System Builder Marathon, March 2012: $650 Gaming PC

Deals
By Paul Henningsen
published

Benchmark Results: DiRT 3 And The Elder Scrolls V: Skyrim

DiRT 3

Unfortunately, we encountered a few kinks as we started testing with our new suite, resulting in data on this page getting excluded from our final performance tally. First, last quarter's machine shipped to its winner before I realized something was wrong with one set of results. Apparently, something changed with Patch 1.2 that threw consistency issues into the benchmark we were using for testing.

Clearly, DiRT 3 is well-threaded, as the Core i5 delivers a lead at every resolution, even at the High quality graphics preset. Very little scaling is seen on this quarter's PC because the Radeon HD 6950 is held back by our Core i3 processor.

It’s unfortunate we do not have comparison data because this story takes a massive change of direction at Ultra details with 8x AA. Based on other tests we've run, we’d expect the older machine to be limited by its Radeon HD 6870, averaging in the 40 FPS range at its stock settings.

The Elder Scrolls V: Skyrim

Patch 1.4 of The Elder Scrolls V: Skyrim, released in between last quarter and this one, delivered interesting performance gains. We’re going to include data for both rigs, but keep in mind that we can't compare the machines themselves, but rather just the improvements introduced by the update.

Prior to patch 1.4, Skyrim was quite CPU-limited. Our 25-second run though the town of Riverwood is demanding. However, exploring the Dwemer City of Markarth dropped frame rates even lower. So, it’s amazing that a Core i5-2400 was limiting our average performance to just over 40 FPS at Ultra settings.

After patch 1.4, this quarter's PC demonstrates host processor limitations at High detail settings. However, GPU limitations emerge at the Ultra preset with 8x MSAA. We included both sets of Skyrim data specifically to show how much the game’s performance profile has changed.

Game testing was planned for just these four new titles. But because we built this machine to improve the gaming experience, we couldn't draw suitable conclusions based on the two titles that gave us good data. Next, we’ll see how well the current PC matches up in our previous suite.

Paul Henningsen
87 Comments Comment from the forums
  • yukijin
    so now that all the 6950's are deactivated or $289+, is this build invalid? because a 7850 is looking really good right now...
    Reply
  • tristan_b
    What yukijin said.
    Reply
  • whysobluepandabear
    I appreciate what they're doing, but at some points, I can't help but feel like a cheap bitch.

    Making decisions overly measly amounts of money ($10) is just dumb. Work an extra day and just get the hardware you want. Or, don't go to the movies or out to eat for a few weeks.

    To me, there's a certain area, at which being cheap, just rips you off - you'd be better off spending a little more, and getting a much better item.
    Reply
  • How do I win this????
    Reply
  • mortsmi7
    Let me get this straight... you raised the budget $150 "as a result of steep price hikes on mechanical storage", then only spent $85 on a HDD. You really just wanted a more expensive graphics card. You could have taken the $70 processor savings and the $65 under-budget HDD savings and nearly have had a $500 build.
    Reply
  • de5_Roy
    very good read.
    nice to see where core i3's limits lie.
    i wonder if you guys will consider amd's new fx 6200 or fx 8120 for the $1200 build, with 78xx series in cfx.
    Reply
  • serhat359
    if I had $600 for a PC, I would go with i3-21xx, 6870, a better mobo and a better case.
    it is probably the best thing to do
    Reply
  • SpadeM
    whysobluepandabearI appreciate what they're doing, but at some points, I can't help but feel like a cheap bitch. Making decisions overly measly amounts of money ($10) is just dumb. Work an extra day and just get the hardware you want. Or, don't go to the movies or out to eat for a few weeks. To me, there's a certain area, at which being cheap, just rips you off - you'd be better off spending a little more, and getting a much better item.It's not an issues of whether they had the money or not, it's a matter of principle, you set your budget and goals at a certain point and then you make choices. Sure, not everyone will be happy with what they chose but that's what forums are for.

    Anyways, anything a bit over 60fps (on a 60hz monitor) really isn't that bad, i mean you might lack the bragging rights but at the end of the day, it's about gaming and feeling satisfied that you shot enough monsters. To further empathize that having 70 fps constant is not total shit because another GPU can serve you 130 (as if you're going to notice without watching the fps counter) my one suggestion for this SBM would be to introduce a different style of graphs. Below 30fps all the colors of the bars to be grey and over 60 the same thing. This to focus the attention on most relevant (to my opinion) segment. I've seen a lot of ppl chase those fps numbers, buying expensive GPUs only to have them sit in a bad enclosure, sub par motherboard or weak CPU.
    Even in gaming, i believe balance is key.
    Reply
  • confish21
    Great Job! These builds keep me at Tomshardware!

    Only thing 1 thing, you said an I3 was used instead of an I5 on this page...
    http://www.tomshardware.com/reviews/build-gaming-pc-overclock,3159-8.html
    You can check the 600 dec build here...
    http://www.tomshardware.com/reviews/core-i5-overclock-performance-gaming,3097.html
    Pretty sure an I5-2400 was used.
    Reply
  • jerreddredd
    I'm glad they used a i3 2120 for the CPU, but I wish they would have used some of the newer cards like the HD 7950 or the GTX 560 Ti 448. these are roughly the same price. Spending and extra $20 on a PSU was a waste. the EA430D and 380W are the core of the budget build. I would like to see some testing of a few of the less expensive PSU ($50 or less) to see which are junk and which aren't bad.
    Reply