgrade it can even reach 5.6 GHZ !

    Is that a joke or what ?

    AMD's site says the Ryzen 9 3900 boosts to 4.3GHz max... I wonder how you're reaching 5.6Ghz 🤔

    I am also surprised, i don't have any idea how it goes to 5.4GHZ while the AMD website says maximum is Up to 4.3GHz :neutral:

    processor : 17

    vendor_id : AuthenticAMD
    cpu family : 23
    model : 113
    model name : AMD Ryzen 9 3900 12-Core Processor
    stepping : 0
    microcode : 0x8701021
    cpu MHz : 5279.687
    cache size : 512 KB
    physical id : 0
    siblings : 24
    core id : 6
    cpu cores : 12
    apicid : 13
    initial apicid : 13
    fpu : yes
    fpu_exception : yes
    cpuid level : 16
    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 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate sme ssbd mba sev ibpb stibp vmmcall sev_es fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif umip rdpid overflow_recov succor smca
    bugs : sysret_ss_attrs spectre_v1 spectre_v2 spec_store_bypass
    bogomips : 6187.96
    TLB size : 3072 4K pages
    clflush size : 64
    cache_alignment : 64
    address sizes : 43 bits physical, 48 bits virtual
    power management: ts ttp tm hwpstate cpb eff_freq_ro [13] [14]

    Need benchmark to confirm the increased performance.
    Could you compare geekbench 5?

  • @jlay said:
    wat

    Model name:                      AMD Ryzen Threadripper 3960X 24-Core Processor
    Stepping:                        0
    Frequency boost:                 enabled
    CPU MHz:                         5785.351
    CPU max MHz:                     6935.7422
    

    Race to the 10ghz?

    Thanked by 1jlay
  • jsgjsg Member, Resident Benchmarker

    @serv_ee said:
    Race to the 10ghz?

    Why stop there? Why not aiming right away at 100 GHz?

  • @notarobo said:
    only on the node? if hypervisor old kernel, any boost for KVM VM?

    That would interest me as well. Is it worth it to risk a rock-solid battle hardned 4.19 kernel for those juicy extra % if the host remains on ancient tech?

  • serv_eeserv_ee Member
    edited February 2021

    @jsg said:

    @serv_ee said:
    Race to the 10ghz?

    Why stop there? Why not aiming right away at 100 GHz?

    Once nodes are done it's either that, "moar cores" or graphite which seems to have died off lately.

    (As a side note and as a personal opinion - no ARM is not the future. Intel once thought x86 is the next best thing after sliced bread but come AMD...people really think they will just sit on their asses and wait for ARM to take over? hah.)

  • jsgjsg Member, Resident Benchmarker

    @serv_ee said:

    @jsg said:

    @serv_ee said:
    Race to the 10ghz?

    Why stop there? Why not aiming right away at 100 GHz?

    Once nodes are done it's either that, "moar cores" or graphite which seems to have died off lately.

    (As a side note and as a personal opinion - no ARM is not the future. Intel once thought x86 is the next best thing after sliced bread but come AMD...people really think they will just sit on their asses and wait for ARM to take over? hah.)

    My remark was pure sarcasm.

  • GHz to the moon babay! 🚀🚀🚀🚀

  • @seriesn said:
    I hope it doesn’t break LVM and libvirt again

    That is why you should ignore "immediately switch" instructions that are for a performance benefit. Let other people gamma test it first...

    Different for updates purely for actively exploited or easily exploitable (so will be actively exploited in no time at all) security issues though.

  • @MeAtExampleDotCom said:

    @seriesn said:
    I hope it doesn’t break LVM and libvirt again

    That is why you should ignore "immediately switch" instructions that are for a performance benefit. Let other people gamma test it first...

    Different for updates purely for actively exploited or easily exploitable (so will be actively exploited in no time at all) security issues though.

    Oh yeah! Facts. That's what homelabs are for!

  • mirocmiroc Member
    edited February 2021

    @seriesn said:

    @MeAtExampleDotCom said:

    @seriesn said:
    I hope it doesn’t break LVM and libvirt again

    That is why you should ignore "immediately switch" instructions that are for a performance benefit. Let other people gamma test it first...

    Different for updates purely for actively exploited or easily exploitable (so will be actively exploited in no time at all) security issues though.

    Oh yeah! Facts. That's what homelabs are for!

    Talking about homelabs: I have an HP dl360e gen8 at home but being a 1u server its so loud i can't even use it. :D not sure how others can bear it

  • Geekbench anyone on 5.11?

  • emreemre Member, LIR

    lol, just checked out , my score is unbelievable!

    I guess something wrong with geekbench and 5.11

    no way my pc can be that fast...

  • Waiting for reports of CPUs bursting in flames. :)

  • I'm surprised more people aren't running Clear Linux OS. Especially, since they optimize for Intel which helps all those old machines run better, but the optimization they do often help AMD big core stuff as well.

    It is somewhat surprising to see people surprised you need current software to work with current hardware.

  • @emre said:

    lol, just checked out , my score is unbelievable!

    I guess something wrong with geekbench and 5.11

    no way my pc can be that fast...

    Do you have an old score (link) that we can compare with?

  • eva2000eva2000 Veteran
    edited February 2021

    @hackerman said: Does anyone use 5.x kernel on CentOS 7?

    I do on some of my Centmin Mod LEMP stack servers with CentOS 7 and 5.x kernels. Also Linode has it's own custom 5.x kernels on offer too. Tested some Linode AMD EPYC with 5.x way back for VPS comparison at https://community.centminmod.com/threads/13-way-vps-server-benchmark-comparison-tests-upcloud-vs-digitalocean-vs-linode-vs-vultr-vs-hetzner.17742/

    folks with insane high cpuinfo/lscpu care to share the output for command cpupower monitor ?

    On an AMD EPYC 7452 32C/64T with CentOS 7 with 3.10 kernel output shows correct frequency between idle and base 2.35 and turbo 3.35Ghz

    cpupower monitor
        |Mperf               || Idle_Stats         
    CPU | C0   | Cx   | Freq || POLL | C1   | C2   
       0| 13.43| 86.57|  1476||  0.00| 26.48| 59.95
      32|  1.64| 98.36|  1482||  0.00|  2.61| 95.18
       1|  6.69| 93.31|  1763||  0.00| 23.46| 71.16
      33|  6.91| 93.09|  1749||  0.00|  1.54| 90.85
       2|  4.33| 95.67|  3226||  0.00| 24.32| 72.93
      34|  0.78| 99.22|  3329||  0.00|  1.25| 97.39
       3|  5.08| 94.92|  1461||  0.00| 22.44| 73.12
      35|  0.08| 99.92|  1464||  0.00|  0.76| 98.61
       4| 23.47| 76.53|  1871||  0.05| 19.77| 56.37
      36|  5.32| 94.68|  1756||  0.00| 11.77| 82.61
       5| 16.09| 83.91|  1490||  0.00| 31.58| 51.90
      37|  7.22| 92.78|  1503||  0.00| 11.22| 81.46
       6| 32.23| 67.77|  2459||  0.00| 30.68| 36.25
      38|  5.57| 94.43|  1983||  0.00|  6.37| 87.41
       7| 12.33| 87.67|  1460||  0.00| 24.64| 63.15
      39|  3.43| 96.57|  1497||  0.00|  3.25| 92.70
       8| 19.80| 80.20|  1568||  0.00| 31.08| 50.58
      40|  5.86| 94.14|  1500||  0.00|  5.92| 88.48
       9|  7.97| 92.03|  1654||  0.03| 10.43| 82.58
      41|  2.13| 97.87|  1670||  0.00|  2.39| 95.90
      10|  5.65| 94.35|  1937||  0.00| 14.82| 79.65
      42|  0.83| 99.17|  1994||  0.00|  0.71| 97.92
      11| 20.62| 79.38|  1892||  0.00| 20.57| 57.74
      43|  0.29| 99.71|  1334||  0.00|  0.49| 98.08
      12|  8.91| 91.09|  3281||  0.00|  4.32| 87.35
      44|  0.01| 99.99|  2941||  0.00|  0.00| 98.85
      13|  3.20| 96.80|  1495||  0.00| 10.26| 85.55
      45|  8.27| 91.73|  1527||  0.00| 18.20| 73.26
      14|  2.13| 97.87|  1490||  0.00| 11.94| 85.90
      46|  1.14| 98.86|  1481||  0.00| 58.52| 39.25
      15|  0.48| 99.52|  1489||  0.00|  2.09| 96.82
      47|  0.00|100.00|  1491||  0.00|  0.00| 98.86
      16|  2.70| 97.30|  1497||  0.08| 35.14| 61.08
      48|  0.31| 99.69|  1497||  0.00|  1.34| 97.20
      17|  1.90| 98.10|  1496||  0.00| 16.73| 80.27
      49|  0.02| 99.98|  1498||  0.00|  0.00| 98.84
      18|  2.10| 97.90|  1496||  0.00| 31.94| 66.74
      50|  0.89| 99.11|  1497||  0.00| 10.74| 88.37
      19|  0.52| 99.48|  1497||  0.00|  5.97| 94.21
      51|  0.01| 99.99|  1499||  0.00|  0.00| 98.85
      20| 20.56| 79.44|  1728||  0.00| 13.73| 64.13
      52|  1.48| 98.52|  1807||  0.00|  2.09| 95.27
      21|  5.69| 94.31|  1686||  0.00| 12.63| 81.68
      53|  1.37| 98.63|  1587||  0.00| 16.38| 82.40
      22|  3.14| 96.86|  2977||  0.00| 14.23| 83.19
      54|  0.01| 99.99|  2342||  0.00|  0.00| 99.15
      23| 10.71| 89.29|  1605||  0.00| 20.92| 68.22
      55|  1.65| 98.35|  1719||  0.01|  3.36| 94.12
      24|  1.12| 98.88|  1489||  0.00|  3.72| 95.62
      56|  0.16| 99.84|  1497||  0.00|  0.57| 98.42
      25|  2.08| 97.92|  1476||  0.00| 10.73| 87.30
      57|  0.71| 99.29|  1492||  0.00| 14.33| 84.19
      26| 16.99| 83.01|  1739||  0.02|  4.91| 77.88
      58|  0.00|100.00|  1612||  0.00|  0.00| 99.15
      27|  1.65| 98.35|  1448||  0.00| 11.07| 86.73
      59|  0.00|100.00|  1440||  0.00|  0.00| 99.15
      28|  1.95| 98.05|  1498||  0.00|  2.76| 95.07
      60|  3.15| 96.85|  1810||  0.00|  0.18| 96.79
      29|  0.45| 99.55|  1497||  0.00|  0.75| 98.26
      61|  0.00|100.00|  1443||  0.00|  0.00|100.00
      30|  0.55| 99.45|  1495||  0.00|  0.00| 98.89
      62|  0.01| 99.99|  1470||  0.00|  0.00| 99.99
      31|  0.62| 99.38|  1497||  0.00|  2.11| 96.72
      63|  0.25| 99.75|  1497||  0.00|  0.57| 99.18
    
    Thanked by 1hackerman
Sign In or Register to comment.