Announcement

Announcement Module
Collapse
No announcement yet.

3570K vs 8350 Gaming Benchmarks w/ 8350

Page Title Module
Move Remove Collapse
X
Conversation Detail Module
Collapse
  • Filter
  • Time
  • Show
Clear All
new posts

  • 3570K vs 8350 Gaming Benchmarks w/ 8350

    AMD FX 8350 OC vs i5 3570k OC Using an EVGA GTX 670 - YouTube Thoughts?

  • #2
    Some numbers on that looks OK. Others look pretty wrong TBH. I haven't tested it myself, but pretty much anything over 1280x1024 is pretty much GPU dependent. It makes no sense why the 8350 would run a game 50% faster than another given identical graphics cards.

    Here's a review on the FX 8350 by Techreport. I think their results are more accurate. For example in Crysis 2...

    i5 3570K - 88 FPS
    i7 3960X - 87 FPS
    FX-8350 - 86 FPS
    i3-3225 - 82 FPS

    AMD's FX-8350 processor reviewed - The Tech Report - Page 8

    Even an i3 will do almost what the much, much, much faster i7 3960X will do at 1080P which just so happens to do just about same thing a FX 8350 does...

    I'm sure the guys in the video did their video with good intentions, but yeah... I'd at least like to see someone do the exact same testing they did and come up with the exact/similar results. Otherwise, some of the numbers I don't trust at all.

    Comment


    • #3
      Same. Hard to believe that the 8350 pushed 57 FPS on max Far Cry 3 where as the 3570K had 25 FPS, and the 3960x gets about 35 FPS on Far Cry 3. Both should be almost neck to neck. The EVGA Stinger is most likely throttling the CPU, and the results are just beyond questionable.
      Logan even said that most tech reviewers most likely lied and he was the only one that was right. I doubt Toms, Anandtech, etc would lie about their results. He said Hardware Canucks was a reputable site, but their results contradict HC's, not to mention HC's results are identically proportional to other review sites.

      I like Logan, but his testing methodology is just ridiculous. I remember in a 7850 review he did, the 7850 pushed over 50 FPS on Metro 2033, all max settings 1080p.

      Comment


      • #4
        Originally posted by eBombzor View Post
        Same. Hard to believe that the 8350 pushed 57 FPS on max Far Cry 3 where as the 3570K had 25 FPS, and the 3960x gets about 35 FPS on Far Cry 3. Both should be almost neck to neck. The EVGA Stinger is most likely throttling the CPU, and the results are just beyond questionable.
        Logan even said that most tech reviewers most likely lied and he was the only one that was right. I doubt Toms, Anandtech, etc would lie about their results. He said Hardware Canucks was a reputable site, but their results contradict HC's, not to mention HC's results are identically proportional to other review sites.

        I like Logan, but his testing methodology is just ridiculous. I remember in a 7850 review he did, the 7850 pushed over 50 FPS on Metro 2033, all max settings 1080p.
        Yeah, I'd definitely wait for someone else to do the exact same testing and come out with the exact same numbers before believing anything that dude says. The 8350 definitely isn't a bad part and would serve quite well for a gaming PC, but it would be a lie to say it outperforms the 3570K the way it did in the video. A couple frames faster here and there? Sure. That's margin of error. 2x the FPS? Nah.

        Wish I had the hardware to test this...

        Comment


        • #5
          Originally posted by sam.cpcr View Post
          Yeah, I'd definitely wait for someone else to do the exact same testing and come out with the exact same numbers before believing anything that dude says. The 8350 definitely isn't a bad part and would serve quite well for a gaming PC, but it would be a lie to say it outperforms the 3570K the way it did in the video. A couple frames faster here and there? Sure. That's margin of error. 2x the FPS? Nah.

          Wish I had the hardware to test this...
          Yea I would be suspicious if the 3570K had double the frames of the 8350. Both CPUs have similar performance so there shouldn't be a huge performance margin anywhere.

          Comment

          Working...
          X