Announcement

Collapse
No announcement yet.

Best AMD OpenCL/OpenGL Budget Development Box

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

  • Best AMD OpenCL/OpenGL Budget Development Box

    I'm looking for an AMD system (and an Intel and an Nvidia, but those are posted elsewhere) that will provide the best range of current and near-future OpenGL/OpenCL/etc. capabilities for experimental development and testing. I am **not** looking for wicked fast gaming performance, compute performance, or anything like that. I want maximum capabilities for the least money. To put it in specifics, I want the following:

    - AMD CPU/APU
    - Motherboard for above (Assume $90.00)
    - 8GB RAM (Assume $50.00)
    - AMD GPU (Optional with APU)
    - Total Price >= $300.00 @ newegg
    - OS: (Arch) Linux with either Radeon or Catalyst

    The reason I ask is that most discussion of APUs/GPUs seems to be geared towards games or compute use. I'm just looking for a development and test box, not something for daily use.

  • #2
    Originally posted by justinzane View Post
    I'm looking for an AMD system (and an Intel and an Nvidia, but those are posted elsewhere) that will provide the best range of current and near-future OpenGL/OpenCL/etc. capabilities for experimental development and testing. I am **not** looking for wicked fast gaming performance, compute performance, or anything like that. I want maximum capabilities for the least money. To put it in specifics, I want the following:

    - AMD CPU/APU
    - Motherboard for above (Assume $90.00)
    - 8GB RAM (Assume $50.00)
    - AMD GPU (Optional with APU)
    - Total Price >= $300.00 @ newegg
    - OS: (Arch) Linux with either Radeon or Catalyst

    The reason I ask is that most discussion of APUs/GPUs seems to be geared towards games or compute use. I'm just looking for a development and test box, not something for daily use.
    The price range is really restrictive, but I do know of a few good options...

    Option 1: AMD APU ( For this, Wikipedia has an excellent article over these chips. ) - I can only suggest keeping an eye out for the GCN based chips that will be released early next year. However if you don't want to wait, the following specification should work great.
    Option 2: AMD CPU w/ Radeon Graphics.

    On both systems I did choose Gigabyte because the experience I have with this maker generally has great support on linux. I don't know about the EFI/UEFI support on these boards, but it should work out well if you place the system into legacy mode.

    Comment


    • #3
      Originally posted by Dandel View Post
      I can only suggest keeping an eye out for the GCN based chips that will be released early next year
      I've done a bit of searching to figure out what, besides marketing hype, GCN is. And, as far as I can tell, it is an internal change from VLIW MIMD to RISC SIMD. While this may be nice for performance or efficiency, it does not seem to be at all relevant to what APIs the APU/GPU can support. Can anyone explain why, from the perspective of available APIs, GCN matters?

      The same goes for TrueAudio... From a developer's perspective, what does it offer?

      Originally posted by Dandel View Post
      I did choose Gigabyte because the experience I have with this maker generally has great support on linux.
      I've only ever had trouble with laptops; but, I'll keep your recommendation in mind. I don't care about overclocking, goofy "quick boot" stuff or just about any other feature, my tendency is to assume that any board with supported IO chips is fine. Thanks.

      Comment


      • #4
        Originally posted by justinzane View Post
        I've done a bit of searching to figure out what, besides marketing hype, GCN is. And, as far as I can tell, it is an internal change from VLIW MIMD to RISC SIMD. While this may be nice for performance or efficiency, it does not seem to be at all relevant to what APIs the APU/GPU can support. Can anyone explain why, from the perspective of available APIs, GCN matters?
        GCN matters since it is what appears in both major consoles this time around. The XBOX One and PS4 This leads to more benefits elsewhere. As far as API Based stuff, the main things that are listed is the new Mantle API that amd is releasing for the GCN chips. This means that you haft to have any of the following...
        • Radeon HD 77xx Series or newer
        • Radeon HD 8550M ( or newer )
        • Radeon Rx 200
        • AMD APU based on Jaguar Cores


        Originally posted by justinzane View Post
        The same goes for TrueAudio... From a developer's perspective, what does it offer?
        There is not a lot of information on TrueAudio yet, but I do know that it'll probably have a lot of help with improving overall Audio quality. As far as cards that support TrueAudio, the only cards I currently know that support this API are as follows... Radeon R7 260X, Radeon R9 290 and Radeon R9 290X.

        Originally posted by justinzane View Post
        I've only ever had trouble with laptops; but, I'll keep your recommendation in mind. I don't care about overclocking, goofy "quick boot" stuff or just about any other feature, my tendency is to assume that any board with supported IO chips is fine. Thanks.
        I have had similar experience on laptops and the latest AMD APU based laptops work a lot better than previous generations. ( The only Tip I have is don't expect open source graphics to be fully functional for a few years seeing as how there is bugs on this ).

        Comment


        • #5
          Mantle, TrueAudio, etc.

          Well, I seem to have mentally dismissed Mantle as I have CUDA. I know that for some problem sets, CUDA is a great tool. But, like CUDA, mantle seems to be a single-vendor lock-in attempt with limited general appeal. If Mantle support starts appearing on Nvidia GPUs or Intel CPU or a significant variety of ARM-ISA implementations, it may be of consideration **to me**. Since AMD seems to have limited ability to effectively implement software support for anything lately, I'm not willing to spend even a few dollars extra on the off chance that Mantle becomes prevalent and useful for non-Windows developers.

          As far as TrueAudio goes, I spent last winter building an almost-audiophile grade subwoofer (~12 foot horn) and FFTFF center-channel. I was sufficiently motivated to write a plugin for pulseaudio to handle the crossover digitally, but I cannot fathom what AMD can do to improve the "quality" of the bitstream on S/PDIF? Either the bits arrive at the amplifier's decoder as intended or they do not. 1990's technology can handle this. Sh|t, even well designed analog equipment can do extremely well. Just like friggin $100/ft gold plated Cat6 will not make gigabit ethernet faster, I'm pretty sure that nothing AMD has to offer will improve audio in any noticeable way. And, if there happens to be some API attached to "leverage" TrueAudio, what other hardware is it useful for?

          Besides, I've already got a system for actual daily use and one as a media center. I'm just looking for a test box to do compile/debug/profile cycles on.

          I appreciate the info, though!

          Comment


          • #6
            I would get a good FM2+ motherboard along the lines that Dandel suggested, fast RAM, and either wait for Kaveri or put a basic Trinity or Richland APU in for now then drop in a Kaveri when they become available. I wouldn't bother with a separate GPU unless you want to do graphics driver development.

            From a development perspective, GCN is better suited for complex arbitrary compute kernels/shaders and has more versatile memory addressing. Kaveri should be a good platform for OpenCL 2.0 features.

            Comment


            • #7
              Originally posted by justinzane View Post
              Well, I seem to have mentally dismissed Mantle as I have CUDA. I know that for some problem sets, CUDA is a great tool. But, like CUDA, mantle seems to be a single-vendor lock-in attempt with limited general appeal. If Mantle support starts appearing on Nvidia GPUs or Intel CPU or a significant variety of ARM-ISA implementations, it may be of consideration **to me**. Since AMD seems to have limited ability to effectively implement software support for anything lately, I'm not willing to spend even a few dollars extra on the off chance that Mantle becomes prevalent and useful for non-Windows developers.
              I agree with you on this point, and I honestly do not believe AMD would limit mantle to windows developers only. I just don't know a lot about this.
              Originally posted by justinzane View Post
              As far as TrueAudio goes, I spent last winter building an almost-audiophile grade subwoofer (~12 foot horn) and FFTFF center-channel. I was sufficiently motivated to write a plugin for pulseaudio to handle the crossover digitally, but I cannot fathom what AMD can do to improve the "quality" of the bitstream on S/PDIF? Either the bits arrive at the amplifier's decoder as intended or they do not. 1990's technology can handle this. Sh|t, even well designed analog equipment can do extremely well. Just like friggin $100/ft gold plated Cat6 will not make gigabit ethernet faster, I'm pretty sure that nothing AMD has to offer will improve audio in any noticeable way. And, if there happens to be some API attached to "leverage" TrueAudio, what other hardware is it useful for?

              Besides, I've already got a system for actual daily use and one as a media center. I'm just looking for a test box to do compile/debug/profile cycles on.

              I appreciate the info, though!
              I understand, and I will say to look at exactly What I pointed out regardless. This is to guarantee you have a lot of leeway. Since the system is for testing you can get a cheaper cpu and motherboard on both cases, and this should drop the cost to a lot closer to the range you where asking for originally.

              Also, This should provide a good test platform since the APU uses a similar architecture to the Radeon R6900 line, and the R7 260X is the latest GCN chip with as many features as possible.

              Also I believe it is worth noting that AMD has a lot better graphics support overall at least on open source drivers that matches ( or exceeds) intel since AMD has regular gpu documentation releases.

              Comment


              • #8
                Nice analysis!

                Thanks. I appreciate your insight. Now, someone please fix the Hynix fire related DRAM price issue. I hate buying when I know prices are temporarily inflated.

                Comment


                • #9
                  Honestly, if you want an APU, wait for next month because early in January the first HSA APU goes on sale, Kaveri. That's when OpenCL will really kick off with OpenCL 2.0. The single greatest issue with OpenCL and APUs today is that there is no HSA hardware yet.

                  Comment


                  • #10
                    HSA sans Intel and Nvidia

                    Originally posted by mmstick View Post
                    Honestly, if you want an APU, wait for next month because early in January the first HSA APU goes on sale, Kaveri. That's when OpenCL will really kick off with OpenCL 2.0. The single greatest issue with OpenCL and APUs today is that there is no HSA hardware yet.
                    [link]http://hsafoundation.com/[/link] The one thing that I note is that Neither Intel or Nvidia are members of HSA; and, since they are huge players in the financial, scientific and entertainment compute markets, I question what HSA will amount to. And, it seems to be significantly led my Microsoft and their track record is not encouraging.

                    Another thing that I am not clear about is what HSA provides -- in real terms -- to open source developers. From what I can tell, it is simply a shim layer that distributes ready-compiled compute tasks to various hardware devices. Which seems like a solution in search of a problem. Or a replacement for an integrated MMU.

                    Add to that that newly released chip families are always marked up to take advantage of price-insensitive early adopters, I wonder what the specific advantage is.

                    Comment


                    • #11
                      Originally posted by justinzane View Post
                      [link]http://hsafoundation.com/[/link] The one thing that I note is that Neither Intel or Nvidia are members of HSA; and, since they are huge players in the financial, scientific and entertainment compute markets, I question what HSA will amount to. And, it seems to be significantly led my Microsoft and their track record is not encouraging.

                      Another thing that I am not clear about is what HSA provides -- in real terms -- to open source developers. From what I can tell, it is simply a shim layer that distributes ready-compiled compute tasks to various hardware devices. Which seems like a solution in search of a problem. Or a replacement for an integrated MMU.

                      Add to that that newly released chip families are always marked up to take advantage of price-insensitive early adopters, I wonder what the specific advantage is.
                      It solves a very real problem -- memory bottlenecks. Especially, it serves the biggest problem holding GPGPU back -- memory bottlenecks. It therefore also makes it possible for GPUs to be 'smarter' since it shares the same memory pool as the CPU, which means it will be able to run a wider variety of different instructions that OpenCL currently cannot do or does very inefficiently. What this means to open source developers is that it will be possible to use GPGPU in a wider variety of programs to speed up processing faster than it would have been with CPU alone. What this means for PC gaming is it will be possible to get highly effecient AI, physics, raytracing, etc. in games. Looking at it right now, because of the memory bottleneck, the only things that run exceedingly well in GPGPU is simple math like crunching prime numbers. This will have a profound effect on the medical, biological, scientific, business, entertainment, gaming, and more industries. What Microsoft does is their business; they are probably looking for the next big thing to keep people from switching to Linux.
                      Last edited by mmstick; 12-11-2013, 07:18 PM.

                      Comment


                      • #12
                        HSA Defines a Hardware Spec?

                        Originally posted by mmstick View Post
                        It solves a very real problem -- memory bottlenecks. Especially, it serves the biggest problem holding GPGPU back -- memory bottlenecks. It therefore also makes it possible for GPUs to be 'smarter' since it shares the same memory pool as the CPU, which means it will be able to run a wider variety of different instructions that OpenCL currently cannot do or does very inefficiently. What this means to open source developers is that it will be possible to use GPGPU in a wider variety of programs to speed up processing faster than it would have been with CPU alone. What this means for PC gaming is it will be possible to get highly effecient AI, physics, raytracing, etc. in games. Looking at it right now, because of the memory bottleneck, the only things that run exceedingly well in GPGPU is simple math like crunching prime numbers. This will have a profound effect on the medical, biological, scientific, business, entertainment, gaming, and more industries. What Microsoft does is their business; they are probably looking for the next big thing to keep people from switching to Linux.
                        Perhaps you can tell that I have developed a serious allergy to the output of PR and Marketing folks. So, I did not get very far wading through the HSA docs which leaves me excessively ignorant...

                        But, your statement seems to imply that HSA defines and mandates a certain level of multi-compute-device memory-management hardware functionality. While I agree that this would be excellent for the performance of compliant multi-device implementations, it seems to be strongly restrictive in terms of vendor participation. Without the participation of Intel CPUs, GPUs and MICs and Nvidia GPUs, and assuming the continued incredibly poor open-source support for ARM-oriented chipsets and GPUs, what is the value?

                        Please know I'm not criticizing you, just trying to be skeptical.

                        Comment


                        • #13
                          Originally posted by justinzane View Post
                          Perhaps you can tell that I have developed a serious allergy to the output of PR and Marketing folks. So, I did not get very far wading through the HSA docs which leaves me excessively ignorant...

                          But, your statement seems to imply that HSA defines and mandates a certain level of multi-compute-device memory-management hardware functionality. While I agree that this would be excellent for the performance of compliant multi-device implementations, it seems to be strongly restrictive in terms of vendor participation. Without the participation of Intel CPUs, GPUs and MICs and Nvidia GPUs, and assuming the continued incredibly poor open-source support for ARM-oriented chipsets and GPUs, what is the value?

                          Please know I'm not criticizing you, just trying to be skeptical.
                          I knew about HSA before it was a PR thing to begin with. AMD has been headed in that direction ever since they bought ATI. This was their end-goal, essentially. There's nothing stopping Intel or NVIDIA from joining the HSA foundation or competing against it, except for the fact that neither NVIDIA or Intel have proper HSA-capable hardware in development that I know of.

                          The problem with NVIDIA and Intel is that they aren't so great at designing chips they have no experience in designing. Neither has a strong suit in both graphics and processor designs like AMD. And especially for NVIDIA, HSA implies a design that uses OpenCL 2.0+ when NVIDIA wishes to keep using their proprietary CUDA.

                          As for ARM, I can't say much about that since ARM has a variety of different designs from a variety of different companies and I know even less about the mobile GPU designs. Someone would just have to design a chip that has the GPU/CPU integrated as one that uses the same shared memory.

                          In other words, only AMD cares about development of an HSA solution right now, but maybe Intel will change their minds once they see AMD chips knocking them out of benchmarks regularly for a cheaper price and running some cool applications that resulted from the new design. From what I've seen, AMD has strong connections with Microsoft right now, which is kind of confusing when Windows/Intel was originally the big thing, so it's not surprising that Microsoft would be heavily backing it. I'm sure they want some Windows Phones using the mobile HSA APU chips next year.

                          Comment


                          • #14
                            Originally posted by justinzane View Post
                            Perhaps you can tell that I have developed a serious allergy to the output of PR and Marketing folks. So, I did not get very far wading through the HSA docs which leaves me excessively ignorant...

                            But, your statement seems to imply that HSA defines and mandates a certain level of multi-compute-device memory-management hardware functionality. While I agree that this would be excellent for the performance of compliant multi-device implementations, it seems to be strongly restrictive in terms of vendor participation. Without the participation of Intel CPUs, GPUs and MICs and Nvidia GPUs, and assuming the continued incredibly poor open-source support for ARM-oriented chipsets and GPUs, what is the value?

                            Please know I'm not criticizing you, just trying to be skeptical.
                            I can see why you'd be overly cautious about PR and Marketing folks. I personally would look into either the AMD A10-7850K or the AMD A10-7700K APU that have an expected release date as Jan 2014. This is because of the highly versatile memory architecture that has been mentioned multiple times. The Kaveri series of chips has three main features that are worth while...
                            • Integrated GCN based GPU ( R7 series w/ Trueaudio and mantle support )
                            • Integrated Cortex A5 coprocessor ( used for the arm security features )
                            • Steamroller based processor. ( This has a lot of performance improvements over the trinity and Richland based processors )
                            • PCIe 3.0 support ( this is a feature for FM2+ boards )

                            Comment


                            • #15
                              I'd recommend this:
                              AMD A8-6500 (3,5GHz, 65W)
                              ASUS A88XM-A (Micro-ATX, FM2+, with Kaveri support)
                              2 x 4GB DDR3-1866 or faster (I use Crucial Ballistix DDR3-1600 but with low timings, 8-8-8-24)
                              Last edited by leonmaxx; 12-13-2013, 04:45 PM.

                              Comment

                              Working...
                              X