SPEC CPU®2026 Floating Point Speed Result

Copyright 2026 Standard Performance Evaluation Corporation

GIGA-BYTE TECHNOLOGY CO., LTD. (Test Sponsor: Ampere Computing, Inc.)

R2A3-T40-AAV1
(3.20 GHz AmpereOneM A192-32M)

SPECspeed®2026_fp_base = 5.59

SPECspeed®2026_fp_energy_base = 2.58

SPECspeed®2026_fp_peak = 5.69

SPECspeed®2026_fp_energy_peak = 2.62
CPU2026 License: 6412 Test Date: Feb-2026
Test Sponsor: Ampere Computing, Inc. Hardware Availability: Nov-2025
Tested by: Ampere Computing, Inc. Software Availability: Oct-2025
Benchmark results graph
Hardware
CPU Name: AmpereOneM A192-32M
  Max MHz: 3200
  Nominal: 3200
Enabled: 192 cores, 1 chip
Orderable: 1 chip
Cache L1: 16 KB I + 64 KB D on chip per core
  L2: 2 MB I+D on chip per core
  L3: 64 MB I+D on chip per chip
  Other: None
Memory: 768 GB (12 x 64 GB 2Rx4 PC5-5600B-R)
Storage: 1 x 960 GB Samsung M.2 (MZ1L2960HCJR-00A07)
Cooling: Air
Other: None
Software
OS: Fedora Linux 43 (Server Edition)
6.17.5-200.64K_PS.fc42.aarch64
Compiler: C/C++/Fortran: Version 15.2.0 of GCC
Compiler Category: Community
Firmware: Version 5.4.5.1 released Oct-2025
File System: xfs
System State: Run level 3 (multi-user)
Base Pointers: 64-bit
Peak Pointers: 64-bit
Other: Jemalloc memory allocator library v5.3.0
Power Management: OS CPU governor set to "performance"
Power
Max. Power (W): 652.6
Idle Power (W): 200.2
Min. Temperature (C): 20.56
Elevation (m): 60
Line Standard: 120 V / 60 Hz / 1 phase / 2 wire
Provisioning: Line powered
Power Settings
Management FW: Version 4.06 of MegaRAC BMC
Memory Mode: Normal
Power-Relevant Hardware
Power Supply: 1 x 1600 W (non-redundant)
  Details: 1+1 1600W 80 PLUS Titanium redundant power
supplies
AC Power Supply
Backplane: PCIe Gen5 x4
Other Storage: Embedded SATA Controller
Storage Model #s: R2A3-T40-AAV1
NICs Installed: 1 x Mellanox Technologies MT27710 Family @ 25 GbE
(2 ports ethernet)
NICs Enabled (FW/OS): 2 / 1
NICs Connected/Speed: 1 @ 25 Gbps
Other HW Model #s: Fan configuration is 4 x 80x80x38mm
Power Analyzer
Power Analyzer: cpu-reference-ptd:8000
Hardware Vendor: Yokogawa
Model: WT-310
Serial Number: T11733285
Input Connection: Serial over USB
Metrology Institute: NIST
Calibration By: Yokogawa USA
Calibration Label: T126622
Calibration Date: 13-Aug-2025
PTDaemon® Version: 1.11.3 (0c074d7d; 2025-10-15)
Setup Description: Directly Connected to PSU 1
Current Ranges Used: 10A
Voltage Range Used: 150V
Temperature Meter
Temperature Meter: cpu-reference-ptd:9000
Hardware Vendor: PCSensor
Model: USB9097+DS18B20
Serial Number: N/A
Input Connection: USB
PTDaemon Version: 1.11.3 (0c074d7d; 2025-10-15)
Setup Description: In front of SUT front panel primary air inlet

Base Results Table

Benchmark Threads Seconds Ratio Energy
(kJ)		 Energy
Ratio		 Average
Power		 Maximum
Power		 Seconds Ratio Energy
(kJ)		 Energy
Ratio		 Average
Power		 Maximum
Power		 Seconds Ratio Energy
(kJ)		 Energy
Ratio		 Average
Power		 Maximum
Power
800.pot3d_s 192 443 1.52 1350 0.937 304 307 440 1.53 1340 0.943 304 307 435 1.55 1320 0.956 304 306
803.sph_exa_s 192 105 11.80 43.0 4.870 410 543 106 11.70 43.0 4.870 407 538 105 11.80 43.0 4.870 411 528
809.cactus_s 192 190 5.89 60.9 3.330 320 351 192 5.86 61.4 3.300 321 353 193 5.81 61.9 3.270 320 355
811.tealeaf_s 192 201 2.77 83.4 1.670 415 563 219 2.54 90.3 1.550 412 552 234 2.38 95.4 1.460 408 550
816.nab_s 192 137 8.19 51.0 3.790 371 395 137 8.20 50.9 3.800 370 396 137 8.21 50.8 3.800 371 394
820.cloverleaf_s 192 128 6.68 66.3 2.510 517 550 127 6.75 65.8 2.530 519 553 129 6.62 66.2 2.510 512 539
822.palm_s 192 272 4.51 91.9 2.260 338 473 269 4.57 90.7 2.290 337 490 274 4.48 92.1 2.250 336 465
849.fotonik3d_s 192 131 5.04 62.9 1.960 481 505 131 5.05 62.8 1.960 480 504 131 5.05 63.2 1.950 484 506
857.namd_s 192 225 6.45 87.5 3.040 389 481 226 6.43 87.7 3.030 389 483 226 6.44 87.9 3.020 390 483
865.roms_s 192 116 9.44 63.7 3.640 551 577 116 9.39 64.1 3.610 552 577 116 9.39 64.1 3.610 552 578
867.nest_s 192 184 11.70 80.3 4.820 436 653 172 12.50 76.6 5.050 444 643 171 12.60 75.8 5.100 443 637
872.marian_s 192 197 5.48 64.2 2.510 325 574 196 5.51 63.7 2.530 325 575 197 5.49 63.8 2.530 324 563
881.neutron_s 192 220 3.70 79.8 1.650 362 553 221 3.69 80.2 1.640 363 555 220 3.71 79.9 1.650 363 556
SPECspeed®2026_fp_base 5.59
SPECspeed®2026_fp_energy_base 2.58
Results appear in the order in which they were run. Bold underlined text indicates a median measurement.

Peak Results Table

Benchmark Threads Seconds Ratio Energy
(kJ)		 Energy
Ratio		 Average
Power		 Maximum
Power		 Seconds Ratio Energy
(kJ)		 Energy
Ratio		 Average
Power		 Maximum
Power		 Seconds Ratio Energy
(kJ)		 Energy
Ratio		 Average
Power		 Maximum
Power
800.pot3d_s 192 443 1.52 1350 0.937 304 307 440 1.53 1340 0.943 304 307 435 1.55 1320 0.956 304 306
803.sph_exa_s 192 105 11.80 43.0 4.870 410 543 106 11.70 43.0 4.870 407 538 105 11.80 43.0 4.870 411 528
809.cactus_s 192 189 5.94 60.7 3.330 322 356 186 6.03 60.3 3.360 324 355 187 5.99 60.3 3.360 322 356
811.tealeaf_s 192 199 2.79 83.2 1.680 417 580 230 2.42 94.6 1.480 411 560 185 3.01 78.1 1.790 422 577
816.nab_s 192 137 8.19 51.0 3.790 371 395 137 8.20 50.9 3.800 370 396 137 8.21 50.8 3.800 371 394
820.cloverleaf_s 192 126 6.79 65.3 2.550 518 540 127 6.75 66.0 2.520 519 550 127 6.74 66.1 2.520 519 551
822.palm_s 192 269 4.56 91.6 2.260 340 494 264 4.65 89.7 2.310 340 477 274 4.49 92.9 2.230 339 504
849.fotonik3d_s 192 131 5.04 62.9 1.960 481 505 131 5.05 62.8 1.960 480 504 131 5.05 63.2 1.950 484 506
857.namd_s 192 220 6.61 86.4 3.080 393 490 219 6.63 86.5 3.070 395 488 218 6.65 86.0 3.090 394 490
865.roms_s 192 117 9.33 63.9 3.620 547 574 116 9.41 64.0 3.610 553 577 116 9.40 63.8 3.630 550 573
867.nest_s 192 162 13.40 71.5 5.410 442 651 161 13.50 71.2 5.430 444 633 166 13.00 73.3 5.270 443 643
872.marian_s 192 197 5.48 64.2 2.510 325 574 196 5.51 63.7 2.530 325 575 197 5.49 63.8 2.530 324 563
881.neutron_s 192 220 3.70 79.8 1.650 362 553 221 3.69 80.2 1.640 363 555 220 3.71 79.9 1.650 363 556
SPECspeed®2026_fp_peak 5.69
SPECspeed®2026_fp_energy_peak 2.62
Results appear in the order in which they were run. Bold underlined text indicates a median measurement.

Compiler Notes

 Binaries were compiled on a system with 1x AmpereOneM CPU
 chip + 768 GB Memory using Fedora Linux 42

Operating System Notes

 'ulimit -s unlimited' was used to set environment stack size
 Set dirty_ratio=8 to limit dirty cache to 8% of memory
      echo 8 | sudo tee /proc/sys/vm/dirty_ratio
 Set swappiness=1 to swap only if necessary
      echo 1 | sudo tee /proc/sys/vm/swappiness
 Set zone_reclaim_mode=1 to free local node memory and avoid remote memory
      echo 1 | sudo tee /proc/sys/vm/zone_reclaim_mode
 Set drop_caches=3 to reset caches before invoking runcpu
      echo 3 | sudo tee /proc/sys/vm/drop_caches
 Set numa_balancing=0 to disable automatic numa balancing
      echo 0 | sudo tee /proc/sys/kernel/numa_balancing
 Switch off all ktune and tuned settings
      sudo tuned-adm off
 Transparent huge pages set to 'never'
      sudo bash -c "echo never > /sys/kernel/mm/transparent_hugepage/enabled"

 runcpu command invoked through numactl i.e.
 numactl --interleave=all runcpu <etc>

Environment Variables Notes

Environment variables set by runcpu before the start of the run:
LD_LIBRARY_PATH =
     "/home/amptest/workspace/sir2026_pkg_verify/ampere_spec2026/spec2026/jem
     alloc/install/lib:/home/amptest/workspace/sir2026_pkg_verify/ampere_spec
     2026/spec2026/gcc/install/lib64:/home/amptest/ampere_spec2026/spec2026/j
     emalloc/install/lib:/home/amptest/ampere_spec2026/spec2026/gcc/install/l
     ib64"
MALLOC_CONF = "thp:never,dirty_decay_ms:-1,muzzy_decay_ms:-1,retain:true"
OMP_STACKSIZE = "120M"

General Notes

 Jemalloc v5.3.0 is available via
 https://github.com/jemalloc/jemalloc/releases/download/5.3.0/jemalloc-5.3.0.tar.bz2
 It was built on Fedora Linux 42 using Version 15.2.0 of GCC
 The configure options are
 "--with-lg-page=16" for building libjemalloc.so, and
 "--with-lg-quantum=3 --with-lg-page=16" for building libjemalloc_ext.so
 Tuned MALLOC_CONF in terms of https://jemalloc.net/jemalloc.3.html

  NA: The test sponsor attests, as of date of publication, that CVE-2017-5754 (Meltdown)
  is mitigated in the system as tested and documented.
  Yes: The test sponsor attests, as of date of publication, that CVE-2017-5753 (Spectre variant 1)
  is mitigated in the system as tested and documented.
  Yes: The test sponsor attests, as of date of publication, that CVE-2017-5715 (Spectre variant 2)
  is mitigated in the system as tested and documented.

Platform Notes

 Note: lscpu is not able to detect the SLC.
 SLC is defined at https://developer.arm.com/documentation/100180/0103/bry1436285730281

 Sysinfo program /home/amptest/workspace/sir2026_pkg_verify/ampere_spec2026/spec2026/bin/sysinfo
 Rev: 069f95da7e7f5d81b2ce48a82150e54f
 running on ampereone Thu Feb  5 20:38:32 2026

 SUT (System Under Test) info as seen by some common utilities.

 ------------------------------------------------------------
 Table of contents
 ------------------------------------------------------------
  1. uname -srvm
  2. w
  3. Username
  4. ulimit -a
  5. sysinfo process ancestry
  6. /proc/cpuinfo
  7. lscpu
  8. numactl --hardware
  9. /proc/meminfo
 10. Systemd service manager version: systemd 258 (258-1.fc43)
 11. Services, from systemctl list-unit-files
 12. Linux kernel boot-time arguments, from /proc/cmdline
 13. cpupower frequency-info
 14. tuned-adm active
 15. sysctl
 16. /sys/kernel/mm/transparent_hugepage
 17. /sys/kernel/mm/transparent_hugepage/khugepaged
 18. OS release
 19. Disk information
 20. /sys/devices/virtual/dmi/id
 21. dmidecode
 22. BIOS
 ------------------------------------------------------------

 ------------------------------------------------------------
 1. uname -srvm
   Linux 6.17.5-200.64K_PS.fc42.aarch64 #1 SMP PREEMPT_DYNAMIC Wed Jan 21 08:54:52 EST 2026 aarch64

 ------------------------------------------------------------
 2. w
    20:38:32 up  6:46,  2 users,  load average: 103.83, 40.91, 32.69
   USER     TTY        LOGIN@   IDLE   JCPU   PCPU WHAT
   amptest            13:52    6:46m  0.00s  0.02s sshd-session: amptest [priv]
   amptest            13:52    6:46m  0.00s  0.21s /usr/lib/systemd/systemd --user

 ------------------------------------------------------------
 3. Username
   From environment variable $USER:  root
   From the command 'logname':       amptest

 ------------------------------------------------------------
 4. ulimit -a
   real-time non-blocking time  (microseconds, -R) unlimited
   core file size              (blocks, -c) 0
   data seg size               (kbytes, -d) unlimited
   scheduling priority                 (-e) 0
   file size                   (blocks, -f) unlimited
   pending signals                     (-i) 784608
   max locked memory           (kbytes, -l) 8192
   max memory size             (kbytes, -m) unlimited
   open files                          (-n) 524288
   pipe size                (512 bytes, -p) 8
   POSIX message queues         (bytes, -q) 819200
   real-time priority                  (-r) 0
   stack size                  (kbytes, -s) unlimited
   cpu time                   (seconds, -t) unlimited
   max user processes                  (-u) 784608
   virtual memory              (kbytes, -v) unlimited
   file locks                          (-x) unlimited

 ------------------------------------------------------------
 5. sysinfo process ancestry
  /usr/lib/systemd/systemd --switched-root --system --deserialize=57
  sshd: /usr/sbin/sshd -D [listener] 0 of 10-100 startups
  sshd-session: amptest [priv]
  sshd-session: amptest@notty
  java -jar remoting.jar -workDir /home/amptest -jar-cache /home/amptest/remoting/jarCache
  /bin/bash /tmp/jenkins17214592009497033566.sh
  sudo -S -E UPDATE_UTIL=false /home/amptest/util/jenkins/speccpu_pkg_verify.sh
    GreenSIR2026/without_report/spec2026_fpspeed_gcc15_Ofast_static_ampereonem_power_112.tgz
  /bin/bash /home/amptest/util/jenkins/speccpu_pkg_verify.sh
    GreenSIR2026/without_report/spec2026_fpspeed_gcc15_Ofast_static_ampereonem_power_112.tgz
  /bin/bash /home/amptest/workspace/sir2026_pkg_verify/ampere_spec2026/run_spec2026.sh --iterations 3
    --nobuild --action run --threads 192 --reportable --tune=base,peak fpspeed
  runcpu --config=ampere_aarch64 --define numasize=192 --define
    gcc_dir=/home/amptest/ampere_spec2026/spec2026/gcc/install --define
    llvm_dir=/home/amptest/ampere_spec2026/spec2026/llvm/install --define
    jemalloc_dir=/home/amptest/ampere_spec2026/spec2026/jemalloc/install --define glibc_dir=/ --iterations 3
    --nobuild --action run --threads 192 --reportable --tune=base,peak fpspeed
  runcpu --configfile ampere_aarch64 --define numasize=192 --define
    gcc_dir=/home/amptest/ampere_spec2026/spec2026/gcc/install --define
    llvm_dir=/home/amptest/ampere_spec2026/spec2026/llvm/install --define
    jemalloc_dir=/home/amptest/ampere_spec2026/spec2026/jemalloc/install --define glibc_dir=/ --iterations 3
    --nobuild --action run --threads 192 --reportable --tune base,peak --runmode speed --tune base:peak --size
    refspeed fpspeed --nopreenv --note-preenv --logfile
    $SPEC/tmp/CPU2026.003/templogs/preenv.fpspeed.003.0.log --lognum 003.0 --from_runcpu 2
  specperl $SPEC/bin/sysinfo -f
 $SPEC = /home/amptest/workspace/sir2026_pkg_verify/ampere_spec2026/spec2026

 ------------------------------------------------------------
 6. /proc/cpuinfo
     CPU implementer : 0xc0
     CPU architecture: 8
     CPU variant     : 0x0
     CPU part        : 0xac4
     CPU revision    : 0
     Features        : fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt
                       fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 asimdfhm dit uscat ilrcpc flagm ssbs sb
                       paca pacg dcpodp flagm2 frint i8mm bf16 rng bti ecv

 ------------------------------------------------------------
 7. lscpu

 From lscpu from util-linux 2.41.3:
   Architecture:                            aarch64
   CPU op-mode(s):                          64-bit
   Byte Order:                              Little Endian
   CPU(s):                                  192
   On-line CPU(s) list:                     0-191
   Vendor ID:                               Ampere
   BIOS Vendor ID:                          Ampere (R)
   Model name:                              Ampere-1a
   BIOS Model name:                         AmpereOne (R) A192-32M CPU @ 3.2GHz
   BIOS CPU family:                         257
   Model:                                   0
   Thread(s) per core:                      1
   Core(s) per socket:                      192
   Socket(s):                               1
   Stepping:                                0x0
   Frequency boost:                         disabled
   CPU(s) scaling MHz:                      100%
   CPU max MHz:                             3200.0000
   CPU min MHz:                             1000.0000
   BogoMIPS:                                2000.00
   Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp
                                            cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512
                                            asimdfhm dit uscat ilrcpc flagm ssbs sb paca pacg dcpodp flagm2
                                            frint i8mm bf16 rng bti ecv
   L1d cache:                               12 MiB (192 instances)
   L1i cache:                               3 MiB (192 instances)
   L2 cache:                                384 MiB (192 instances)
   NUMA node(s):                            1
   NUMA node0 CPU(s):                       0-191
   Vulnerability Gather data sampling:      Not affected
   Vulnerability Ghostwrite:                Not affected
   Vulnerability Indirect target selection: Not affected
   Vulnerability Itlb multihit:             Not affected
   Vulnerability L1tf:                      Not affected
   Vulnerability Mds:                       Not affected
   Vulnerability Meltdown:                  Not affected
   Vulnerability Mmio stale data:           Not affected
   Vulnerability Old microcode:             Not affected
   Vulnerability Reg file data sampling:    Not affected
   Vulnerability Retbleed:                  Not affected
   Vulnerability Spec rstack overflow:      Not affected
   Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
   Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
   Vulnerability Spectre v2:                Not affected
   Vulnerability Srbds:                     Not affected
   Vulnerability Tsa:                       Not affected
   Vulnerability Tsx async abort:           Not affected
   Vulnerability Vmscape:                   Not affected

 From lscpu --cache:
      NAME ONE-SIZE ALL-SIZE WAYS TYPE        LEVEL SETS PHY-LINE COHERENCY-SIZE
      L1d       64K      12M    4 Data            1  256                      64
      L1i       16K       3M    4 Instruction     1   64                      64
      L2         2M     384M    8 Unified         2 4096                      64

 ------------------------------------------------------------
 8. numactl --hardware
 NOTE: a numactl 'node' might or might not correspond to a physical chip.
   available: 1 nodes (0)
   node 0 cpus: 0-191
   node 0 size: 784583 MB
   node 0 free: 774508 MB
   node distances:
   node     0
      0:   10

 ------------------------------------------------------------
 9. /proc/meminfo
    MemTotal:       803413888 kB

 'who -r' did not return a run level

 ------------------------------------------------------------
 10. Systemd service manager version: systemd 258 (258-1.fc43)
   Default Target  Status
   multi-user      running

 ------------------------------------------------------------
 11. Services, from systemctl list-unit-files
   STATE            UNIT FILES
   enabled          ModemManager NetworkManager NetworkManager-dispatcher abrt-journal-core abrt-oops
                    abrt-vmcore abrt-xorg abrtd atd audit-rules auditd authselect-apply-changes bluetooth
                    crond dbus-broker fips-crypto-policy-overlay firewalld getty@ irqbalance iscsi-onboot
                    iscsi-starter lm_sensors lvm2-monitor mdmonitor multipathd rpmdb-rebuild rsyslog
                    rtkit-daemon selinux-autorelabel-mark smartd sshd sssd systemd-boot-clear-sysfail
                    systemd-confext systemd-resolved systemd-sysext systemd-tpm2-clear tuned udisks2 upower
   enabled-runtime  systemd-remount-fs
   disabled         NetworkManager-wait-online abrt-pstoreoops arp-ethers blk-availability
                    canberra-system-bootup canberra-system-shutdown canberra-system-shutdown-reboot
                    chrony-wait chronyd chronyd-restricted console-getty containerd cpupower debug-shell
                    dnf-system-upgrade dnf5-offline-transaction dnsmasq docker fancontrol fsidd gssproxy
                    iscsi-init iscsid iscsiuio kvm_stat lastlog2-import low-memory-monitor lvm-devices-import
                    man-db-restart-cache-update netavark-dhcp-proxy netavark-firewalld-reload
                    netavark-nftables-reload nfs-blkmap nfs-server nftables nis-domainname
                    nvmefc-boot-connections nvmf-autoconnect openhpid podman podman-auto-update
                    podman-clean-transient podman-kube@ podman-restart psacct rmtfs rpcbind
                    selinux-check-proper-disable ssh-host-keys-migration sshd-keygen@ svnserve
                    systemd-boot-check-no-failures systemd-boot-update systemd-homed systemd-homed-activate
                    systemd-homed-firstboot systemd-network-generator systemd-oomd systemd-pcrlock-file-system
                    systemd-pcrlock-firmware-code systemd-pcrlock-firmware-config systemd-pcrlock-machine-id
                    systemd-pcrlock-make-policy systemd-pcrlock-secureboot-authority
                    systemd-pcrlock-secureboot-policy systemd-pstore systemd-time-wait-sync systemd-timesyncd
                    systemd-udev-load-credentials systemd-userdb-load-credentials wpa_supplicant zvbid
   indirect         iscsi pcscd serial-getty@ sshd@ sssd-autofs sssd-kcm sssd-nss sssd-pac sssd-pam sssd-ssh
                    sssd-sudo systemd-sysupdate systemd-sysupdate-reboot systemd-userdbd

 ------------------------------------------------------------
 12. Linux kernel boot-time arguments, from /proc/cmdline
   BOOT_IMAGE=(hd5,gpt2)/vmlinuz-6.17.5-200.64K_PS.fc42.aarch64
   root=/dev/mapper/fedora-root
   ro
   rd.lvm.lv=fedora/root
   nokaslr
   kaslr.disabled=1

 ------------------------------------------------------------
 13. cpupower frequency-info
   analyzing CPU 20:
     current policy: frequency should be within 1000 MHz and 3.20 GHz.
                     The governor "performance" may decide which speed to use
                     within this range.

 ------------------------------------------------------------
 14. tuned-adm active
   No current active profile.

 ------------------------------------------------------------
 15. sysctl
   kernel.numa_balancing               0
   kernel.randomize_va_space           2
   vm.compaction_proactiveness        20
   vm.dirty_background_bytes           0
   vm.dirty_background_ratio          10
   vm.dirty_bytes                      0
   vm.dirty_expire_centisecs        3000
   vm.dirty_ratio                      8
   vm.dirty_writeback_centisecs      500
   vm.dirtytime_expire_seconds     43200
   vm.extfrag_threshold              500
   vm.min_unmapped_ratio               1
   vm.nr_hugepages                     0
   vm.nr_hugepages_mempolicy           0
   vm.nr_overcommit_hugepages          0
   vm.swappiness                       1
   vm.watermark_boost_factor       15000
   vm.watermark_scale_factor          10
   vm.zone_reclaim_mode                1

 ------------------------------------------------------------
 16. /sys/kernel/mm/transparent_hugepage
   defrag          always defer defer+madvise madvise [never]
   enabled         always madvise [never]
   hpage_pmd_size  536870912
   shmem_enabled   always within_size advise [never] deny force

 ------------------------------------------------------------
 17. /sys/kernel/mm/transparent_hugepage/khugepaged
   alloc_sleep_millisecs   60000
   defrag                      1
   max_ptes_none            8191
   max_ptes_shared          4096
   max_ptes_swap            1024
   pages_to_scan           65536
   scan_sleep_millisecs    10000

 ------------------------------------------------------------
 18. OS release
   From /etc/*-release /etc/*-version
   os-release     Fedora Linux 43 (Server Edition)
   fedora-release Fedora release 43 (Forty Three)
   redhat-release Fedora release 43 (Forty Three)
   system-release Fedora release 43 (Forty Three)

 ------------------------------------------------------------
 19. Disk information
 SPEC is set to: /home/amptest/workspace/sir2026_pkg_verify/ampere_spec2026/spec2026
   Filesystem              Type  Size  Used Avail Use% Mounted on
   /dev/mapper/fedora-root xfs   891G  258G  634G  29% /

 ------------------------------------------------------------
 20. /sys/devices/virtual/dmi/id
     Vendor:         GIGACOMPUTING
     Product:        R2A3-T40-AAV1
     Product Family: R2A3-T40-AAV1
     Serial:         0000000000000000000000000001

 ------------------------------------------------------------
 21. dmidecode
   Additional information from dmidecode 3.6 follows.  WARNING: Use caution when you interpret this section.
   The 'dmidecode' program reads system data which is "intended to allow hardware to be accurately
   determined", but the intent may not be met, as there are frequent changes to hardware, firmware, and the
   "DMTF SMBIOS" standard.
   Memory:
     12x Micron Technology MTC40F2046S1RC56BD2 MLCC 64 GB 2 rank 5600


 ------------------------------------------------------------
 22. BIOS
 (This section combines info from /sys/devices and dmidecode.)
    BIOS Vendor:       Ampere Computing LLC
    BIOS Version:      5.4.5.1
    BIOS Date:         10/22/2025
    BIOS Revision:     5.37

Compiler Version Notes

==============================================================================
C       | 811.tealeaf_s(base, peak) 816.nab_s(base, peak) 881.neutron_s(base,
        | peak)
------------------------------------------------------------------------------
gcc (GCC) 15.2.0
Copyright (C) 2025 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
------------------------------------------------------------------------------

==============================================================================
C++     | 803.sph_exa_s(base, peak) 857.namd_s(base, peak) 867.nest_s(base,
        | peak) 872.marian_s(base, peak)
------------------------------------------------------------------------------
g++ (GCC) 15.2.0
Copyright (C) 2025 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
------------------------------------------------------------------------------

==============================================================================
C++, C  | 809.cactus_s(base pass 0, peak pass 1, peak pass 2)
------------------------------------------------------------------------------
g++ (GCC) 15.2.0
Copyright (C) 2025 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
------------------------------------------------------------------------------

==============================================================================
C++, C  | 809.cactus_s(base pass 0, peak pass 1, peak pass 2)
------------------------------------------------------------------------------
gcc (GCC) 15.2.0
Copyright (C) 2025 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
------------------------------------------------------------------------------

==============================================================================
C++, C  | 809.cactus_s(base pass 0, peak pass 1, peak pass 2)
------------------------------------------------------------------------------
g++ (GCC) 15.2.0
Copyright (C) 2025 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
------------------------------------------------------------------------------

==============================================================================
C++, C  | 809.cactus_s(base pass 0, peak pass 1, peak pass 2)
------------------------------------------------------------------------------
gcc (GCC) 15.2.0
Copyright (C) 2025 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
------------------------------------------------------------------------------

==============================================================================
Fortran | 800.pot3d_s(base, peak) 820.cloverleaf_s(base, peak)
        | 822.palm_s(base, peak) 849.fotonik3d_s(base, peak) 865.roms_s(base,
        | peak)
------------------------------------------------------------------------------
GNU Fortran (GCC) 15.2.0
Copyright (C) 2025 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
------------------------------------------------------------------------------

Base Compiler Invocation

C benchmarks:

 gcc 

C++ benchmarks:

 g++ 

Fortran benchmarks:

 gfortran 

Benchmarks using both C and C++:

 g++   gcc 

Base Portability Flags

800.pot3d_s:  -DSPEC_SUPPRESS_LOCAL_AND_REDUCE 
867.nest_s:  -fno-finite-math-only 

Base Optimization Flags

C benchmarks:

 -mabi=lp64   -std=c18   -Wl,-allow-multiple-definition   -static   -fwhole-program   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib64   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib   -L/home/amptest/ampere_spec2026/spec2026/jemalloc/install/lib   -g   -Ofast   -mcpu=ampere1a   -flto=32   -funroll-loops   --param early-inlining-insns=96   --param max-inline-insns-auto=64   --param inline-unit-growth=96   -fopenmp   -DSPEC_OPENMP   -u malloc   -ljemalloc 

C++ benchmarks:

803.sph_exa_s:  -mabi=lp64   -std=c++17   -Wl,-allow-multiple-definition   -static   -fwhole-program   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib64   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib   -L/home/amptest/ampere_spec2026/spec2026/jemalloc/install/lib   -g   -Ofast   -mcpu=ampere1a   -flto=32   -funroll-loops   --param early-inlining-insns=256   --param max-inline-insns-auto=128   --param inline-unit-growth=256   -ffinite-loops   -fopenmp   -DSPEC_OPENMP   -u malloc   -ljemalloc_ext 
857.namd_s:  Same as 803.sph_exa_s 
867.nest_s:  Same as 803.sph_exa_s 
872.marian_s:  -mabi=lp64   -std=c++17   -Wl,-allow-multiple-definition   -static   -fwhole-program   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib64   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib   -L/home/amptest/ampere_spec2026/spec2026/jemalloc/install/lib   -g   -Ofast   -mcpu=ampere1a   -flto=32   -funroll-loops   --param early-inlining-insns=256   --param max-inline-insns-auto=128   --param inline-unit-growth=256   -ffinite-loops   -pthread   -u malloc   -ljemalloc_ext 

Fortran benchmarks:

800.pot3d_s:  -mabi=lp64   -std=f2008   -Wl,-allow-multiple-definition   -static   -fwhole-program   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib64   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib   -L/home/amptest/ampere_spec2026/spec2026/jemalloc/install/lib   -g   -Ofast   -mcpu=ampere1a   -flto=32   -funroll-loops   -fallow-argument-mismatch   -ftree-parallelize-loops=16   -u malloc   -ljemalloc 
820.cloverleaf_s:  -mabi=lp64   -std=f2008   -Wl,-allow-multiple-definition   -static   -fwhole-program   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib64   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib   -L/home/amptest/ampere_spec2026/spec2026/jemalloc/install/lib   -g   -Ofast   -mcpu=ampere1a   -flto=32   -funroll-loops   -fallow-argument-mismatch   -DSPEC_OPENMP   -fopenmp   -u malloc   -ljemalloc 
822.palm_s:  Same as 820.cloverleaf_s 
849.fotonik3d_s:  Same as 820.cloverleaf_s 
865.roms_s:  Same as 820.cloverleaf_s 

Benchmarks using both C and C++:

 -mabi=lp64   -std=c++17   -std=c18   -Wl,-allow-multiple-definition   -static   -fwhole-program   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib64   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib   -L/home/amptest/ampere_spec2026/spec2026/jemalloc/install/lib   -g   -Ofast   -mcpu=ampere1a   -flto=32   -funroll-loops   --param early-inlining-insns=96   --param max-inline-insns-auto=64   --param inline-unit-growth=96   --param early-inlining-insns=256   --param max-inline-insns-auto=128   --param inline-unit-growth=256   -ffinite-loops   -fopenmp   -DSPEC_OPENMP   -u malloc   -ljemalloc_ext 

Base Other Flags

C benchmarks:

 -w   -Wl,-Map,mapfile 

C++ benchmarks:

 -Wl,-Map,mapfile 

Fortran benchmarks:

 -Wl,-Map,mapfile 

Benchmarks using both C and C++:

 -w   -Wl,-Map,mapfile 

Peak Compiler Invocation

C benchmarks:

 gcc 

C++ benchmarks:

 g++ 

Fortran benchmarks:

 gfortran 

Benchmarks using both C and C++:

 g++   gcc 

Peak Portability Flags

800.pot3d_s:  -DSPEC_SUPPRESS_LOCAL_AND_REDUCE 
867.nest_s:  -fno-finite-math-only 

Peak Optimization Flags

C benchmarks:

811.tealeaf_s:  -mabi=lp64   -std=c18   -Wl,-allow-multiple-definition   -static   -fwhole-program   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib64   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib   -L/home/amptest/ampere_spec2026/spec2026/jemalloc/install/lib   -fprofile-generate   -fprofile-use   -fprofile-partial-training   -g   -Ofast   -mcpu=ampere1a   -flto=32   -funroll-loops   --param early-inlining-insns=96   --param max-inline-insns-auto=64   --param inline-unit-growth=96   -fopenmp   -DSPEC_OPENMP   -u malloc   -ljemalloc 
816.nab_s:  basepeak = yes 
881.neutron_s:  basepeak = yes 

C++ benchmarks:

803.sph_exa_s:  basepeak = yes 
857.namd_s:  -mabi=lp64   -std=c++17   -Wl,-allow-multiple-definition   -static   -fwhole-program   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib64   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib   -L/home/amptest/ampere_spec2026/spec2026/jemalloc/install/lib   -fprofile-generate   -fprofile-use   -fprofile-partial-training   -g   -Ofast   -mcpu=ampere1a   -flto=32   -funroll-loops   --param early-inlining-insns=256   --param max-inline-insns-auto=128   --param inline-unit-growth=256   -ffinite-loops   -fopenmp   -DSPEC_OPENMP   -u malloc   -ljemalloc_ext 
867.nest_s:  Same as 857.namd_s 
872.marian_s:  basepeak = yes 

Fortran benchmarks:

800.pot3d_s:  basepeak = yes 
820.cloverleaf_s:  -mabi=lp64   -std=f2008   -Wl,-allow-multiple-definition   -static   -fwhole-program   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib64   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib   -L/home/amptest/ampere_spec2026/spec2026/jemalloc/install/lib   -fprofile-generate   -fprofile-use   -fprofile-partial-training   -g   -Ofast   -mcpu=ampere1a   -flto=32   -funroll-loops   -fallow-argument-mismatch   -DSPEC_OPENMP   -fopenmp   -u malloc   -ljemalloc 
822.palm_s:  Same as 820.cloverleaf_s 
849.fotonik3d_s:  basepeak = yes 
865.roms_s:  Same as 820.cloverleaf_s 

Benchmarks using both C and C++:

 -mabi=lp64   -std=c++17   -std=c18   -Wl,-allow-multiple-definition   -static   -fwhole-program   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib64   -L/home/amptest/ampere_spec2026/spec2026/gcc/install/lib   -L/home/amptest/ampere_spec2026/spec2026/jemalloc/install/lib   -fprofile-generate   -fprofile-use   -fprofile-partial-training   -g   -Ofast   -mcpu=ampere1a   -flto=32   -funroll-loops   --param early-inlining-insns=96   --param max-inline-insns-auto=64   --param inline-unit-growth=96   --param early-inlining-insns=256   --param max-inline-insns-auto=128   --param inline-unit-growth=256   -ffinite-loops   -fopenmp   -DSPEC_OPENMP   -u malloc   -ljemalloc_ext 

Peak Other Flags

C benchmarks:

 -w   -Wl,-Map,mapfile 

C++ benchmarks:

 -Wl,-Map,mapfile 

Fortran benchmarks:

 -Wl,-Map,mapfile 

Benchmarks using both C and C++:

 -w   -Wl,-Map,mapfile 

The flags files that were used to format this result can be browsed at
http://www.spec.org/cpu2026/results/flags/gcc-rev-A2.html,
http://www.spec.org/cpu2026/results/flags/GIGA-BYTE-platform-settings-AmpereOneM-rev.1.html.

You can also download the XML flags sources by saving the following links:
http://www.spec.org/cpu2026/results/flags/gcc-rev-A2.xml,
http://www.spec.org/cpu2026/results/flags/GIGA-BYTE-platform-settings-AmpereOneM-rev.1.xml.

PTDaemon, SPEC CPU, and SPECspeed are registered trademarks of the Standard Performance Evaluation Corporation. All other brand and product names appearing in this result are trademarks or registered trademarks of their respective holders.

For questions about this result, please contact the tester. For other inquiries, please contact info@spec.org
Copyright 2026 Standard Performance Evaluation Corporation
Tested with SPEC CPU®2026 v0.902.0 on 2026-02-05 23:38:31-0500.
Report generated on 2026-05-04 23:26:10 by SPEC CPU®2026 HTML formatter (f598b8d7).
Originally published on 2026-05-05.