2*Intel(R) Xeon(R) Gold 5122 CPU @ 3.60GHz
12*HPE SmartMemory DDR4-2666 RDIMM 16GiB
<\/p>\n\n\n\n
iLO 5 1.37 Oct 25 2018
System ROM U30 v1.46 (10\/02\/2018)
Intelligent Platform Abstraction Data 7.2.0 Build 30
System Programmable Logic Device 0x2A
Power Management Controller Firmware 1.0.4
NVMe Backplane Firmware 1.20
Power Supply Firmware 1.00
Power Supply Firmware 1.00
Innovation Engine (IE) Firmware 0.1.6.1
Server Platform Services (SPS) Firmware 4.0.4.288
Redundant System ROM U30 v1.42 (06\/20\/2018)
Intelligent Provisioning 3.20.154
Power Management Controller FW Bootloader 1.1
HPE Smart Storage Battery 1 Firmware 0.60
HPE Eth 10\/25Gb 2p 631FLR-SFP28 Adptr 212.0.103001
HPE Ethernet 1Gb 4-port 331i Adapter – NIC 20.12.41
HPE Smart Array P816i-a SR Gen10 1.65
HPE 100Gb 1p OP101 QSFP28 x16 OPA Adptr 1.5.2.0.0
HPE InfiniBand EDR\/Ethernet 100Gb 2-port 840QSF 12.22.40.30
Embedded Video Controller 2.5<\/p>\n\n\n\n
CentOS Linux release 7.6.1810 (Core)
Linux yaoge123 3.10.0-957.el7.x86_64 #1 SMP Thu Nov 8 23:39:32 UTC 2018 x86_64 x86_64 x86_64 GNU\/Linux
Intel(R) Memory Latency Checker – v3.6<\/p>\n\n\n\n\n\n\n\n
.\/mlc_avx512;.\/mlc_avx512 -Y;.\/mlc_avx512 -Z
\u53ea\u53d6\u540e\u4e24\u6b21\u6d4b\u8bd5\u7ed3\u679c<\/p>\n\n\n\n
Workload Profile: General Power Efficient Compute
Intel (R) Hyperthreading Options: Disabled
Enabled Cores per Processor: 0
Processor x2APIC Support: Disabled
Intel(R) Virtualization Technology (Intel VT): Disabled
Intel (R) VT-d: Disabled
SR-IOV: Disabled
Advanced Memory Protection: Advanced ECC
Power Regulator: Dynamic Power Savings Mode<\/p>\n\n\n\n
total used free shared buff\/cache available\nMem: 188G 1.5G 186G 9.7M 222M 186G\nSwap: 4.0G 0B 4.0G\n\n.\/mlc_avx512 -Y\nIntel(R) Memory Latency Checker - v3.6\nMeasuring idle latencies (in ns)...\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t 81.7\t 136.6\t\n 1\t 133.1\t 76.9\t\n\nMeasuring Peak Injection Memory Bandwidths for the system\nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing traffic with the following read-write ratios\nALL Reads :\t99954.8\t\n3:1 Reads-Writes :\t145189.0\t\n2:1 Reads-Writes :\t151137.4\t\n1:1 Reads-Writes :\t150854.9\t\nStream-triad like:\t101469.8\t\n\nMeasuring Memory Bandwidths between nodes within system \nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t49976.5\t31093.5\t\n 1\t30873.7\t50050.0\t\n\nMeasuring Loaded Latencies for the system\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\nInject\tLatency\tBandwidth\nDelay\t(ns)\tMB\/sec\n==========================\n 00000\t 88.68\t 89235.1\n 00002\t 88.73\t 89335.3\n 00008\t 88.65\t 89217.6\n 00015\t 88.76\t 88243.5\n 00050\t 89.40\t 85207.4\n 00100\t 84.77\t 66373.2\n 00200\t 83.76\t 45796.8\n 00300\t 83.08\t 35407.3\n 00400\t 82.67\t 28352.7\n 00500\t 82.29\t 23883.3\n 00700\t 81.76\t 17950.3\n 01000\t 81.61\t 13123.4\n 01300\t 81.48\t 10414.7\n 01700\t 81.54\t 8249.4\n 02500\t 81.43\t 5930.9\n 03500\t 81.43\t 4481.2\n 05000\t 81.47\t 3384.8\n 09000\t 81.63\t 2226.1\n 20000\t 81.44\t 1440.1\n\nMeasuring cache-to-cache transfer latency (in ns)...\nLocal Socket L2->L2 HIT latency\t47.4\nLocal Socket L2->L2 HITM latency\t47.5\nRemote Socket L2->L2 HITM latency (data address homed in writer socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 112.5\t\n 1\t 114.0\t -\t\nRemote Socket L2->L2 HITM latency (data address homed in reader socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 178.3\t\n 1\t 180.4\t -\t\n\n.\/mlc_avx512 -Z\nIntel(R) Memory Latency Checker - v3.6\nMeasuring idle latencies (in ns)...\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t 81.6\t 137.7\t\n 1\t 133.3\t 77.5\t\n\nMeasuring Peak Injection Memory Bandwidths for the system\nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing traffic with the following read-write ratios\nALL Reads :\t104428.9\t\n3:1 Reads-Writes :\t138213.4\t\n2:1 Reads-Writes :\t148680.5\t\n1:1 Reads-Writes :\t148760.9\t\nStream-triad like:\t97657.5\t\n\nMeasuring Memory Bandwidths between nodes within system \nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t52281.3\t31156.8\t\n 1\t31071.8\t51930.0\t\n\nMeasuring Loaded Latencies for the system\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\nInject\tLatency\tBandwidth\nDelay\t(ns)\tMB\/sec\n==========================\n 00000\t 88.14\t 92987.7\n 00002\t 88.07\t 92861.1\n 00008\t 86.46\t 87868.3\n 00015\t 87.76\t 87068.6\n 00050\t 89.06\t 84397.9\n 00100\t 84.80\t 65169.7\n 00200\t 83.98\t 45178.1\n 00300\t 83.05\t 34780.6\n 00400\t 82.47\t 27854.5\n 00500\t 82.08\t 23249.2\n 00700\t 81.80\t 17532.0\n 01000\t 81.58\t 12815.7\n 01300\t 81.51\t 10190.1\n 01700\t 81.51\t 8062.1\n 02500\t 81.35\t 5794.3\n 03500\t 81.40\t 4386.5\n 05000\t 81.29\t 3314.9\n 09000\t 81.06\t 2193.8\n 20000\t 80.98\t 1427.2\n\nMeasuring cache-to-cache transfer latency (in ns)...\nLocal Socket L2->L2 HIT latency\t47.4\nLocal Socket L2->L2 HITM latency\t47.4\nRemote Socket L2->L2 HITM latency (data address homed in writer socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 112.5\t\n 1\t 114.1\t -\t\nRemote Socket L2->L2 HITM latency (data address homed in reader socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 178.7\t\n 1\t 180.7\t -\t<\/code><\/pre>\n\n\n\nOnline Spare Memory with Advanced ECC Support
<\/h4>\n\n\n\n
Workload Profile: General Power Efficient Compute
Intel (R) Hyperthreading Options: Disabled
Enabled Cores per Processor: 0
Processor x2APIC Support: Disabled
Intel(R) Virtualization Technology (Intel VT): Disabled
Intel (R) VT-d: Disabled
SR-IOV: Disabled
Advanced Memory Protection: Online Spare Memory with Advanced ECC Support
Power Regulator: Dynamic Power Savings Mode <\/p>\n\n\n\n
total used free shared buff\/cache available\nMem: 93G 998M 92G 9.7M 169M 92G\nSwap: 4.0G 0B 4.0G\n\n.\/mlc_avx512 -Y\nIntel(R) Memory Latency Checker - v3.6\nMeasuring idle latencies (in ns)...\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t 81.5\t 137.3\t\n 1\t 132.9\t 81.6\t\n\nMeasuring Peak Injection Memory Bandwidths for the system\nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing traffic with the following read-write ratios\nALL Reads :\t101550.7\t\n3:1 Reads-Writes :\t122423.5\t\n2:1 Reads-Writes :\t123643.4\t\n1:1 Reads-Writes :\t123946.9\t\nStream-triad like:\t100338.1\t\n\nMeasuring Memory Bandwidths between nodes within system \nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t51795.6\t30250.4\t\n 1\t30170.0\t50789.5\t\n\nMeasuring Loaded Latencies for the system\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\nInject\tLatency\tBandwidth\nDelay\t(ns)\tMB\/sec\n==========================\n 00000\t 93.47\t 91710.7\n 00002\t 93.39\t 91768.5\n 00008\t 93.48\t 91771.5\n 00015\t 93.49\t 91460.7\n 00050\t 93.60\t 86147.5\n 00100\t 88.45\t 66769.2\n 00200\t 85.89\t 45695.5\n 00300\t 85.45\t 35209.2\n 00400\t 85.12\t 28174.0\n 00500\t 84.18\t 23784.2\n 00700\t 82.89\t 17896.0\n 01000\t 82.07\t 13104.8\n 01300\t 81.90\t 10377.9\n 01700\t 81.71\t 8247.6\n 02500\t 81.66\t 5928.2\n 03500\t 81.65\t 4451.3\n 05000\t 81.60\t 3383.5\n 09000\t 81.45\t 2228.4\n 20000\t 81.19\t 1443.0\n\nMeasuring cache-to-cache transfer latency (in ns)...\nLocal Socket L2->L2 HIT latency\t47.4\nLocal Socket L2->L2 HITM latency\t47.4\nRemote Socket L2->L2 HITM latency (data address homed in writer socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 112.6\t\n 1\t 114.1\t -\t\nRemote Socket L2->L2 HITM latency (data address homed in reader socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 179.0\t\n 1\t 178.1\t -\t\n\n.\/mlc_avx512 -Z\nIntel(R) Memory Latency Checker - v3.6\nMeasuring idle latencies (in ns)...\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t 81.7\t 136.3\t\n 1\t 132.6\t 75.8\t\n\nMeasuring Peak Injection Memory Bandwidths for the system\nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing traffic with the following read-write ratios\nALL Reads :\t103224.0\t\n3:1 Reads-Writes :\t117884.1\t\n2:1 Reads-Writes :\t123197.6\t\n1:1 Reads-Writes :\t122909.9\t\nStream-triad like:\t95929.5\t\n\nMeasuring Memory Bandwidths between nodes within system \nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t51349.0\t30597.5\t\n 1\t30340.8\t51559.5\t\n\nMeasuring Loaded Latencies for the system\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\nInject\tLatency\tBandwidth\nDelay\t(ns)\tMB\/sec\n==========================\n 00000\t 92.20\t 91895.3\n 00002\t 92.26\t 91902.1\n 00008\t 91.58\t 87681.6\n 00015\t 92.70\t 88728.1\n 00050\t 91.73\t 83071.0\n 00100\t 87.27\t 64608.4\n 00200\t 87.20\t 44479.3\n 00300\t 85.37\t 34295.7\n 00400\t 85.13\t 27746.6\n 00500\t 83.84\t 22966.0\n 00700\t 82.59\t 17352.7\n 01000\t 81.45\t 12820.3\n 01300\t 80.62\t 10185.6\n 01700\t 80.86\t 8072.0\n 02500\t 80.20\t 5805.8\n 03500\t 79.78\t 4404.5\n 05000\t 79.43\t 3335.5\n 09000\t 78.53\t 2219.9\n 20000\t 77.91\t 1458.6\n\nMeasuring cache-to-cache transfer latency (in ns)...\nLocal Socket L2->L2 HIT latency\t47.4\nLocal Socket L2->L2 HITM latency\t47.5\nRemote Socket L2->L2 HITM latency (data address homed in writer socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 112.6\t\n 1\t 114.1\t -\t\nRemote Socket L2->L2 HITM latency (data address homed in reader socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 176.1\t\n 1\t 178.1\t -\t<\/code><\/pre>\n\n\n\n Static High Performance Mode<\/h4>\n\n\n\n
Workload Profile: High Performance Compute (HPC)
Intel (R) Hyperthreading Options: Disabled
Enabled Cores per Processor: 0
Processor x2APIC Support: Disabled
Intel(R) Virtualization Technology (Intel VT): Disabled
Intel (R) VT-d: Disabled
SR-IOV: Disabled
Advanced Memory Protection: Advanced ECC
Power Regulator: Static High Performance Mode
NUMA Group Size Optimization: Clustered
Sub-NUMA Clustering: Disabled<\/p>\n\n\n\n
total used free shared buff\/cache available\nMem: 188G 1.5G 186G 9.7M 174M 186G\nSwap: 4.0G 0B 4.0G\n\n.\/mlc_avx512 -Y\nIntel(R) Memory Latency Checker - v3.6\nMeasuring idle latencies (in ns)...\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t 81.5\t 136.6\t\n 1\t 133.0\t 80.8\t\n\nMeasuring Peak Injection Memory Bandwidths for the system\nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing traffic with the following read-write ratios\nALL Reads :\t99671.7\t\n3:1 Reads-Writes :\t144913.0\t\n2:1 Reads-Writes :\t151828.4\t\n1:1 Reads-Writes :\t150604.2\t\nStream-triad like:\t101510.8\t\n\nMeasuring Memory Bandwidths between nodes within system \nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t50065.0\t30928.7\t\n 1\t30822.4\t49750.0\t\n\nMeasuring Loaded Latencies for the system\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\nInject\tLatency\tBandwidth\nDelay\t(ns)\tMB\/sec\n==========================\n 00000\t 88.36\t 88426.0\n 00002\t 88.28\t 88513.4\n 00008\t 88.39\t 88376.7\n 00015\t 88.86\t 87479.0\n 00050\t 91.10\t 84842.5\n 00100\t 85.64\t 65520.4\n 00200\t 84.52\t 45690.0\n 00300\t 84.02\t 35335.0\n 00400\t 82.86\t 28333.1\n 00500\t 82.59\t 23871.6\n 00700\t 81.90\t 17939.0\n 01000\t 81.59\t 13122.1\n 01300\t 81.42\t 10415.4\n 01700\t 81.48\t 8250.4\n 02500\t 81.36\t 5931.9\n 03500\t 81.35\t 4482.6\n 05000\t 81.22\t 3387.9\n 09000\t 81.06\t 2233.0\n 20000\t 80.91\t 1445.8\n\nMeasuring cache-to-cache transfer latency (in ns)...\nLocal Socket L2->L2 HIT latency\t47.2\nLocal Socket L2->L2 HITM latency\t47.3\nRemote Socket L2->L2 HITM latency (data address homed in writer socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 112.6\t\n 1\t 113.9\t -\t\nRemote Socket L2->L2 HITM latency (data address homed in reader socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 176.2\t\n 1\t 177.6\t -\t\n\n.\/mlc_avx512 -Z\nIntel(R) Memory Latency Checker - v3.6\nMeasuring idle latencies (in ns)...\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t 81.5\t 136.6\t\n 1\t 133.0\t 80.8\t\n\nMeasuring Peak Injection Memory Bandwidths for the system\nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing traffic with the following read-write ratios\nALL Reads :\t103610.7\t\n3:1 Reads-Writes :\t138387.8\t\n2:1 Reads-Writes :\t149094.5\t\n1:1 Reads-Writes :\t149883.0\t\nStream-triad like:\t97603.1\t\n\nMeasuring Memory Bandwidths between nodes within system \nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\n\t\tNuma node\nNuma node\t 0\t 1\t\n 0\t51908.0\t31130.9\t\n 1\t31021.3\t51681.4\t\n\nMeasuring Loaded Latencies for the system\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\nInject\tLatency\tBandwidth\nDelay\t(ns)\tMB\/sec\n==========================\n 00000\t 87.97\t 92148.1\n 00002\t 87.94\t 91946.7\n 00008\t 86.79\t 87489.0\n 00015\t 88.32\t 87027.6\n 00050\t 90.93\t 84764.9\n 00100\t 85.72\t 66075.6\n 00200\t 84.60\t 45281.3\n 00300\t 83.64\t 34671.7\n 00400\t 83.08\t 27878.4\n 00500\t 82.57\t 23317.8\n 00700\t 81.79\t 17534.7\n 01000\t 81.58\t 12821.8\n 01300\t 81.56\t 10186.6\n 01700\t 81.61\t 8061.2\n 02500\t 81.48\t 5795.0\n 03500\t 81.40\t 4387.2\n 05000\t 81.27\t 3317.6\n 09000\t 81.06\t 2194.2\n 20000\t 80.94\t 1427.7\n\nMeasuring cache-to-cache transfer latency (in ns)...\nLocal Socket L2->L2 HIT latency\t47.2\nLocal Socket L2->L2 HITM latency\t47.3\nRemote Socket L2->L2 HITM latency (data address homed in writer socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 112.6\t\n 1\t 113.9\t -\t\nRemote Socket L2->L2 HITM latency (data address homed in reader socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t\n 0\t -\t 176.5\t\n 1\t 178.1\t -\t<\/code><\/pre>\n\n\n\nStatic High Performance Mode – Sub-NUMA<\/h4>\n\n\n\n
Workload Profile: High Performance Compute (HPC)
Intel (R) Hyperthreading Options: Disabled
Enabled Cores per Processor: 0
Processor x2APIC Support: Disabled
Intel(R) Virtualization Technology (Intel VT): Disabled
Intel (R) VT-d: Disabled
SR-IOV: Disabled
Advanced Memory Protection: Advanced ECC
Power Regulator: Static High Performance Mode
NUMA Group Size Optimization: Clustered
Sub-NUMA Clustering: Enabled<\/p>\n\n\n\n
total used free shared buff\/cache available\nMem: 188G 1.5G 186G 9.7M 176M 186G\nSwap: 4.0G 0B 4.0G\n\n.\/mlc_avx512 -Y\nIntel(R) Memory Latency Checker - v3.6\nMeasuring idle latencies (in ns)...\n\t\tNuma node\nNuma node\t 0\t 1\t 2\t 3\t\n 0\t 73.2\t 81.9\t 131.4\t 139.6\t\n 1\t 81.4\t 73.2\t 132.8\t 140.7\t\n 2\t 128.6\t 138.5\t 73.1\t 81.4\t\n 3\t 132.4\t 140.6\t 81.4\t 80.0\t\n\nMeasuring Peak Injection Memory Bandwidths for the system\nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing traffic with the following read-write ratios\nALL Reads :\t113325.1\t\n3:1 Reads-Writes :\t153852.5\t\n2:1 Reads-Writes :\t158538.8\t\n1:1 Reads-Writes :\t163349.2\t\nStream-triad like:\t106597.2\t\n\nMeasuring Memory Bandwidths between nodes within system \nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\n\t\tNuma node\nNuma node\t 0\t 1\t 2\t 3\t\n 0\t28097.4\t26815.3\t18072.4\t17292.1\t\n 1\t27017.8\t28428.9\t18125.0\t17314.9\t\n 2\t18003.6\t17296.8\t28663.1\t26939.6\t\n 3\t17851.0\t17180.2\t26850.6\t28090.9\t\n\nMeasuring Loaded Latencies for the system\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\nInject\tLatency\tBandwidth\nDelay\t(ns)\tMB\/sec\n==========================\n 00000\t 77.56\t 100595.6\n 00002\t 77.57\t 100674.4\n 00008\t 77.55\t 100718.8\n 00015\t 77.55\t 99986.2\n 00050\t 77.43\t 92388.9\n 00100\t 75.33\t 76544.3\n 00200\t 74.09\t 45915.4\n 00300\t 74.25\t 36109.2\n 00400\t 74.10\t 28748.0\n 00500\t 74.02\t 24010.9\n 00700\t 73.74\t 18067.8\n 01000\t 73.61\t 13237.0\n 01300\t 73.53\t 10535.6\n 01700\t 73.40\t 8344.6\n 02500\t 73.25\t 6023.4\n 03500\t 73.16\t 4569.0\n 05000\t 73.13\t 3474.4\n 09000\t 73.11\t 2318.4\n 20000\t 73.07\t 1530.5\n\nMeasuring cache-to-cache transfer latency (in ns)...\nLocal Socket L2->L2 HIT latency\t45.5\nLocal Socket L2->L2 HITM latency\t45.6\nRemote Socket L2->L2 HITM latency (data address homed in writer socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t 2\t 3\t\n 0\t -\t 47.9\t 108.0\t 110.0\t\n 1\t 47.5\t -\t 114.1\t 116.1\t\n 2\t 108.4\t 109.3\t -\t 45.4\t\n 3\t 114.7\t 115.4\t 45.6\t -\t\nRemote Socket L2->L2 HITM latency (data address homed in reader socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t 2\t 3\t\n 0\t -\t 47.3\t 169.9\t 177.3\t\n 1\t 45.6\t -\t 168.8\t 176.3\t\n 2\t 173.2\t 177.3\t -\t 48.4\t\n 3\t 173.7\t 177.7\t 48.6\t -\t\n\n.\/mlc_avx512 -Z\nIntel(R) Memory Latency Checker - v3.6\nMeasuring idle latencies (in ns)...\n\t\tNuma node\nNuma node\t 0\t 1\t 2\t 3\t\n 0\t 73.2\t 81.6\t 131.0\t 139.6\t\n 1\t 81.4\t 73.2\t 132.8\t 140.7\t\n 2\t 128.6\t 138.5\t 73.1\t 81.4\t\n 3\t 132.4\t 140.6\t 81.4\t 80.0\t\n\nMeasuring Peak Injection Memory Bandwidths for the system\nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing traffic with the following read-write ratios\nALL Reads :\t116063.6\t\n3:1 Reads-Writes :\t146797.3\t\n2:1 Reads-Writes :\t155021.3\t\n1:1 Reads-Writes :\t160334.0\t\nStream-triad like:\t102302.4\t\n\nMeasuring Memory Bandwidths between nodes within system \nBandwidths are in MB\/sec (1 MB\/sec = 1,000,000 Bytes\/sec)\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\n\t\tNuma node\nNuma node\t 0\t 1\t 2\t 3\t\n 0\t28901.7\t27470.7\t17819.7\t17063.9\t\n 1\t27885.7\t29067.0\t17830.9\t17038.7\t\n 2\t17677.7\t16976.0\t29378.9\t27692.9\t\n 3\t17582.0\t16918.8\t27642.7\t28928.8\t\n\nMeasuring Loaded Latencies for the system\nUsing all the threads from each core if Hyper-threading is enabled\nUsing Read-only traffic type\nInject\tLatency\tBandwidth\nDelay\t(ns)\tMB\/sec\n==========================\n 00000\t 76.67\t 103298.6\n 00002\t 76.67\t 103133.2\n 00008\t 76.59\t 97928.7\n 00015\t 77.14\t 99307.7\n 00050\t 77.64\t 91830.6\n 00100\t 75.32\t 76645.1\n 00200\t 74.28\t 45323.2\n 00300\t 74.23\t 35625.3\n 00400\t 74.10\t 28168.6\n 00500\t 74.55\t 23432.3\n 00700\t 73.77\t 17619.9\n 01000\t 73.59\t 12901.8\n 01300\t 73.53\t 10270.0\n 01700\t 73.48\t 8151.9\n 02500\t 73.26\t 5880.3\n 03500\t 73.16\t 4477.2\n 05000\t 73.12\t 3403.2\n 09000\t 73.10\t 2279.6\n 20000\t 73.09\t 1512.5\n\nMeasuring cache-to-cache transfer latency (in ns)...\nLocal Socket L2->L2 HIT latency\t45.5\nLocal Socket L2->L2 HITM latency\t45.6\nRemote Socket L2->L2 HITM latency (data address homed in writer socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t 2\t 3\t\n 0\t -\t 47.9\t 108.0\t 109.9\t\n 1\t 47.4\t -\t 114.1\t 116.1\t\n 2\t 108.4\t 109.3\t -\t 45.4\t\n 3\t 114.7\t 115.4\t 45.6\t -\t\nRemote Socket L2->L2 HITM latency (data address homed in reader socket)\n\t\t\tReader Numa Node\nWriter Numa Node 0\t 1\t 2\t 3\t\n 0\t -\t 47.3\t 169.9\t 177.4\t\n 1\t 45.6\t -\t 168.8\t 176.3\t\n 2\t 173.3\t 177.3\t -\t 48.4\t\n 3\t 173.7\t 177.7\t 48.6\t -\t<\/code><\/pre>\n","protected":false},"excerpt":{"rendered":"\u786c\u4ef6\u73af\u5883 2*Intel(R) Xeon(R) Gold 5122 CPU @ 3.60GHz12*HPE S […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":false,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2}},"categories":[87,198,11],"tags":[],"class_list":["post-2553","post","type-post","status-publish","format-standard","hentry","category-hardware","category-hpc","category-xnix"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/paOwEq-Fb","_links":{"self":[{"href":"https:\/\/www.yaoge123.com\/blog\/wp-json\/wp\/v2\/posts\/2553","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.yaoge123.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.yaoge123.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.yaoge123.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.yaoge123.com\/blog\/wp-json\/wp\/v2\/comments?post=2553"}],"version-history":[{"count":17,"href":"https:\/\/www.yaoge123.com\/blog\/wp-json\/wp\/v2\/posts\/2553\/revisions"}],"predecessor-version":[{"id":2581,"href":"https:\/\/www.yaoge123.com\/blog\/wp-json\/wp\/v2\/posts\/2553\/revisions\/2581"}],"wp:attachment":[{"href":"https:\/\/www.yaoge123.com\/blog\/wp-json\/wp\/v2\/media?parent=2553"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.yaoge123.com\/blog\/wp-json\/wp\/v2\/categories?post=2553"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.yaoge123.com\/blog\/wp-json\/wp\/v2\/tags?post=2553"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}