Test Configuration

Software Configuration

Windows 2003 was configured with /3GB and /PAE switches in the boot.ini to support the 8GB of memory used for our tests. SQL Server Enterprise was set to use AWE extensions, and a maximum memory limit was set at 6144MB.

Intel Bensley 3.46 Pre-Production System
Dual 3.46GHz Dual-Core Dempsey Processors
Pre-production Blackford based Intel Motherboard
8GB FBDIMM DDR-2 533Mhz
Windows 2003 Enterprise Server (32 Bit) SP1
8 x 36GB 15,000RPM Ultra320 SCSI drives in RAID-0
LSI Logic 320-2 SCSI Raid Controller

Opteron 280 System
Tyan S2882 K8S Motherboard
Dual Opteron 280 (Dual-Core) Processors
8GB Corsair PC3200 DDR
Windows 2003 Enterprise Server (32 Bit) SP1
8 x 36GB 15,000RPM Ultra320 SCSI drives in RAID-0
LSI Logic 320-2 SCSI Raid Controller

Measuring Power

To measure power consumption of each system, we used an EXTECH Instruments Power Analyzer Model 380803. This power analyzer allows us to view current power consumption, and log the consumption at various intervals during a test to a text file. For this test, we used the same Power Supply for both systems, although we recorded the difference between a 750W power supply and a 550W power supply, and it was less than 3 Watts. We should note that the Raid Array was powered by a separate power supply that was not plugged into our analyzer, so we were measuring strictly bare system power consumption. If you’re curious, the Raid Array used about 98 Watts spun up, and averaged 110 Watts during the database tests.

Idle – To measure a system at idle, we booted each system into Windows and let it stabilize by watching the task manager in Windows and the Wattage readings. Once we were at a stable reading, we began recording for 100 iterations of our data logger (which logs every 2 seconds). We then took those numbers and averaged them to get the idle power reading.

50% – We used our database benchmark to measure a loaded system, by adjusting the thread count for the test to a level that produced a half loaded system. Then, we would run our database test for its duration while recording to the data logger. Finally, we averaged those results.

100% – To produce a fully loaded system, we used the same technique as above, except increasing the number of threads until we achieved a fully loaded system.

The future is performance per Watt. Database Benchmark Results
POST A COMMENT

67 Comments

View All Comments

  • gjmck - Monday, December 19, 2005 - link

    I'm curious that the numbers dont reflect the true difference between equivalently configured Intel vs. Opteron systems.

    The Dempsey processor TDP max is 130W and the Opteron is 95W. That difference is only 35W. The memory controller needed by Dempsey should only consume 60 - 80W. Using 80W that gives the maximum total difference between two eqivalently configured systems as 80 + 35 = 115W.

    Yet in the max processor utilization tests the difference was 214 Watts. So where is the extra 99 Watts being used? FBD? If so then when Opteron uses similar memory technology the delta will not be as great.

    Gregg McKnight
    Reply
  • Furen - Thursday, December 22, 2005 - link

    Intel's TDPs reflect "typical" power draw, while AMD's reflects the "worst-case scenario" power consumption, so they're not directly comparable. I very much doubt the memory controller uses even close to 80W, I'd say something like 20-30W for the whole northbridge is reasonable. FB does use more power, but that shouldn't be more than 5-10W per dimm. The rest is just the CPU being insanely power-inefficient. Reply
  • dannybin1742 - Friday, December 16, 2005 - link

    to keep anthing at a constant temperature, the heat going into the system must equal the heat being taken away. so if one system uses 200W of power, first you have the cost of the 200W, then you have the cost to remove the 200W of heat given off by the use of the system. on top of this air, conditioners are 20-25% efficient at best (if i remember correctly), so the amount of power used to remove the heat generated would take 3-4X more energy to remove. so in essence you are looking at at LEAST 2X amount of money calculated in the article. (i took a year of thermodynamics at school here, when i was an undergrad) in reality, you are probably looking at 4-6X to run and remove the heat from the data center. they should have looked at the opteron 2.2ghz HE (low voltage) i'd be interested to see what power numbers those put up.

    also, was winxp 2003 server 64 bit? or were all the tests run in 32 bit? i just skimmed over the article. how about linux?
    Reply
  • coldpower27 - Friday, December 16, 2005 - link


    Opteron 270 HE is the highest of the lower wattage 2 Way Opterons and it runs at 2.0GHZ.

    Reply
  • Viditor - Friday, December 16, 2005 - link

    quote:

    Opteron 270 HE is the highest of the lower wattage 2 Way Opterons and it runs at 2.0GHZ

    You mean 2 way dual core...
    The 250 HE is single core at 2.4 GHz...
    Reply
  • coldpower27 - Friday, December 16, 2005 - link

    Yes, I assume 2 Dual Core vs 2 Dual Core. Reply
  • haris - Friday, December 16, 2005 - link

    One question that kept nagging me was "How many "threads" were required to get the systems to each load level?" How much of a difference would it make to performance/watt if you have to take into account that processor 1 is also handling x% more/less threads then processor 2? Reply
  • Jason Clark - Friday, December 16, 2005 - link

    That will teach me for just taking a $1,000 measurement devices reported figures :) It actually figures out the cost, which obviously was wrong. I've updated the numbers, they should be correct.


    Again, sorry :)
    Reply
  • coldpower27 - Friday, December 16, 2005 - link

    Thanks alot.:) Reply
  • Biffa - Friday, December 16, 2005 - link

    With over a 1Ghz defecit (yes I know) in processor speed, and with only 1MB of cache per core rather than 2MB, I think we can safely say that Intel is still clutching at straws at this level of the game.

    Good PR on their part (always admired them for that) however its a crying shame that after all this time this is the best they can do.
    Reply

Log in

Don't have an account? Sign up now