Announcement

Collapse
No announcement yet.

Strange cases not representing reality.

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Strange cases not representing reality.

    7700XT is a weird one, according to passmark it's equivalent to 6700XT, but according to reviewers it's equivalent to 6800.
    Click image for larger version  Name:	1.png Views:	0 Size:	69.6 KB ID:	55968Click image for larger version

Name:	2.gif
Views:	355
Size:	81.9 KB
ID:	55970

  • #2
    The results are an average of all submissions. As such the more samples we get the more accurate the results become, whereas videocards with only a few samples have less accurate results. Radeon RX 7700 XT was recently released this month and there are only 8 samples (as of this post) and the other two videocards have thousands of benchmarks submitted.

    There is a wide range among the 8 samples and the rating should level out once more submissions are received.

    Click image for larger version

Name:	image.png
Views:	338
Size:	24.9 KB
ID:	55973

    Comment


    • #3
      Why is gtx 660 ti listed as having 3072mb memory when the cards are 2048?

      Comment


      • #4
        Note that the page says 'Max Memory Size', so the one GPU model may have multiple memory configurations but we only display the max.

        Edit: there are GTX 660 ti GPUs with 3GB of memory:
        https://www.techpowerup.com/gpu-spec...-ii-3-gb.b1630
        https://www.evga.com/products/specs/...c-e499a214e7aa

        Comment


        • #5
          So here's an example of large sample not representing reviewers(hardwareunboxed) results. This is almost a 2x difference in favor of B580 from RX 6650 XT. But with 1000 samples your data shows the complete opposite.
          Click image for larger version  Name:	1.jpg Views:	4 Size:	278.0 KB ID:	59261

          Like if I were to use this data to guide my purchase this would make no sense.
          Last edited by Freenoise; Jun-09-2025, 10:14 PM.

          Comment


          • #6
            With Intel cards, it is very dependent on what games and device drivers you use. The initial release of Intel's drivers was really awful. After dozens of releases they have now got better. But our results are a mix of all driver versions. The DirectX9,10 and 11 results were particularly awful, as Intel emulated some of the old APIs and put most of their optimization effort into DirectX12.

            So if you only look at the latest device drivers and only at DirectX12 games, the Intel card is great (as shown in our results).
            If you look at a mix of driver versions and DirectX versions, it doesn't look as good.

            We are planning on discontinuing our DIrectX9 test in our next major release, as DX9 is getting rather old now.

            Comment

            Working...
            X