THE SQL Server Blog Spot on the Web

Welcome to - The SQL Server blog spot on the web Sign in | |
in Search

John Paul Cook

CPU benchmarking and time for an upgrade

Is your SQL Server running slower than you’d like? Is it your SQL Server configuration or your slightly old hardware? We need to use tools to gather information.

My subjective impression is that my 2.2 GHz laptop outperforms my 3.0 GHz overclocked desktop in CPU bound tasks. Both are quad core machines. The desktop has a first generation quad and the laptop has a second generation quad core processor. To introduce objectivity, PassMark’s Performance Test was used to benchmark the two machines.


If you look under CPU Mark, you can see that the laptop’s second generation 2.2 GHz processor significantly outperforms the desktop’s first generation 3.0 GHz processor. The point here is that stated processor speeds aren’t everything. Comparing clock speeds from one generation of processor to another or from AMD to Intel is like comparing apples to oranges. We can drill down into the processor details and see where the differences are.


If you look under Disk Mark in the first screen capture, you can see that the laptop’s SSD is rated as being faster than the desktop’s 7200 rpm hard drive. Drilling down into the details indicates that the SSD’s advantages aren’t as pronounced for sequential reads.


It’s wise to periodically benchmark hardware. At a client site a few years ago, I used Performance Test to trace a performance problem back to the hardware. Before doing extensive performance tuning of a SQL Server, make sure that the hardware isn’t the culprit. You can’t tune away hardware problems.

Published Saturday, January 21, 2012 4:59 PM by John Paul Cook

Comment Notification

If you would like to receive an email when updates are made to this post, please register here

Subscribe to this post's comments using RSS



GrumpyOldDBA said:

It's good but I'm not sure you can conclude from a cpu test that a SQL Server will run faster. One critical factor is the speed of the memory for this is where access to the buffer cache will be. Linchi Shea has been running a series of very interesting posts on SQL Server performance, he's just posted results for a dual 6 core, if you compare these results to the results from the quad 10 core it make interetsing reading. Most of my prod servers run on quad quad core 2.9GHz servers, another team has a quad 10 core box ( at 2.2GHz ) with 8 cores my box was much faster at making backups - it was only when the backup ran across 40 cores that it outstripped mine ( using Litespeed )

I agree new generation cpus are better but it's a little more complex that that I suspect.

January 23, 2012 8:04 AM

John Paul Cook said:

I'm not suggesting that a SQL Server will run faster only because of a cpu test. Like you, I'm saying that the issue is complex. We need to use tools instead of making assumptions.

January 23, 2012 9:55 AM

Abraham said:

I want a tool to measure a system to check how fast SQL Server can run.

April 28, 2014 8:30 PM

Leave a Comment


About John Paul Cook

John Paul Cook is a database and Azure specialist in Houston. He previously worked as a Data Platform Solution Architect in Microsoft's Houston office. Prior to joining Microsoft, he was a SQL Server MVP. He is experienced in SQL Server and Oracle database application design, development, and implementation. He has spoken at many conferences including Microsoft TechEd and the SQL PASS Summit. He has worked in oil and gas, financial, manufacturing, and healthcare industries. John is also a Registered Nurse currently studying to be a psychiatric nurse practitioner. Contributing author to SQL Server MVP Deep Dives and SQL Server MVP Deep Dives Volume 2. Connect on LinkedIn

This Blog



Privacy Statement