Computer Ed Radio

Green Gaming PC Conclusion (Part 1)

The concept was simple, the goal was to find the best mix of components for providing good gaming performance, energy efficient and reasonably priced. I must admit that I approached this build with some pre-conceived ideas of how I thought energy usage of a PC would work. Those where quickly destroyed as I began my testing.

When thinking of a Green PC most people would start with the PSU however that is actually one of the easiest things to look at. I decided to begin with the CPU.

For my initial testing I began with a base system as follows:

• CPU: Phenom II 965
• 4 Gig DDR3 RAM
• Gigabyte 770 based Motherboard
• Sapphire Vapor-X 5770
• Antec Earthwatt 500 PSU
• WB Caviar Black 640 Gig HD

This base system can provide an outstanding gaming experience at resolutions of 1680×1050, so I knew it had the horse power I needed. With this in mind I set about taking readings of the wattage pull the system provided from the wall. For measuring the wattage I used a Kill-A-Watt between the PC and the surge protector.

Once I attained my baseline numbers I found my preconceptions quickly falling away. This system according to a PSU calculator is supposed to max out at near 400 watts. With this in mind I expected that a run of a hard video game would come at least close to full load. However running Dragon Age and Dirt 2 to provide some examples of newer games I was surprised at little little load was put on the PSU. So after double and triple checking my numbers I came to realize my original thoughts where wrong and moved on to test my next CPU.

The method of testing was fairly simple. Each CPU was first tested to see if it could meet minimum performance requirements of providing over 30 FPS at 1680×1050 for the gaming with a good detail level and AA enabled. Once a CPU passed that test I moved on the power usage testing to see what difference in total system load was made.

The CPUs used for this test where the Phenom II 965, 945 (95 TDP) and 910e. I also tested the Athlon II X4 630. Some quick power consumption reports are below.

 

Now looking at those numbers you would be quick to conclude that the 910e delivered by far the most energy efficient load and you would be right. However a close examinations of what those numbers mean when compared to price might lead to a different conclusion. According to the Department of Energy the average cost of power in the US for a home was $0.12 per KWH. Using this number lets see what we can save with the 910 over the other end of the scale, the 965.

If we assume a 100% for an 8 hour period every day we will see a savings of $25.93. Not as much as you would think is it? The kicker here though is that it is HARD to push a PC to a full load for extended periods. So lets try again with some new numbers, lets assume you play Dragon Age for 8 hours a day, what is the difference then? Eight hours of DAO played 365 days a year on a 910e will save you $15.77 over using a 965 processor.

Okay but you say, I still save money and use less energy, that was the point right? Really? Lets consider these numbers but this time lets use the 910e and compare it to the Athlon II 630. The Athlon chip can be bought at the price of about $100, the 910e is coming in at around $175. So lets look again at that 8 hour a day full load run. Running at full load the 910e will save you an average of $7.71 over the 630. When you consider the price difference is $75 you can see you would need about 10 years for the 910e to save you enough money in power usage to make up for the cost difference of the chip.  So much for saving power saving you money.

A comparison of the two chips found the 630 was actually faster in most testing. The lack of an L3 cache was offset by the higher clock speed of the 630. Both are quad core and provide excellent multitasking and both can easily handle most of today’s games at the 1680 resolution with a good graphics card.

While the 910e does deliver a lower power load and produces less heat as well the fact that you can save almost half the price get equal or better performance and show only a minor difference in CPU power consumption the clear choice for me was the Athlon II X4 630.

Now I want you to understand the numbers I used for this calculation where a bit extreme. To understand what I mean I setup my Kill-A-Watt to record and report the KWH usage of my PC over a 48 period of Saturday and Sunday. During my weekends I do a lot of gaming, being it MMOs or games like Dragon Age, Demigod or others. I also do a lot of work with audio files preparing for the show on Sunday and a ton of web usage. In other words my usage over the weekend is likely a bit harder than the average user but will still give solid base data. With this in mind my 965 based system came in at 3.5 KWH, the 910e delivered 3.3 KWH. Using these numbers and extending out to a year we see that the 910e will save us $4.50. In other words not enough to matter.

The reason for these close numbers is simple, I used Cool-n-Quiet. This combined with sleep modes using Windows 7 set at Balanced provides some excellent energy savings and brings the processor more out of the equation for energy savings. In fact so much so that The 945 which is faster and cheaper than the 910e is able to deliver very close to the same energy load as the 910e and 630.

At the end of the day the processor can make some difference in the power but the way the system is used makes a much bigger difference in the power consumption of the system than the CPU. With this in mind the extra cost of the “e” series of CPUs is not justified in everyday use, let alone a gaming machine.

On Monday I will post my findings when comparing GPUs and PSUs and on Wednesday I will post my final completed build in it’s entirety.

March 20, 2010 Posted by Computer Ed | Uncategorized | Leave a Comment