Recouping Expense in the Datacenter Achieving Greater Efficiencies and Cost Savings   Presented by:  Mark Ellis IDC Enterprise Data Centre  London May 22nd 2008
THE OUTLOOK? Credit Crunch  Capacity Crunch Energy Crunch
Energy costs are driving TCO increases 70% of server TCO associated with power & cooling (vs. 70% Capex, Opex 10 years ago) - IDC
360 o  data centre efficiency – brown field examples Infrastructure Facilities  Commercial  Demand  Hot & cold aisle containment Combined heat & power Operating temperature M&E efficiencies Consolidation & virtualisation Cool refresh Global Dynamic Grids Agility  (provisioning &  workload automation) Whitebox  vs.  blade Infrastructure efficiency Set power budgets Demand challenge process Policy enforcement  (faster provisioning helps) Developer mindset shift  (box hugging to utility via OSIs) Active decom programs Organisational behaviour Fast RFP-less vendor discovery Annual supply-chain consolidation Contract & vendor management
Infrastructure efficiency - examples DIFFICULTY  (Tech and Org) low high ROI  (all positive) high Lower 1b Realistic distributed compression ratios a. Virtualisation: Dev 10:1 b. Consolidation:  DB 4.5:1 Web 3:1 MDS 2:1 N.B. Better than linear power-related savings  1. Global dynamic grids Many jobs aren’t latency sensitive no need to be in Metro tier 3 or 4 data centres.  50% saving migrating to tier 2 non metro area 2. Cool refresh e.g. Intel ‘Caneland’ 2 x performance per watt compared to Xeon MP.  3. White box preference over blade c30% Capex save. Debatable Opex delta Blade technology optimization around density rarely appropriate 4. 1a 2 4 3 5 Agile / Real-Time Infrastructure Potential for huge increases in utilisation through near-time provisioning  SOI etc 5.
M&E efficiency - examples DIFFICULTY  (Tech and Org) low high ROI  (all positive) high Lower Avoid host exhaust mixing with cold air a. Simple band aid steps,  add blanking plates, don’t use perforated tiles in hot isles, remove dead under floor cables if possible b. For high density equipment consider containment units such as APC  InfraStruXure   Operating temperature 1 o  temperature increase against typical data hall levels (20-22 o ) can = 4% energy saving – Sun Don’t exceed 25 o 1. Combined Heat & Power (CHP) Potential for 70 - 75% efficiencies at point of use versus central power station of circa 45% Heat by-product can improve cooling efficiencies Reduced C02 foot print 2. Hot & cold aisle containment 3. 1 2 3a 3b
Fastest returns from initiatives that minimise impact on application platform e.g. Cool refresh Further virtualisation Summary Controlling demand is easier than future clean-up Challenge placement and stove-piping of Grids
Thank you
About Citihub Citihub is a specialist IT infrastructure consultancy for the financial services industry  Our people are proven industry practitioners with an appetite for complex challenges, particularly around applications infrastructure We're passionate about our clients' success; the industry’s leading companies come back to us time and again because we take ownership and we deliver results [email_address]

Recouping Expense in the Datacentre - Citihub IDC Presentation May 2008

  • 1.
    Recouping Expense inthe Datacenter Achieving Greater Efficiencies and Cost Savings Presented by: Mark Ellis IDC Enterprise Data Centre London May 22nd 2008
  • 2.
    THE OUTLOOK? CreditCrunch Capacity Crunch Energy Crunch
  • 3.
    Energy costs aredriving TCO increases 70% of server TCO associated with power & cooling (vs. 70% Capex, Opex 10 years ago) - IDC
  • 4.
    360 o data centre efficiency – brown field examples Infrastructure Facilities Commercial Demand Hot & cold aisle containment Combined heat & power Operating temperature M&E efficiencies Consolidation & virtualisation Cool refresh Global Dynamic Grids Agility (provisioning & workload automation) Whitebox vs. blade Infrastructure efficiency Set power budgets Demand challenge process Policy enforcement (faster provisioning helps) Developer mindset shift (box hugging to utility via OSIs) Active decom programs Organisational behaviour Fast RFP-less vendor discovery Annual supply-chain consolidation Contract & vendor management
  • 5.
    Infrastructure efficiency -examples DIFFICULTY (Tech and Org) low high ROI (all positive) high Lower 1b Realistic distributed compression ratios a. Virtualisation: Dev 10:1 b. Consolidation: DB 4.5:1 Web 3:1 MDS 2:1 N.B. Better than linear power-related savings 1. Global dynamic grids Many jobs aren’t latency sensitive no need to be in Metro tier 3 or 4 data centres. 50% saving migrating to tier 2 non metro area 2. Cool refresh e.g. Intel ‘Caneland’ 2 x performance per watt compared to Xeon MP. 3. White box preference over blade c30% Capex save. Debatable Opex delta Blade technology optimization around density rarely appropriate 4. 1a 2 4 3 5 Agile / Real-Time Infrastructure Potential for huge increases in utilisation through near-time provisioning SOI etc 5.
  • 6.
    M&E efficiency -examples DIFFICULTY (Tech and Org) low high ROI (all positive) high Lower Avoid host exhaust mixing with cold air a. Simple band aid steps, add blanking plates, don’t use perforated tiles in hot isles, remove dead under floor cables if possible b. For high density equipment consider containment units such as APC InfraStruXure Operating temperature 1 o temperature increase against typical data hall levels (20-22 o ) can = 4% energy saving – Sun Don’t exceed 25 o 1. Combined Heat & Power (CHP) Potential for 70 - 75% efficiencies at point of use versus central power station of circa 45% Heat by-product can improve cooling efficiencies Reduced C02 foot print 2. Hot & cold aisle containment 3. 1 2 3a 3b
  • 7.
    Fastest returns frominitiatives that minimise impact on application platform e.g. Cool refresh Further virtualisation Summary Controlling demand is easier than future clean-up Challenge placement and stove-piping of Grids
  • 8.
  • 9.
    About Citihub Citihubis a specialist IT infrastructure consultancy for the financial services industry Our people are proven industry practitioners with an appetite for complex challenges, particularly around applications infrastructure We're passionate about our clients' success; the industry’s leading companies come back to us time and again because we take ownership and we deliver results [email_address]