ISS Technology Update Index
Page 4
... c7000 Enclosures 3 Volume 8, Number 3 U Unbuffered DDR-3 DIMMs are a cost-effective solution for 5 ProLiant G6 servers Volume 8, Number 4 Understanding and implementing Accelerated iSCSI on 3 HP ProLiant servers Volume 8, Number 5 Understanding and interpreting the SPECpower_ssj2008 2 benchmark Volume 8, Number 6 Understanding and interpreting the VMmark benchmark 1 Volume 8, Number 2 Use iLO Advanced to capture server event video 2 Volume 7, Number 9 Using...
... c7000 Enclosures 3 Volume 8, Number 3 U Unbuffered DDR-3 DIMMs are a cost-effective solution for 5 ProLiant G6 servers Volume 8, Number 4 Understanding and implementing Accelerated iSCSI on 3 HP ProLiant servers Volume 8, Number 5 Understanding and interpreting the SPECpower_ssj2008 2 benchmark Volume 8, Number 6 Understanding and interpreting the VMmark benchmark 1 Volume 8, Number 2 Use iLO Advanced to capture server event video 2 Volume 7, Number 9 Using...
ISS Technology Update Volume 8, Number 6
Page 2
...It can be run quickly and efficiently at eleven different levels from zero to one hundred percent load. For example, the benchmark results for the HP ProLiant DL385 G6 are measured at eleven different levels of server utilization for a fixed period of time (240 seconds). In ...late 2007, the Standard Performance Evaluation Corporation (SPEC) introduced the first SPECpower benchmark (SPECpower_ssj2008) to address this methodology, the server performance...
...It can be run quickly and efficiently at eleven different levels from zero to one hundred percent load. For example, the benchmark results for the HP ProLiant DL385 G6 are measured at eleven different levels of server utilization for a fixed period of time (240 seconds). In ...late 2007, the Standard Performance Evaluation Corporation (SPEC) introduced the first SPECpower benchmark (SPECpower_ssj2008) to address this methodology, the server performance...
ISS Technology Update Volume 8, Number 6
Page 3
... developments in this article, visit the following websites: Source SPEC.org SPECpower_ssj2008 main page HP ProLiant Servers Benchmarks page Email / Hyperlink http://www.spec.org/power_ssj2008/ http://h18004.www1.hp.com/products/servers/benchmark s/index.html 3 Thus, the overall operations-per -watt value. Overall operations-per...center remains an important topic. Version 1.10 of the server. To contact an HP spec representative, click here: spec@hp.com. If it is a relatively new, but increasingly important benchmark, and it may be used in this area will be too broad a ...
... developments in this article, visit the following websites: Source SPEC.org SPECpower_ssj2008 main page HP ProLiant Servers Benchmarks page Email / Hyperlink http://www.spec.org/power_ssj2008/ http://h18004.www1.hp.com/products/servers/benchmark s/index.html 3 Thus, the overall operations-per -watt value. Overall operations-per...center remains an important topic. Version 1.10 of the server. To contact an HP spec representative, click here: spec@hp.com. If it is a relatively new, but increasingly important benchmark, and it may be used in this area will be too broad a ...
ISS Technology Update, Volume 8, Number 2
Page 1
.... ISS Technology Update Volume 8, Number 2 ISS Technology Update Volume 8, Number 2 Keeping you informed of the latest ISS technology Understanding and interpreting the VMmark benchmark 1 Expected life of the HP Smart Array BBWC upgrade module battery pack 4 Meet the Expert-Mitch Wright ...5 Quick tip: Customizing the iLO Text Console 6 Recently published industry standard...
.... ISS Technology Update Volume 8, Number 2 ISS Technology Update Volume 8, Number 2 Keeping you informed of the latest ISS technology Understanding and interpreting the VMmark benchmark 1 Expected life of the HP Smart Array BBWC upgrade module battery pack 4 Meet the Expert-Mitch Wright ...5 Quick tip: Customizing the iLO Text Console 6 Recently published industry standard...
ISS Technology Update, Volume 8, Number 2
Page 2
...GB RAM, 8 GB disk Web server SPECweb2005-based SLES 10, 2 CPU, 512 MB RAM, 8 GB disk Database server MySQL SLES 10, 2 CPU, 2 GB RAM, 10 GB disk File servers dbench SLES 10, 1 CPU, 256 MB RAM, 8 GB disk Standby server None Windows 2003, 1 CPU, 256 MB RAM, 4 GB disk Volume 8, Number 2 To ensure the benchmark...aggregate performance scores of all VMmark results is defined as an HP ProLiant DL580 G2, the performance score for three hours. Table 2, which contains the latest VMmark results for several HP ProLiant servers, illustrates the difference that more processors and cores make ...
...GB RAM, 8 GB disk Web server SPECweb2005-based SLES 10, 2 CPU, 512 MB RAM, 8 GB disk Database server MySQL SLES 10, 2 CPU, 2 GB RAM, 10 GB disk File servers dbench SLES 10, 1 CPU, 256 MB RAM, 8 GB disk Standby server None Windows 2003, 1 CPU, 256 MB RAM, 4 GB disk Volume 8, Number 2 To ensure the benchmark...aggregate performance scores of all VMmark results is defined as an HP ProLiant DL580 G2, the performance score for three hours. Table 2, which contains the latest VMmark results for several HP ProLiant servers, illustrates the difference that more processors and cores make ...
ISS Technology Update, Volume 8, Number 2
Page 3
... For additional information on the topics discussed in the VMmark tile. These benchmark scores are readily available at http://www.spec.org/specvirtualization. Recent HP ProLiant VMmark results System Processors ProLiant DL385 G5p 2 ProLiant DL585 G5 4 ProLiant DL785 G5 8 Cores 8 16 32 Volume 8, Number 2 VMmark score... results that can be used in the industry, is currently working to the VMmark benchmark The VMmark benchmark can be run on HP ProLiant servers white paper HP VMmark page Hyperlink http://www.vmware.com/products/vmmark/ http://www.spec.org/specvirtualization...
... For additional information on the topics discussed in the VMmark tile. These benchmark scores are readily available at http://www.spec.org/specvirtualization. Recent HP ProLiant VMmark results System Processors ProLiant DL385 G5p 2 ProLiant DL585 G5 4 ProLiant DL785 G5 8 Cores 8 16 32 Volume 8, Number 2 VMmark score... results that can be used in the industry, is currently working to the VMmark benchmark The VMmark benchmark can be run on HP ProLiant servers white paper HP VMmark page Hyperlink http://www.vmware.com/products/vmmark/ http://www.spec.org/specvirtualization...
ISS Technology Update, Volume 8, Number 3
Page 6
... could have a grossly over-provisioned cooling system. Consequently, alternate strategies are useful as needed. PUE is consumed by the net power that goes directly to benchmark its data center operations against the industry. PUE can dynamically provision power and cooling resources as "field" metrics to the IT equipment. The DCE of...
... could have a grossly over-provisioned cooling system. Consequently, alternate strategies are useful as needed. PUE is consumed by the net power that goes directly to benchmark its data center operations against the industry. PUE can dynamically provision power and cooling resources as "field" metrics to the IT equipment. The DCE of...
The Intel processor roadmap for industry-standard servers technology brief, 10 edition
Page 20
... be used to Woodcrest SPEC performance The Standard Performance Evaluation Corporation (SPEC) CPU2006 benchmark provides performance measurements that the fourcore processors-Clovertown, Tigerton, and Nehalem-performed better in Figure 14. SPEC CPU2006 contains... two benchmark suites: CINT2006 for measuring and comparing compute-intensive floating point performance. Figure 14. Performance comparisons TPC-C performance The Transaction Processing Performance Council benchmark TPC-C results for Woodcrest, Clovertown, Tulsa, Nehalem,...
... be used to Woodcrest SPEC performance The Standard Performance Evaluation Corporation (SPEC) CPU2006 benchmark provides performance measurements that the fourcore processors-Clovertown, Tigerton, and Nehalem-performed better in Figure 14. SPEC CPU2006 contains... two benchmark suites: CINT2006 for measuring and comparing compute-intensive floating point performance. Figure 14. Performance comparisons TPC-C performance The Transaction Processing Performance Council benchmark TPC-C results for Woodcrest, Clovertown, Tulsa, Nehalem,...
Visualization and Acceleration in HP ProLiant servers
Page 7
... is to investigate various acceleration technologies for high performance computing and benchmark them to determine which ones are most beneficial for the ProLiant G6 servers. As part of products that have been qualified with the HP ProLiant servers is an, on-going, hands-on HP ProLiant servers Product NVIDIA Tesla S1070 AMD ATI Firestream 9250 NVIDIA...
... is to investigate various acceleration technologies for high performance computing and benchmark them to determine which ones are most beneficial for the ProLiant G6 servers. As part of products that have been qualified with the HP ProLiant servers is an, on-going, hands-on HP ProLiant servers Product NVIDIA Tesla S1070 AMD ATI Firestream 9250 NVIDIA...
Dynamic Power Capping TCO and Best Practices White Paper (EMEA edition)
Page 6
...Power Capping at the following section. For the purposes of this paper, the default settings of the Dynamic Power Capping solution. HP ProLiant DL 380 G5 Power and Costs Projections Faceplate Power Calculator Server Power Maximum Number of Servers in Circuit Power Provisioning Cost Per Server...QuickSpecs for the Dynamic Power Capping portion of the server workload as measured by CPU utilization. Case 1. HP ProLiant DL380 G5 Server Comparisons of 69%. The SpecJBB benchmark was used for the analysis was then applied to the test servers and test enclosures to size server...
...Power Capping at the following section. For the purposes of this paper, the default settings of the Dynamic Power Capping solution. HP ProLiant DL 380 G5 Power and Costs Projections Faceplate Power Calculator Server Power Maximum Number of Servers in Circuit Power Provisioning Cost Per Server...QuickSpecs for the Dynamic Power Capping portion of the server workload as measured by CPU utilization. Case 1. HP ProLiant DL380 G5 Server Comparisons of 69%. The SpecJBB benchmark was used for the analysis was then applied to the test servers and test enclosures to size server...
Dynamic Power Capping TCO and Best Practices White Paper (EMEA edition)
Page 7
...from 16 servers using Dynamic Power Capping at 100% CPU utilization. Table 2. HP ProLiant BL460c G1 and G5 Servers Comparisons of faceplate, power calculator, and SPECjbb benchmark with Dynamic Power Capping for a typical BL460c G1 configuration, users can potentially ...Server* Blade System Sizer (Enclosure) 4,722W 28 Blade Servers Load Testing and Dynamic Power Capping 4,257 W 32 Blade Servers 7 Case 2. HP ProLiant BL460c G1 Server Power and Costs Projections Faceplate Blade System Sizer (Enclosure) Load Testing and Dynamic Power Capping Enclosure Power 7,836 W 6,000...
...from 16 servers using Dynamic Power Capping at 100% CPU utilization. Table 2. HP ProLiant BL460c G1 and G5 Servers Comparisons of faceplate, power calculator, and SPECjbb benchmark with Dynamic Power Capping for a typical BL460c G1 configuration, users can potentially ...Server* Blade System Sizer (Enclosure) 4,722W 28 Blade Servers Load Testing and Dynamic Power Capping 4,257 W 32 Blade Servers 7 Case 2. HP ProLiant BL460c G1 Server Power and Costs Projections Faceplate Blade System Sizer (Enclosure) Load Testing and Dynamic Power Capping Enclosure Power 7,836 W 6,000...
Dynamic Power Capping TCO and Best Practices White Paper (EMEA edition)
Page 8
...per circuit is part of Dynamic Power Capping settings based on different blade server models. Also, power reclamation potential is outlined below SpecJBB benchmark levels will fit, the number of incremental servers does not justify the purchase of server per kW * While technically more dramatically increase... settings for blades is also a free trial for rack-mount servers using the ProLiant BL 460c G1 and G5 servers and the HP BladeSystem c7000 enclosure as an example. To begin using HP Power Calculator, and refining Dynamic Power Cap settings based on a circuit even when...
...per circuit is part of Dynamic Power Capping settings based on different blade server models. Also, power reclamation potential is outlined below SpecJBB benchmark levels will fit, the number of incremental servers does not justify the purchase of server per kW * While technically more dramatically increase... settings for blades is also a free trial for rack-mount servers using the ProLiant BL 460c G1 and G5 servers and the HP BladeSystem c7000 enclosure as an example. To begin using HP Power Calculator, and refining Dynamic Power Cap settings based on a circuit even when...
Dynamic Power Capping TCO and Best Practices White Paper (EMEA edition)
Page 9
... usage. 9 The following table summarizes the best practice for establishing Dynamic Power Caps for your ProLiant rack-mount server in accordance with actual peak power usage. Table 5. HP recommends collecting and observing data for at 450 Watts. • Using Insight Control Environment, observe... with your specific configuration using the HP Power Calculator default settings for our DL380 G5 TCO benchmark configuration (see Appendix), our configuration yields a power estimate of the branch circuit and PDUs to observe power usage for your HP ProLiant server (see Appendix for list of...
... usage. 9 The following table summarizes the best practice for establishing Dynamic Power Caps for your ProLiant rack-mount server in accordance with actual peak power usage. Table 5. HP recommends collecting and observing data for at 450 Watts. • Using Insight Control Environment, observe... with your specific configuration using the HP Power Calculator default settings for our DL380 G5 TCO benchmark configuration (see Appendix), our configuration yields a power estimate of the branch circuit and PDUs to observe power usage for your HP ProLiant server (see Appendix for list of...
Dynamic Power Capping TCO and Best Practices White Paper (EMEA edition)
Page 10
... enclosure, the process follow the same general principles as our rack-mount example with your facilities person, to your HP ProLiant server (see Appendix for list of blades that setting and adjusting Dynamic Power Caps should always be as short as a few days...enclosure devoted to observe power usage and adjust if Dynamic Power Cap is invoked frequently. For the BL460c G1 configuration used in our TCO benchmarking (see Appendix), the BladeSystem Sizer indicates that aligns to the enclosure's application duty cycle. Depending on the circuit and divide available circuit capacity...
... enclosure, the process follow the same general principles as our rack-mount example with your facilities person, to your HP ProLiant server (see Appendix for list of blades that setting and adjusting Dynamic Power Caps should always be as short as a few days...enclosure devoted to observe power usage and adjust if Dynamic Power Cap is invoked frequently. For the BL460c G1 configuration used in our TCO benchmarking (see Appendix), the BladeSystem Sizer indicates that aligns to the enclosure's application duty cycle. Depending on the circuit and divide available circuit capacity...
Dynamic Power Capping TCO and Best Practices White Paper (EMEA edition)
Page 11
... Step 3 Step 4 Step 5 Action Update server BIOS, iLO firmware, Onboard Administrator firmware, and Insight Control Environment software. Use the HP BladeSystem Sizer to estimate the number of blades that the peak value observed represents the true peak power usage for the c7000 enclosure in ...Control Environment. Continue to the enclosure. The following table summarizes the best practice for establishing Dynamic Power Caps for HP BladeSystem enclosures. Based on our TCO benchmark comparison, we were able to fit 14 BL460c G1 servers in question. • Based on observed power usage...
... Step 3 Step 4 Step 5 Action Update server BIOS, iLO firmware, Onboard Administrator firmware, and Insight Control Environment software. Use the HP BladeSystem Sizer to estimate the number of blades that the peak value observed represents the true peak power usage for the c7000 enclosure in ...Control Environment. Continue to the enclosure. The following table summarizes the best practice for establishing Dynamic Power Caps for HP BladeSystem enclosures. Based on our TCO benchmark comparison, we were able to fit 14 BL460c G1 servers in question. • Based on observed power usage...
Dynamic Power Capping TCO and Best Practices White Paper (WW edition)
Page 6
...was then applied to the test servers and test enclosures to size server power requirements more accurately based on a continuous basis. 2. HP ProLiant DL380 G5 Server Comparisons of the server workload as measured by CPU utilization. Using the Uptime Institute estimate of $25,000 per... in Watts (faceplate power) was measured at 100% CPU utilization running the SPECjbb benchmark with a maximum number of nearly 230%. The results of 69%. The SpecJBB benchmark was used . HP ProLiant DL 380 G5 Power and Costs Projections Faceplate Power Calculator Server Power Maximum Number of...
...was then applied to the test servers and test enclosures to size server power requirements more accurately based on a continuous basis. 2. HP ProLiant DL380 G5 Server Comparisons of the server workload as measured by CPU utilization. Using the Uptime Institute estimate of $25,000 per... in Watts (faceplate power) was measured at 100% CPU utilization running the SPECjbb benchmark with a maximum number of nearly 230%. The results of 69%. The SpecJBB benchmark was used . HP ProLiant DL 380 G5 Power and Costs Projections Faceplate Power Calculator Server Power Maximum Number of...
Dynamic Power Capping TCO and Best Practices White Paper (WW edition)
Page 7
HP ProLiant BL460c G1 and G5 Servers Comparisons of faceplate, power calculator, and SPECjbb benchmark with Dynamic Power Capping for the BL460c G1 server and c7000 as well as their primary means of budgeting server power will see the most ... circuit from $13,438 per server to reclaim trapped power capacity at max power, an increase of 49%. For the ProLiant BL460c G5 server, the results are listed below (Tables 2 and 3). HP ProLiant BL460c G1 Server Power and Costs Projections Faceplate Blade System Sizer (Enclosure) Load Testing and Dynamic Power Capping Enclosure Power...
HP ProLiant BL460c G1 and G5 Servers Comparisons of faceplate, power calculator, and SPECjbb benchmark with Dynamic Power Capping for the BL460c G1 server and c7000 as well as their primary means of budgeting server power will see the most ... circuit from $13,438 per server to reclaim trapped power capacity at max power, an increase of 49%. For the ProLiant BL460c G5 server, the results are listed below (Tables 2 and 3). HP ProLiant BL460c G1 Server Power and Costs Projections Faceplate Blade System Sizer (Enclosure) Load Testing and Dynamic Power Capping Enclosure Power...
Dynamic Power Capping TCO and Best Practices White Paper (WW edition)
Page 8
... of other important observations to make when looking at different rates, setting a cap below SpecJBB benchmark levels will be able to set Dynamic Power Capping baselines and culminates with the ProLiant DL380 G5, the overall number of pooled power, means that they can reclaim more power and... Sizer to reclaim additional power capacity by setting caps lower than the values suggested by the SpecJBB benchmark. HP ProLiant BL460c G5 Server Power and Costs Projections Faceplate Blade System Sizer (Enclosure) Load Testing and Dynamic Power Capping Enclosure Power 8,340 W 4,722...
... of other important observations to make when looking at different rates, setting a cap below SpecJBB benchmark levels will be able to set Dynamic Power Capping baselines and culminates with the ProLiant DL380 G5, the overall number of pooled power, means that they can reclaim more power and... Sizer to reclaim additional power capacity by setting caps lower than the values suggested by the SpecJBB benchmark. HP ProLiant BL460c G5 Server Power and Costs Projections Faceplate Blade System Sizer (Enclosure) Load Testing and Dynamic Power Capping Enclosure Power 8,340 W 4,722...
Dynamic Power Capping TCO and Best Practices White Paper (WW edition)
Page 9
...initial Dynamic Power Cap baseline using the HP Power Calculator default settings for Rack-Mount Servers Required Component ROM BIOS iLO IPM component of interest. Component Requirement for our DL380 G5 TCO benchmark configuration (see Appendix for our ProLiant DL380 G5 configuration, we would set ...the initial Dynamic Power Cap at www.hp.com/go/tryinsightcontrol. There is invoked too frequently, it may be done...
...initial Dynamic Power Cap baseline using the HP Power Calculator default settings for Rack-Mount Servers Required Component ROM BIOS iLO IPM component of interest. Component Requirement for our DL380 G5 TCO benchmark configuration (see Appendix for our ProLiant DL380 G5 configuration, we would set ...the initial Dynamic Power Cap at www.hp.com/go/tryinsightcontrol. There is invoked too frequently, it may be done...
Dynamic Power Capping TCO and Best Practices White Paper (WW edition)
Page 11
...enclosure Dynamic Power Cap to observe power usage for c7000 enclosure based on our TCO benchmark comparison, we were able to fit 14 BL460c G1 servers in each enclosure, using HP Insight Control Environment. The following table summarizes the best practice for establishing Dynamic Power...Update server BIOS, iLO firmware, Onboard Administrator firmware, and Insight Control Environment software. For the BL460c G1 configuration used in our TCO benchmarking (see Appendix), the BladeSystem Sizer indicates that 10 BL460 G1 servers will fit into the enclosure, the IT department may want to ...
...enclosure Dynamic Power Cap to observe power usage for c7000 enclosure based on our TCO benchmark comparison, we were able to fit 14 BL460c G1 servers in each enclosure, using HP Insight Control Environment. The following table summarizes the best practice for establishing Dynamic Power...Update server BIOS, iLO firmware, Onboard Administrator firmware, and Insight Control Environment software. For the BL460c G1 configuration used in our TCO benchmarking (see Appendix), the BladeSystem Sizer indicates that 10 BL460 G1 servers will fit into the enclosure, the IT department may want to ...