Summary

Run type
Validate
Certification
PASS - 1 failing test waived by a reviewer
Machine
Custom build Gigabyte Technology Co., Ltd. G493-ZB1-AAP1-000
CPU
AMD EPYC 9965 192-Core Processor · 2 sockets · 384 cores / 384 threads
CPU flags
170 feature flags (informational)
All flags, as reported:
3dnowprefetchabmadxaesamd_lbr_pmc_freezeamd_lbr_v2amd_ppinaperfmperfapicaratavicavxavx2avx512_bf16avx512_bitalgavx512_vbmi2avx512_vnniavx512_vp2intersectavx512_vpopcntdqavx512bwavx512cdavx512dqavx512favx512ifmaavx512vbmiavx512vlavx_vnnibmi1bmi2bpextbus_lock_detectcat_l3cdp_l3clflushclflushoptclwbclzerocmovcmp_legacyconstant_tsccpbcppccpuidcqmcqm_llccqm_mbm_localcqm_mbm_totalcqm_occup_llccr8_legacycx16cx8dedebug_swapdecodeassistsermsextapicextd_apicidf16cflush_l1dflushbyasidfmafpufsgsbasefsrmfxsrfxsr_optgfnihthw_pstateibpbibrsibrs_enhancedibsinvpcidirperfla57lahf_lmlbrvlmmbamcamcemisalignssemmxmmxextmonitormovbemovdir64bmovdirimsrmtrrmwaitxnonstop_tscnoplnptnrip_savenxospkeosvwoverflow_recovpaepatpausefilterpcidpclmulqdqpdpe1gbperfctr_coreperfctr_llcperfctr_nbperfmon_v2pfthresholdpgepkupnipopcntpsepse36raplrdpidrdprurdrandrdseedrdt_ardtscprep_goodsepsha_niskinitsmapsmcasmepssbdssesse2sse4_1sse4_2sse4assse3stibpsuccorsvmsvm_locksyscalltcetopoexttsctsc_adjusttsc_scaleumipuser_shstkv_spec_ctrlv_vmsave_vmloadvaesvgifvmcb_cleanvmevmmcallvnmivpclmulqdqwbnoinvdwdtx2apicx2avicxgetbv1xsavexsavecxsaveerptrxsaveoptxsavesxtopology
GPU
NVIDIA H200 NVL (nvidia 595.91.07)
Memory
755 GB · 24 × DDR5s @ 4800 MT/s
Memory modules
SlotChannelSizeType SpeedRankManufacturerPart number
DIMM_P0_A1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_B1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_C1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_D1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_E1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_F1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_G1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_H1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_I1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_J1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_K1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P0_L1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P1_M1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P1_N1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P1_O1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P1_P1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P1_Q1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P1_R1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPKC
DIMM_P1_S1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPEC
DIMM_P1_T1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPEC
DIMM_P1_U1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPEC
DIMM_P1_V1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPEC
DIMM_P1_W1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPEC
DIMM_P1_X1 BANK 1 32 GB DDR5 4800 MT/s (rated 6400) 2 Samsung M321R4GA3EB2-CCPWC
24 of 48 slots populated
AlmaLinux
10.2
Kernel
6.12.0-211.42.1.el10_2.x86_64_v2 tainted (12288) An out-of-tree module was loaded; An unsigned module was loaded
Suite
0.9.0 (0.9.0-5.el10)
Submitted by
scc26-segfault · 2026-11-18 23:37

Validation tests

TestCategorySeverityStatusDetail
bench.compile.python compilation pass
bench.compress.xz compression pass
bench.compress.zstd compression pass
bench.cpu.stressng-matrix cpu pass
bench.cpu.sysbench-multi cpu pass
bench.cpu.sysbench-single cpu pass
bench.crypto.openssl-aes256gcm crypto pass
bench.crypto.openssl-rsa4096 crypto pass
bench.crypto.openssl-sha256 crypto pass
bench.gpu.clpeak gpu pass
bench.gpu.cuda-bandwidth gpu pass
bench.mem.bandwidth memory pass
bench.mem.latency memory pass
bench.mem.stressng-stream memory pass
bench.net.iperf3-reverse network skip no --peer given (needs an iperf3 server on a second host)
bench.net.iperf3-tcp network skip no --peer given (needs an iperf3 server on a second host)
bench.sched.stressng-switch scheduler pass
validate.cpu.flags cpu informational skip no CPU feature flags were reported by this system
validate.cpu.functional cpu required pass
validate.gpu.cuda-bandwidth gpu required pass host to device 53.65 GiB/s, device to host 52.34 GiB/s
validate.gpu.cuda-devicequery gpu required pass CUDA sees 8 device(s): NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL (runtime 13030, driver 13020)
validate.gpu.cuda-vectoradd gpu required pass 16777216 elements computed correctly on the device
validate.gpu.driver gpu informational pass 8 with a driver (nvidia); 1 management display adapter(s), which need none (ASPEED Graphics Family [2000])
validate.gpu.nvidia-smi gpu required pass 8 NVIDIA GPU(s), driver 595.91.07
validate.gpu.opencl-compute gpu required pass 4194304 elements computed correctly on NVIDIA H200 NVL
validate.gpu.opencl-devices gpu required pass OpenCL sees 8 hardware device(s): NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL
validate.gpu.vulkan-compute gpu required pass 16777216 bytes filled and verified on NVIDIA H200 NVL (discrete)
validate.gpu.vulkan-devices gpu required pass Vulkan sees 8 hardware device(s): NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL, NVIDIA H200 NVL
validate.ipmi.bmc ipmi conditional pass GIGA-BYTE TECHNOLOGY CO., LTD BMC, firmware 13.06, IPMI 2.0
validate.ipmi.health ipmi conditional pass 80 sensor(s) read, none critical
validate.ipmi.sel ipmi informational pass 93 SEL entries, none indicating a hardware failure
validate.kernel.dmesg kernel required fail 7 kernel error message(s) matched failure patterns
validate.kernel.taint kernel required pass tainted (O E) but not by anything hardware-related: out-of-tree module loaded, unsigned module loaded
validate.media.audio media informational pass
validate.media.video media informational pass
validate.memory.edac memory conditional pass
validate.memory.functional memory required pass
validate.network.datapath network conditional skip no --peer given (needs an iperf3 server on a second host)
validate.network.link network required pass
validate.network.pci-drivers network required pass
validate.platform.firmware platform informational pass
validate.platform.pcie-aer platform required pass
validate.platform.rtc platform required pass
validate.platform.secureboot platform informational pass
validate.platform.selinux platform informational pass worth a look: SELinux is disabled, so this run was not performed in the configuration AlmaLinux ships. The hardware results stand; treat them as weaker evidence for anything that depends on driver or firmware-loading paths under enforcing
validate.platform.thermal platform conditional skip no ACPI thermal zones
validate.platform.watchdog platform conditional pass
validate.power.backlight power conditional skip brightness is not adjustable on this system: the kernel exposes no backlight device and the attached display offered no brightness control over DDC/CI (no display answered DDC/CI)
validate.power.cpufreq power conditional pass
validate.storage.io-sanity storage required pass
validate.storage.nvme-errorlog storage conditional pass
validate.storage.smart storage required pass
validate.virt.kvm virtualization conditional pass

Benchmarks

BenchmarkMetricValue
Python build bench.compile.python build_time primary 19.1 s ↓ better
xz compression bench.compress.xz comp_speed primary 236 MB/s ↑ better
zstd compression bench.compress.zstd comp_speed_l19 248 MB/s ↑ better
zstd compression bench.compress.zstd comp_speed_l3 primary 13,290 MB/s ↑ better
zstd compression bench.compress.zstd decomp_speed_l19 10,582 MB/s ↑ better
zstd compression bench.compress.zstd decomp_speed_l3 7,676 MB/s ↑ better
CPU matrix math bench.cpu.stressng-matrix bogo_ops_per_sec primary 1,846,858 bogo-ops/s ↑ better
CPU, all cores bench.cpu.sysbench-multi events_per_sec primary 858,835 events/s ↑ better
CPU, single core bench.cpu.sysbench-single events_per_sec primary 2,540 events/s ↑ better
AES-256-GCM bench.crypto.openssl-aes256gcm throughput primary 10,375 MB/s ↑ better
RSA-4096 signing bench.crypto.openssl-rsa4096 signs_per_sec primary 423 ops/s ↑ better
RSA-4096 signing bench.crypto.openssl-rsa4096 verifies_per_sec 11,442 ops/s ↑ better
SHA-256 bench.crypto.openssl-sha256 throughput primary 1,104 MB/s ↑ better
GPU compute bench.gpu.clpeak OpenCL Compute · Single precision opencl_single_precision_compute NVIDIA H200 NVL 59,022 GFLOPS ↑ better
GPU compute bench.gpu.clpeak OpenCL Compute · Double precision opencl_double_precision_compute NVIDIA H200 NVL 29,915 GFLOPS ↑ better
GPU compute bench.gpu.clpeak OpenCL Compute · Integer opencl_integer_compute NVIDIA H200 NVL 29,924 GOPS ↑ better
GPU compute bench.gpu.clpeak OpenCL Bandwidth · Global memory opencl_global_memory_bandwidth NVIDIA H200 NVL 4,147 GB/s ↑ better
GPU compute bench.gpu.clpeak OpenCL Bandwidth · Local memory opencl_local_memory_bandwidth NVIDIA H200 NVL 25,777 GB/s ↑ better
GPU compute bench.gpu.clpeak OpenCL Bandwidth · Image memory opencl_image_memory_bandwidth NVIDIA H200 NVL 3,351 GB/s ↑ better
GPU compute bench.gpu.clpeak OpenCL Bandwidth · Host transfer opencl_transfer_bandwidth NVIDIA H200 NVL 57.6 GB/s ↑ better
GPU compute bench.gpu.clpeak OpenCL Latency · Kernel launch opencl_kernel_launch_latency NVIDIA H200 NVL 20.8 us ↓ better
GPU compute bench.gpu.clpeak CUDA Compute · Single precision cuda_single_precision_compute NVIDIA H200 NVL primary 59,636 GFLOPS ↑ better
GPU compute bench.gpu.clpeak CUDA Compute · Double precision cuda_double_precision_compute NVIDIA H200 NVL 30,044 GFLOPS ↑ better
GPU compute bench.gpu.clpeak CUDA Compute · Half precision cuda_half_precision_compute NVIDIA H200 NVL 59,806 GFLOPS ↑ better
GPU compute bench.gpu.clpeak CUDA Compute · Mixed precision cuda_mixed_precision_compute NVIDIA H200 NVL 58,534 GFLOPS ↑ better
GPU compute bench.gpu.clpeak CUDA Compute · bfloat16 cuda_bfloat16_compute NVIDIA H200 NVL 57,716 GFLOPS ↑ better
GPU compute bench.gpu.clpeak CUDA Compute · Integer cuda_integer_compute NVIDIA H200 NVL 28,554 GOPS ↑ better
GPU compute bench.gpu.clpeak CUDA Compute · Integer, int8 dot product cuda_integer_compute_int8_dp NVIDIA H200 NVL 113,745 GOPS ↑ better
GPU compute bench.gpu.clpeak CUDA Bandwidth · Global memory cuda_global_memory_bandwidth NVIDIA H200 NVL 3,906 GB/s ↑ better
GPU compute bench.gpu.clpeak CUDA Bandwidth · Local memory cuda_local_memory_bandwidth NVIDIA H200 NVL 24,093 GB/s ↑ better
GPU compute bench.gpu.clpeak CUDA Bandwidth · Image memory cuda_image_memory_bandwidth NVIDIA H200 NVL 3,192 GB/s ↑ better
GPU compute bench.gpu.clpeak CUDA Bandwidth · Host transfer cuda_transfer_bandwidth NVIDIA H200 NVL 56.8 GB/s ↑ better
GPU compute bench.gpu.clpeak CUDA Latency · Kernel launch cuda_kernel_launch_latency NVIDIA H200 NVL 7.42 us ↓ better
GPU compute bench.gpu.clpeak Vulkan Compute · Single precision vulkan_single_precision_compute NVIDIA H200 NVL 59,022 GFLOPS ↑ better
GPU compute bench.gpu.clpeak Vulkan Compute · Double precision vulkan_double_precision_compute NVIDIA H200 NVL 29,523 GFLOPS ↑ better
GPU compute bench.gpu.clpeak Vulkan Compute · Half precision vulkan_half_precision_compute NVIDIA H200 NVL 59,917 GFLOPS ↑ better
GPU compute bench.gpu.clpeak Vulkan Compute · Mixed precision vulkan_mixed_precision_compute NVIDIA H200 NVL 59,132 GFLOPS ↑ better
GPU compute bench.gpu.clpeak Vulkan Compute · bfloat16 vulkan_bfloat16_compute NVIDIA H200 NVL 55,176 GFLOPS ↑ better
GPU compute bench.gpu.clpeak Vulkan Compute · Integer vulkan_integer_compute NVIDIA H200 NVL 30,072 GOPS ↑ better
GPU compute bench.gpu.clpeak Vulkan Compute · Integer, int8 dot product vulkan_integer_compute_int8_dp NVIDIA H200 NVL 29,127 GOPS ↑ better
GPU compute bench.gpu.clpeak Vulkan Bandwidth · Global memory vulkan_global_memory_bandwidth NVIDIA H200 NVL 4,035 GB/s ↑ better
GPU compute bench.gpu.clpeak Vulkan Bandwidth · Local memory vulkan_local_memory_bandwidth NVIDIA H200 NVL 25,357 GB/s ↑ better
GPU compute bench.gpu.clpeak Vulkan Bandwidth · Image memory vulkan_image_memory_bandwidth NVIDIA H200 NVL 3,277 GB/s ↑ better
GPU compute bench.gpu.clpeak Vulkan Bandwidth · Host transfer vulkan_transfer_bandwidth NVIDIA H200 NVL 51.1 GB/s ↑ better
GPU compute bench.gpu.clpeak Vulkan Latency · Kernel launch vulkan_kernel_launch_latency NVIDIA H200 NVL 58.4 us ↓ better
GPU transfer bench.gpu.cuda-bandwidth device_to_device 1,874 GiB/s ↑ better
GPU transfer bench.gpu.cuda-bandwidth device_to_host 52.5 GiB/s ↑ better
GPU transfer bench.gpu.cuda-bandwidth host_to_device primary 53.6 GiB/s ↑ better
Memory bandwidth bench.mem.bandwidth add_bandwidth 178,937 MB/s ↑ better
Memory bandwidth bench.mem.bandwidth copy_bandwidth 119,837 MB/s ↑ better
Memory bandwidth bench.mem.bandwidth scale_bandwidth 119,304 MB/s ↑ better
Memory bandwidth bench.mem.bandwidth triad_bandwidth primary 179,017 MB/s ↑ better
Memory latency bench.mem.latency latency_64m primary 86.5 ns ↓ better
Memory stream bench.mem.stressng-stream compute_rate 117 Mflop/s ↑ better
Memory stream bench.mem.stressng-stream memory_read_rate primary 1,339 MB/s ↑ better
Memory stream bench.mem.stressng-stream memory_write_rate 893 MB/s ↑ better
Context switching bench.sched.stressng-switch ctx_switch_latency 2,716 ns ↓ better
Context switching bench.sched.stressng-switch ctx_switch_rate primary 141,362,355 ops/s ↑ better