Page 1 of 2 12 LastLast
Results 1 to 16 of 30

Thread: Problems with 7950 and windows 8? [tech report]

  1. #1
    Banhammer in peace PeterB kalniel's Avatar
    Join Date
    Aug 2005
    Posts
    30,749
    Thanks
    1,788
    Thanked
    3,288 times in 2,647 posts
    • kalniel's system
      • Motherboard:
      • Gigabyte Z390 Aorus Ultra
      • CPU:
      • Intel i9 9900k
      • Memory:
      • 32GB DDR4 3200 CL16
      • Storage:
      • 1TB Samsung 970Evo+ NVMe
      • Graphics card(s):
      • nVidia GTX 1060 6GB
      • PSU:
      • Seasonic 600W
      • Case:
      • Cooler Master HAF 912
      • Operating System:
      • Win 10 Pro x64
      • Monitor(s):
      • Dell S2721DGF
      • Internet:
      • rubbish

    Problems with 7950 and windows 8? [tech report]

    This stems from a previous tech report review where they found a 7950 performed quite badly under windows 8. That all the tests were conducted under windows 8 wasn't the point of the article, but they've responded to concern and this new article examines whether there is a link:

    http://techreport.com/review/24022/d...e-in-windows-8

  2. Received thanks from:

    scaryjim (12-12-2012)

  3. #2
    Moosing about! CAT-THE-FIFTH's Avatar
    Join Date
    Aug 2006
    Location
    Not here
    Posts
    31,621
    Thanks
    3,757
    Thanked
    5,060 times in 3,912 posts
    • CAT-THE-FIFTH's system
      • Motherboard:
      • Less E-PEEN
      • CPU:
      • Massive E-PEEN
      • Memory:
      • RGB E-PEEN
      • Storage:
      • Not in any order
      • Graphics card(s):
      • EVEN BIGGER E-PEEN
      • PSU:
      • OVERSIZED
      • Case:
      • UNDERSIZED
      • Operating System:
      • DOS 6.22
      • Monitor(s):
      • NOT USUALLY ON....WHEN I POST
      • Internet:
      • FUNCTIONAL

    Re: Problems with 7950 and windows 8? [tech report]

    It looks like the HD7950 struggles even worse under Windows 8 with AC3.

    " Although the Radeon's FPS average is much higher than the GeForce's, repeated latency spikes once again weigh it down, and rolling back to Windows 7 doesn't offer any relief.

    Both of these cards perform quite acceptably here, we should note. We've had to lower our "badness" threshold to 33.3 milliseconds since neither card surpasses the 50-millisecond mark for even a single frame. The thing is, the Radeon's higher FPS average suggests it's easily the faster solution for this workload, but some of its potential is wasted by the stubborn presence of higher-latency frames throughout the test run. "

    So,if both cards are fine,why change the conditions to show a difference which really make,well,no difference in Hitman??

    They are using 50MS for a reason in the other games,and not 33.3MS!
    Last edited by CAT-THE-FIFTH; 12-12-2012 at 04:14 PM.

  4. #3
    Senior Member
    Join Date
    Jun 2012
    Posts
    2,401
    Thanks
    87
    Thanked
    151 times in 145 posts
    • Willzzz's system
      • Motherboard:
      • Gigabyte
      • CPU:
      • 4670K
      • PSU:
      • FD Newton R3 600W
      • Case:
      • Corsair 350D

    Re: Problems with 7950 and windows 8? [tech report]

    33.3ms is the same as 30fps, this is generally regarded as the minimum acceptable frame rate, so it seems a sensible threshold?

    It is curious how inconsistent the data is, some games do perform better in Win7 others perform worse.

  5. #4
    Moosing about! CAT-THE-FIFTH's Avatar
    Join Date
    Aug 2006
    Location
    Not here
    Posts
    31,621
    Thanks
    3,757
    Thanked
    5,060 times in 3,912 posts
    • CAT-THE-FIFTH's system
      • Motherboard:
      • Less E-PEEN
      • CPU:
      • Massive E-PEEN
      • Memory:
      • RGB E-PEEN
      • Storage:
      • Not in any order
      • Graphics card(s):
      • EVEN BIGGER E-PEEN
      • PSU:
      • OVERSIZED
      • Case:
      • UNDERSIZED
      • Operating System:
      • DOS 6.22
      • Monitor(s):
      • NOT USUALLY ON....WHEN I POST
      • Internet:
      • FUNCTIONAL

    Re: Problems with 7950 and windows 8? [tech report]

    Quote Originally Posted by Willzzz View Post
    33.3ms is the same as 30fps, this is generally regarded as the minimum acceptable frame rate, so it seems a sensible threshold?
    The other games are 50MS BTW,and they have only used 33.3MS for Hitman.

    "Both of these cards perform quite acceptably here, we should note. We've had to lower our "badness" threshold to 33.3 milliseconds since neither card surpasses the 50-millisecond mark for even a single frame. "

    So basically a pointless test,otherwise 33.3MS would have been used for all games.

    Edit!!

    Moreover,why not put in the 50MS results too??

  6. #5
    Senior Member
    Join Date
    Jun 2012
    Posts
    2,401
    Thanks
    87
    Thanked
    151 times in 145 posts
    • Willzzz's system
      • Motherboard:
      • Gigabyte
      • CPU:
      • 4670K
      • PSU:
      • FD Newton R3 600W
      • Case:
      • Corsair 350D

    Re: Problems with 7950 and windows 8? [tech report]

    No, that's not the case. You have to set your threshold based on the performance of the card, otherwise you'd end up with either 0% or 100% both of which would be meaningless.

    If they were testing a 7850 then they would use a 50ms test, if it was a 7970 they might use a 16.7ms test.

    The test is determined by the expectations that consumers have of the cards. At this level you would hope to get a minimum of 30fps at all times, if you had paid more you might expect a minimum of 60fps.

  7. #6
    Not a good person scaryjim's Avatar
    Join Date
    Jan 2009
    Location
    Gateshead
    Posts
    15,196
    Thanks
    1,231
    Thanked
    2,291 times in 1,874 posts
    • scaryjim's system
      • Motherboard:
      • Dell Inspiron
      • CPU:
      • Core i5 8250U
      • Memory:
      • 2x 4GB DDR4 2666
      • Storage:
      • 128GB M.2 SSD + 1TB HDD
      • Graphics card(s):
      • Radeon R5 230
      • PSU:
      • Battery/Dell brick
      • Case:
      • Dell Inspiron 5570
      • Operating System:
      • Windows 10
      • Monitor(s):
      • 15" 1080p laptop panel

    Re: Problems with 7950 and windows 8? [tech report]

    Quote Originally Posted by CAT-THE-FIFTH View Post
    So,if both cards are fine,why change the conditions to show a difference which really make,well,no difference ...
    To demonstrate that there is a difference. Given tr's change to latency-based benchmarks is still relatively recent, they're still playing with the best way to present the results. They've previously provided time-beyond graphs at 16.7ms, 33.3ms and 50ms, with some justification, but I imagine there's a lot of work goes into producing their reviews and I don't begrudge them picking charts that show differences. They're very transparent about it, pointing out where they've had to move to a lower threshold to show the difference between cards, and they could've decided just to go for a flat 16.7ms (~ 60fps). Instead, they use a sensible measure for the cards and games they're reviewing, altering it from game to game as necessary (e.g. in the linked review, MoH:Warfighter uses 16.7ms threshold).

    EDIT: CAT; AC3 and Hitman use 33.3ms. MoH and Skyrim - whiterun use 16.7ms. They change threshold to demonstrate a difference. Whilst no frames longer than 50ms is an indication that the card isn't *bad*, no frames longer than 16.7ms would be an indication that the card is excellent. If a review is able to show a difference between cards, then it should do so, IMNSHO

  8. #7
    Moosing about! CAT-THE-FIFTH's Avatar
    Join Date
    Aug 2006
    Location
    Not here
    Posts
    31,621
    Thanks
    3,757
    Thanked
    5,060 times in 3,912 posts
    • CAT-THE-FIFTH's system
      • Motherboard:
      • Less E-PEEN
      • CPU:
      • Massive E-PEEN
      • Memory:
      • RGB E-PEEN
      • Storage:
      • Not in any order
      • Graphics card(s):
      • EVEN BIGGER E-PEEN
      • PSU:
      • OVERSIZED
      • Case:
      • UNDERSIZED
      • Operating System:
      • DOS 6.22
      • Monitor(s):
      • NOT USUALLY ON....WHEN I POST
      • Internet:
      • FUNCTIONAL

    Re: Problems with 7950 and windows 8? [tech report]

    Quote Originally Posted by Willzzz View Post
    No, that's not the case. You have to set your threshold based on the performance of the card, otherwise you'd end up with either 0% or 100% both of which would be meaningless.
    Quote Originally Posted by scaryjim View Post
    To demonstrate that there is a difference. Given tr's change to latency-based benchmarks is still relatively recent, they're still playing with the best way to present the results. They've previously provided time-beyond graphs at 16.7ms, 33.3ms and 50ms, with some justification, but I imagine there's a lot of work goes into producing their reviews and I don't begrudge them picking charts that show differences. They're very transparent about it, pointing out where they've had to move to a lower threshold to show the difference between cards, and they could've decided just to go for a flat 16.7ms (~ 60fps). Instead, they use a sensible measure for the cards and games they're reviewing, altering it from game to game as necessary (e.g. in the linked review, MoH:Warfighter uses 16.7ms threshold).
    Sorry,no it is meaningless what they did.

    "Both of these cards perform quite acceptably here, we should note. We've had to lower our "badness" threshold to 33.3 milliseconds since neither card surpasses the 50-millisecond mark for even a single frame. "

    So,why not show 33.3MS results for all the games and not just one??

    Its like saying a camera which produces a 3010X3010 pixel image can be used to print images bigger than a camera which produces 3000X3000 images.

    Pointless.

  9. #8
    Not a good person scaryjim's Avatar
    Join Date
    Jan 2009
    Location
    Gateshead
    Posts
    15,196
    Thanks
    1,231
    Thanked
    2,291 times in 1,874 posts
    • scaryjim's system
      • Motherboard:
      • Dell Inspiron
      • CPU:
      • Core i5 8250U
      • Memory:
      • 2x 4GB DDR4 2666
      • Storage:
      • 128GB M.2 SSD + 1TB HDD
      • Graphics card(s):
      • Radeon R5 230
      • PSU:
      • Battery/Dell brick
      • Case:
      • Dell Inspiron 5570
      • Operating System:
      • Windows 10
      • Monitor(s):
      • 15" 1080p laptop panel

    Re: Problems with 7950 and windows 8? [tech report]

    Quote Originally Posted by CAT-THE-FIFTH View Post
    Sorry,no it is meaningless what they did.

    "Both of these cards perform quite acceptably here, we should note. We've had to lower our "badness" threshold to 33.3 milliseconds since neither card surpasses the 50-millisecond mark for even a single frame. "
    How is it meaningless when they explain what they've done? They have demonstrated a difference! Neither card is *bad*, but the latencies show the 660Ti is better. The 7950 spends more time at below 30fps. That's a reasonable difference to highlight! It would be more meaningless to publish a chart showing neither card took more than 50ms to render a frame, because that doesn't tell you anything about the difference between the cards!

  10. #9
    Moosing about! CAT-THE-FIFTH's Avatar
    Join Date
    Aug 2006
    Location
    Not here
    Posts
    31,621
    Thanks
    3,757
    Thanked
    5,060 times in 3,912 posts
    • CAT-THE-FIFTH's system
      • Motherboard:
      • Less E-PEEN
      • CPU:
      • Massive E-PEEN
      • Memory:
      • RGB E-PEEN
      • Storage:
      • Not in any order
      • Graphics card(s):
      • EVEN BIGGER E-PEEN
      • PSU:
      • OVERSIZED
      • Case:
      • UNDERSIZED
      • Operating System:
      • DOS 6.22
      • Monitor(s):
      • NOT USUALLY ON....WHEN I POST
      • Internet:
      • FUNCTIONAL

    Re: Problems with 7950 and windows 8? [tech report]

    Quote Originally Posted by scaryjim View Post
    How is it meaningless when they explain what they've done? They have demonstrated a difference! Neither card is *bad*, but the latencies show the 660Ti is better. The 7950 spends more time at below 30fps. That's a reasonable difference to highlight!
    No it isn't. Why not show it for ALL the games tested??

    Its inconsistent.

  11. #10
    Not a good person scaryjim's Avatar
    Join Date
    Jan 2009
    Location
    Gateshead
    Posts
    15,196
    Thanks
    1,231
    Thanked
    2,291 times in 1,874 posts
    • scaryjim's system
      • Motherboard:
      • Dell Inspiron
      • CPU:
      • Core i5 8250U
      • Memory:
      • 2x 4GB DDR4 2666
      • Storage:
      • 128GB M.2 SSD + 1TB HDD
      • Graphics card(s):
      • Radeon R5 230
      • PSU:
      • Battery/Dell brick
      • Case:
      • Dell Inspiron 5570
      • Operating System:
      • Windows 10
      • Monitor(s):
      • 15" 1080p laptop panel

    Re: Problems with 7950 and windows 8? [tech report]

    See my earlier post. They use 16.7ms for two test, 33.3ms for two tests. They *always* choose the highest theshold that produces a difference.

  12. #11
    Banhammer in peace PeterB kalniel's Avatar
    Join Date
    Aug 2005
    Posts
    30,749
    Thanks
    1,788
    Thanked
    3,288 times in 2,647 posts
    • kalniel's system
      • Motherboard:
      • Gigabyte Z390 Aorus Ultra
      • CPU:
      • Intel i9 9900k
      • Memory:
      • 32GB DDR4 3200 CL16
      • Storage:
      • 1TB Samsung 970Evo+ NVMe
      • Graphics card(s):
      • nVidia GTX 1060 6GB
      • PSU:
      • Seasonic 600W
      • Case:
      • Cooler Master HAF 912
      • Operating System:
      • Win 10 Pro x64
      • Monitor(s):
      • Dell S2721DGF
      • Internet:
      • rubbish

    Re: Problems with 7950 and windows 8? [tech report]

    It's not pointless. They just lower the threshold until you start to see a difference. No point showing 33ms if you are already seeing a difference at 50ms. The fact that they had to lower to 33ms should indicate that there is no difference at 50ms.

    The question is whether there is a difference, and they're answering that fine.

    It's bad news for the 7950, with windows 7 showing up bad latencies in almost as many cases as windows 8. They have communicated this with AMD, who have said they are looking into it with some alarm. Let's hope drivers can sort this out. In the meantime I think TR deserve some praise for addressing reader's concerns so quickly and attempting to fairly show an interesting result.

  13. #12
    Senior Member
    Join Date
    Jun 2012
    Posts
    2,401
    Thanks
    87
    Thanked
    151 times in 145 posts
    • Willzzz's system
      • Motherboard:
      • Gigabyte
      • CPU:
      • 4670K
      • PSU:
      • FD Newton R3 600W
      • Case:
      • Corsair 350D

    Re: Problems with 7950 and windows 8? [tech report]

    Acceptably is subjective for most people.

    But you can look at the graph if you want to choose your own 'badness' threshold.

    What would be the point in choosing 50ms in this case?

  14. #13
    Moosing about! CAT-THE-FIFTH's Avatar
    Join Date
    Aug 2006
    Location
    Not here
    Posts
    31,621
    Thanks
    3,757
    Thanked
    5,060 times in 3,912 posts
    • CAT-THE-FIFTH's system
      • Motherboard:
      • Less E-PEEN
      • CPU:
      • Massive E-PEEN
      • Memory:
      • RGB E-PEEN
      • Storage:
      • Not in any order
      • Graphics card(s):
      • EVEN BIGGER E-PEEN
      • PSU:
      • OVERSIZED
      • Case:
      • UNDERSIZED
      • Operating System:
      • DOS 6.22
      • Monitor(s):
      • NOT USUALLY ON....WHEN I POST
      • Internet:
      • FUNCTIONAL

    Re: Problems with 7950 and windows 8? [tech report]

    Quote Originally Posted by scaryjim View Post
    See my earlier post. They use 16.7ms for two test, 33.3ms for two tests. They *always* choose the highest theshold that produces a difference.
    So,when is it a really valid difference?? Testing for the point for the testing is pointless in its own right. They talk about people measurebating about framerates as being a wrong thing,then measurebate about latencies.

    Why not link to 16.7,33 and 50 MS?? Then people can make their own minds up?? At least in some games,you can see a difference as they say it can be felt,ie,stuttering in Skyrim and that has been already mentioned by users elsewhere.

    But,if there is no noticeable difference,why lower the threshold to emphasis a difference which appears to do nothing?? Is the difference noticeable in blind testing??
    Last edited by CAT-THE-FIFTH; 12-12-2012 at 04:48 PM.

  15. #14
    Not a good person scaryjim's Avatar
    Join Date
    Jan 2009
    Location
    Gateshead
    Posts
    15,196
    Thanks
    1,231
    Thanked
    2,291 times in 1,874 posts
    • scaryjim's system
      • Motherboard:
      • Dell Inspiron
      • CPU:
      • Core i5 8250U
      • Memory:
      • 2x 4GB DDR4 2666
      • Storage:
      • 128GB M.2 SSD + 1TB HDD
      • Graphics card(s):
      • Radeon R5 230
      • PSU:
      • Battery/Dell brick
      • Case:
      • Dell Inspiron 5570
      • Operating System:
      • Windows 10
      • Monitor(s):
      • 15" 1080p laptop panel

    Re: Problems with 7950 and windows 8? [tech report]

    Any difference is valid, as long as you point out what the difference is. The graphs are clearly labelled.

    They used to run graphs at all three latencies back when they started doing latency-based reviews. I assume it takes a long time though, and this article has been thrown together pretty quickly in response to people (like me ) who were unhappy at their lackadaisical approach to any difference Win 8 might make in the previous test. They tell you in the commentary that there's no difference at 50ms, so it's not like that information is hidden. They're assuming a certain level of intelligence from their readers, and that they can make their own decision from the information provided. Plus they don't use the "time beyond" graphs in their summary - that's based on 99th percentile measures. The "time beyond" graphs are purely an extra, for people to draw their own conclusions from. Would you prefer that they weren't provided at all?

    If they offered 3 time beyond graphs, how long would it be before people started complaining that there was too big a difference between 16.7ms and 33.3ms, and that they should do 20ms and 25ms graphs too (representing 50fps and 40fps) as that's where the really noticable differences are? They can't do everything for every review. As it is, they present a much better representation of game performance than pretty much any other site, by looking at overall frame times and latencies rather than crude averages.

    If the 33.3ms graph showed the AMD card in a better light, would you be complaining about their use of it? This is coming across a lot like "They say the AMD card is good enough, so why do they have to show it losing in a graph?". Bottom line is the 7950 isn't as smooth at rendering any of those games as the 660Ti, and it's a little worse in Win 7 than in Win 8. It doesn't matter how you measure that - it's a fact.

  16. #15
    Moosing about! CAT-THE-FIFTH's Avatar
    Join Date
    Aug 2006
    Location
    Not here
    Posts
    31,621
    Thanks
    3,757
    Thanked
    5,060 times in 3,912 posts
    • CAT-THE-FIFTH's system
      • Motherboard:
      • Less E-PEEN
      • CPU:
      • Massive E-PEEN
      • Memory:
      • RGB E-PEEN
      • Storage:
      • Not in any order
      • Graphics card(s):
      • EVEN BIGGER E-PEEN
      • PSU:
      • OVERSIZED
      • Case:
      • UNDERSIZED
      • Operating System:
      • DOS 6.22
      • Monitor(s):
      • NOT USUALLY ON....WHEN I POST
      • Internet:
      • FUNCTIONAL

    Re: Problems with 7950 and windows 8? [tech report]

    Quote Originally Posted by scaryjim View Post
    stuff
    You still have not answered my question though.

  17. #16
    Senior Member
    Join Date
    Jun 2012
    Posts
    2,401
    Thanks
    87
    Thanked
    151 times in 145 posts
    • Willzzz's system
      • Motherboard:
      • Gigabyte
      • CPU:
      • 4670K
      • PSU:
      • FD Newton R3 600W
      • Case:
      • Corsair 350D

    Re: Problems with 7950 and windows 8? [tech report]

    Quote Originally Posted by CAT-THE-FIFTH View Post
    Why not link to 16.7,33 and 50 MS?? Then people can make their own minds up?? At least in some games,you can see a difference as they say it can be felt,ie,stuttering in Skyrim and that has been already mentioned by users elsewhere.
    As I said above, you can do this.

    Look at the graph, choose whichever latency you want to pick, read across and you can see which cards have most peaks above that latency.

Page 1 of 2 12 LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •