Radeon Crimson 16.2 drivers are ready for Ashes of the Singularity

AMD has just released its Radeon Software Crimson Edition 16.2 hotfix drivers. This version comes with optimizations for the Ashes of the Singularity Benchmark 2.0. The company points out that this benchmark uses multiple DirectX 12 features, including asynchronous shaders, multi-threaded command buffers, and command buffer reordering.

The Crimson 16.2 hotfix also includes performance optimizations for Rise of the Tomb Raider and the SteamVR Performance Test. AMD says its Radeon R9 390, Nano, and Fury series GPUs all reach "VR recommended" status in the SteamVR test, and it reiterates that CrossFire configurations using its Affinity multiGPU feature deliver significant performance scaling in that test.

This release also includes CrossFire improvements for The Division and XCOM 2. Gamers playing Fallout 4 and Rise of the Tomb Raider on AMD cards should be happy to know that the 16.2 drivers fix multiple bugs with both games, too. AMD also includes a fix for a black-screen issue when booting with an HDMI monitor connected to systems with a mix of Intel and AMD GPUs, as well as gameplay choppiness in CrossFire configurations using FreeSync.

To download the new drivers or check out the release notes, hit this link here.

Comments closed
    • Shobai
    • 4 years ago

    I’m pretty late to the party in commenting on this news item, but hopefully someone can help: the Radeon Settings window doesn’t remember its position. Every time I open it, it pops up in the middle of my primary monitor, so I drag it across to the secondary in order to see it while in game, etc.

    Any ideas on how to make it remember? Is there anything I can do on my end? Or will that be something AMD needs to sort out?

      • brucethemoose
      • 4 years ago

      Use an alternative, maybe? What do you use the Radeon Settings menu for anyway?

        • Shobai
        • 4 years ago

        Downclocking, only. Previous experience has been that the post-Rivatuner apps [Trixx, Afterburner, Asus’] are happy going up, but not down – and even for RT, you had to modify a config file to get it to work.

    • brucethemoose
    • 4 years ago

    I don’t see it in the release notes, but these drivers finally fixed the broken memory clock slider.

      • Shobai
      • 4 years ago

      [edit: clarity]

      Is this in relation to the second last dot point under “Resolved Issues”? Or something else?

        • brucethemoose
        • 4 years ago

        [quote<] The AMD Overdrive™ memory clock slider does not show original clock values if memory speeds are overclocked [/quote<] Yea, that's it, didn't see it before. What that means is you could only increase the memory slider, not decrease it, up until now.

          • Shobai
          • 4 years ago

          That’s quite interesting: I’ve got an R9 290, and I’ve been using the memory slider to drop the RAM down to the lowest I can [150MHz] whenever I’m not playing games [I drop the GPU clock also, but that seems unrelated]. I bring it back up to stock, which is 1260MHz for my card, in order to game.

          I haven’t had any issues with decreasing the slider thus far, but is that because I haven’t tried upping the clock past stock?

            • brucethemoose
            • 4 years ago

            Didn’t work right with my 7950, at least. Whenever I upped the memory slider, whatever I set it to became the new lower limit.

            At 1575mhz (max in the Radeon Settings menu), it disappeared entirely. Seems to be fixed now though.

    • chuckula
    • 4 years ago

    Do these drivers fix the errors in DirectFlip support that are causing frame timing issues with Radeon GPUs?

    [url<]http://www.guru3d.com/articles_pages/ashes_of_singularity_directx_12_benchmark_ii_review,10.html[/url<]

      • Pitabred
      • 4 years ago

      From that link you just posted:

      “Update: hours before the release of this article we got word back from AMD. They have confirmed our findings. Radeon Software [b<]16.1 / 16.2 does not support a DX12 feature called DirectFlip[/b<], which is mandatory and the solve to this specific situation. AMD intends to resolve this issue in a future driver update."

        • chuckula
        • 4 years ago

        So I guess the answer is nyet.

          • morphine
          • 4 years ago

          Yep, nyet = not yet.

          • anotherengineer
          • 4 years ago

          Indeed, and apparently review on hold. Still waiting for Nvidia to fix some stuff also.

          “For Nvidia it seems that ASYNC compute shaders are not working and we have to winder as to why anno February 2016 still is the case ?”

          “AMD intends to resolve this issue in a future driver update, after which we’ll update this article. The same goes for Nvidia with ASYNC compute support, or better yet the lack of it.”

            • DoomGuy64
            • 4 years ago

            I don’t think nvidia can “fix” ASYNC, since their hardware isn’t as flexible as AMD. Support, yes. Fix, no.
            [url<]http://www.overclock.net/t/1572716/directx-12-asynchronous-compute-an-exercise-in-crowd-sourcing[/url<] [quote<] The main problem with nVIDIAs HyperQ implementation is in terms of preemption. It is in part due to the software scheduling aspect of nVIDIAs Maxwell 2 solution as well as the limitations, in terms of Not being able to execute a second compute task until the end of a draw call boundary (lag or latency being introduced) and lack of error verification, from the Asynchronous Warp Schedulers. Example: Say you have a graphic shader, in a frame, that is taking a particularly long time to complete, with Maxwell2, you have to wait for this graphic shader to complete before you can execute another task. If a graphic shader takes 16ms, you have to wait till it completes before executing another graphic or compute command. This is what is called "slow context switching". Every millisecond, brings down your FPS for that frame. Therefore if a graphic shader takes 16ms to execute, a compute task takes 20ms to execute and you've got a copy command taking 5 ms to execute, you end up with a frame which takes 41ms to execute. This introduces a delay, or lag/latency, between executions. While preemption wasn't important for DX11, and nVIDIA primarily designed their Kepler/Maxwell/Maxwell 2 architectures with DX11 in mind, it becomes quite important for DX12 Asynchronous Compute as well as VR. [/quote<]

            • DancinJack
            • 4 years ago

            It’s kind of funny to me. Even though Maxwell 2 (as referred to above) doesn’t possess the resources needed for async compute/shading, regardless if it is software or hardware, I think that just goes to show how well they have done most everything else. Frame times have historically been lower on Nvidia given equal/nearly equal FPS, and IMO, speaks to how quickly they can get work done.

            Anyway, bring on Pascal!

            • DoomGuy64
            • 4 years ago

            Better drivers. Maxwell can do async btw, but only with slow context switching. AMD needs to pick up the pace, which if they did, they could beat nvidia. The thing with AMD drivers is that if you are willing to wait a bit, they can match or exceed nvidia’s performance. AMD keeps up with most titles nowadays, but nvidia still has better initial support with less bugs. This isn’t something that bothers me though, since I very seldom buy games day one. ST:BF is a good example of why you shouldn’t, with the ridiculous pricing and limited content. I’m not about to spend $100+ for the full day1 experience, so by the time I do buy the game, AMD’s drivers will have more than caught up to the game.

            • anotherengineer
            • 4 years ago

            “but nvidia still has better initial support with less bugs”

            Remember vista when it came out?? 😉 I do, I had an Nvidia card at the time, just downgraded back to XP 🙂 Crap happens in the software world……….to everyone………

            • Klimax
            • 4 years ago

            I think Nvidia has learned from that lesson… (BTW: Around that time one of driver series had several nasty bugs too under XP. IIRC 95.xxx or something like that in 9x range)

            • bfar
            • 4 years ago

            Yea, I learned my lesson after Titanfall.

            Although I did buy Alien Isolation on day one. When they get it right, they get it right!

      • DoomGuy64
      • 4 years ago

      DX12 isn’t important ([i<]now[/i<]). What people are really interested in, is if the multimon and general corruption issues are solved. No point in updating if your screen corrupts returning from a low power state, or hdmi doesn't work. (if you use hdmi) What hasn't been fixed though is the clock throttling, which has been going on for a while, and requires a 3rd party tool (clockblocker) to fix. People are pretty ticked that this issue hasn't been solved, but I have a feeling it's more complex than just a simple driver bug, since it isn't something everyone is equally affected by.

      • DoomGuy64
      • 4 years ago

      UPDATE: Guru3d is doing it wrong.
      [url<]http://www.extremetech.com/extreme/223654-instrument-error-amd-fcat-and-ashes-of-the-singularity[/url<] TLDR: AMD is using DWM, and not DirectFlip, which throws off FCAT results.

Pin It on Pinterest

Share This