Blade Server TCO and Architecture – You Cannot Separate Them

I have heard this a lot over the years, in one way or another – “The only price that really counts is what I actually pay for my server.” 

Alright, so why bother with a TCO analysis?  The truth is that server acquisition costs only contribute 20% (or less) to a 3 year server TCO.  Management and other OpEx costs contribute the remaining 80%.  If you go to 5 years, the acquisition cost starts to fade into obscurity.

There are a number of studies you can find online that call out server acquisition cost at 15% to 17% of TCO, or even less.  One is an Information Week report that quotes a 2007 IDC study.  The Information Week article is very good, with multiple sources and definitely worth a read.  Since 2007 there have been myriad improvements in processor performance, as well as, server and architectural innovations (Cisco UCS).  All of these supply ample rationale for a low CapEx component for Server / Data Center Total Cost of Ownership, see the figures below.  

[The WW Server Related Spend… chart is from IDC, “New Econmoinc Model of the Datacenter”; IDC 2011]  [Only the graph is from the cited source, the table is my analysis of the numbers presented by the graph.]

Summary of the figures above:

• Server purchase spend and associated power & cooling spending is flat (red and green bands above)
• Physical server management cost is the down (blue aband bove)
• Virtual server management cost are way up and increasing (orange band above)

 

THE BOLD STATEMENT

The actual acquisition cost of a server is irrelevant.  Let’s say it this way: The acquisition cost difference between server vendors is irrelevant.  Better?  What makes or breaks data center economics are the OpEx costs (primarily management) and these significantly impact business success and profitability.  That’s a pretty bold statement, so let’s look at why I think this is true.

Rationale

Inputs:  percentage allocation between CapEx and OpEx follow the IDC paper cited above.

• $100.00 
• 20% CapEx (server acquisition) and
• 80% OpEx (everything else). 

Analyze:

• Server acquisition cost is 20% higher than a competitor –  very unlikely in today’s competitive marketplace. 

[Note:  A company can always ‘pay you’ to buy their servers, but I prefer to use real world, repeatable cost structures to make important decisions.  Once-in-a-lifetime special deep discount giveaways always have a hidden cost.  In other words a free server is not actually free. As Robert Heinlein said and I paraphrase ‘TANSTAAFL – There ain’t no such thing as a free lunch’.]

• OpEx cost can be lowered 20%  –  and very likely more, since the vast majority of this cost is server management related (both physical and virtual).  Cisco UCS Management  tools  are unparalled in the delivery of benefits here.

Result:

• CapEx (Server) going up only marginally increases TCO (even a 20% cost increase).
• OpEx going down dramatically impacts TCO (primary OpEx driver is management)

     Think about it in a more general way:

• CapEx is a singular event.  A one-time event for a specific server(s).
• OpEx is a recurring activity.  It is the thing that keeps on giving (or taking) every month that makes the real difference.
• TCO is driven almost exclusively by a high contribution, recurring component – OpEx.

Conclusion:

• Server cost is irrelevant because: changing its contribution to total TCO has a vanishingly small impact. 
• It actually pays you to pay more for the correct hardware.  What a concept! 

Value is More Important than Price.

• Value = benefit derived from money spent, or in clear speech –
• Value = More bang for your buck.

I absolutely agree that a server TCO analysis is important and Cisco has the tools.

• An easy to use public tool, the Cisco UCS TCO/ROI Advisor; and,
• A more detailed analytical TCO tool that your Cisco team can complete with your assistance.

Now that we have agreed  that server cost is just not relevant to a TCO analysis and the buying decision process, what is?  

It’s Architecture. 

To me, architecture is the critical combination of form, funtion(ality), and design.  It is the single most important component of OpEx, because architecture defines management capability.  It is crucial to choose a server that is designed from day one as part of an intentional architecture. This architecture must maximize flexibility and scale, increase management capability and collaboration, and reduce complexity. 

In other words, the architecture needs to be flexible, scalable, easily managed and dynamic, so it can respond to the needs of your business.  Take a look at Cisco UCS and UCS Manager and see what a truly architecture built from the ground up, Unified Computing System, can do for you.