Comments Locked

24 Comments

Back to Article

  • hpvd - Monday, March 16, 2015 - link

    just a essential question:
    Where from could a software "know" which different types of cores with which types of features are available in a heterogeneous system?
    How could the software or the OS choose the best fitting types of cores for a sepcial part of the software's code?

    A method which is used for CPUs with big.LITTLE concepts (arm cpus in smartphones) seams not to be suitable/efficient for complex taks
    her the scheduler of the OS "simply" sorts for priorität (high -> big) and foreground/background tasks (foreground >big)
    https://en.wikipedia.org/wiki/ARM_big.LITTLE#Heter...
  • pTmdfx - Tuesday, March 17, 2015 - link

    HSA exposes the platform topology, and software does its own choice (i.e. pick an agent, create a queue associated to the agent, submit a packet to the queue). In other words, if you want automation, you would have to wait for a library (e.g. AMD's Bolt) to target HSA.
  • cyberguyz - Tuesday, March 17, 2015 - link

    Novice tech writer mistake #1:

    Commonly using an acronym (i.e. HSA) without definint it within tghe first paragraph of the article. Noob writers alwasy assume everybody knows what an uncommon actronyn means without botrhering to define it until the article is almost done (if at all).

    Are there no editors here to catch these noob mistakes?
  • cyberguyz - Tuesday, March 17, 2015 - link

    PS: While I can't type worth a crap, even I know this much.
  • hlmcompany - Tuesday, March 17, 2015 - link

    Yup, this happens. Of the different things the acronym may mean, HSA is Head Stack Assembly for Hard Disk Drives.
  • name99 - Tuesday, March 17, 2015 - link

    If you don't know what HSA means in this context, then perhaps you should be reading PCWorld or something rather than AnandTECH?
    Or perhaps you should learn how to use Google and Wikipedia?

    It's a bit pathetic to have, in the same comment thread, one set of guys complaining that freshmen programmers don't get taught multi-threaded programming aggressively enough, while some other guy is complaining that the article uses words he doesn't understand, and that he, apparently, doesn't know how to look up on the internet.

    Filling an article with definitional crap is NOT free --- it wastes the time of readers who already know what is going on, who have already internalized HSA (and other TLAs) as a single concept and who actually stumble in their reading over unexpected expansions of the phrase.
  • cyberguyz - Tuesday, March 17, 2015 - link

    OK Einstein. What makes you think that comp sci grads are the only ones that read articles here? You get all kinds of people reading this site including computer hobbyists as well as folks trying to learn a little bit about the computer tech around them.

    Anandtech is by no means the sole domain of the Computer Science Ms degree holder that just graduated last week and never was. So get your head our of your ass and get with the program.

    FYI I have been in the computer software development business for over 35 years. That's right - no doubt longer than you have been alive. One thing I have been beaten over the head over all those years by MY tech writers is that you never, ever use acronyms without first defining them. As this article is using an acronym within its title, it is excusable to hold off the definition until the first paragraph. You don't believe me? Try being a tech writer for software developer in one of the largest acronym manufacturers in the world. You might actually learn something.
  • iLovefloss - Tuesday, March 17, 2015 - link

    @name99 Nah, you're just being an elitist jerk. By defining it, cyberguyz doing something like this "Heterogeneous System Architecture (HSA)". This tells the reader that HSA means Heterogeneous System Architecture instead of the numerous other things it could mean.

    The HSA Foundation's frontpage (and all of their literature) does this even though they assume everyone reading their page has a grasp on what HSA is and what it is aiming for.

    http://www.hsafoundation.com/

    Pointing out that HSA means Heterogeneous System Architecture will not hurt anyone, and if that trips you up, you have some reading comprehension problems you need to work on that you don't need to drag anyone else in on.
  • cyberguyz - Saturday, March 21, 2015 - link

    That is exactly what I am talking about iLovefloss. It doesn't take much but gives folks something more than an acronym to google with if they want more info. it's just good tech writing practices.
  • R3MF - Tuesday, March 17, 2015 - link

    Has there been any indication from AMD on how their High-End dedicated GPU products fit into their HSA strategy?

    Right now, there is a problem:
    1. HSA is deemed core to AMD's future in obviating the IPC deficit that derives from Intel's enormous R&D and scale of production advantages
    2. The fact that AMD's £500 GPU products are completely unable to make use of HSA, providing no platform advantage for AMD to capiutalise upon

    How is AMD going to square this circle?
  • Oniguma - Tuesday, March 17, 2015 - link

    You're fabricating a problem that doesn't exist.

    The squared circle aka squircle is that they use their GPU development in power consumption, performance and HBM VRAM that are directly integrated from discrete GPU development, over to APU development.

    The two products serve two different needs. While having similar development goals.

    What we really need to see is the IPC on AMD's X86 become competitive against Intel in their APU's. Otherwise all the integrated GPU power in the world won't mean anything in the desktop space.
  • R3MF - Tuesday, March 17, 2015 - link

    seems pretty inefficient, in that the enormous R&D resource they direct towards high-end GPU's does not complement the R&D resource they put into HSA.
  • pTmdfx - Tuesday, March 17, 2015 - link

    If you have gone through the specification, you might actually find the HSA specification *supports* accelerators with its local memory - the memory model, memory management API and the platform topology specification. One of the examples given in the specification even has discrete GPU nodes illustrated.

    Theoretically, discrete GPUs since Sea Islands should be able to support HSA to some degree (Base Profile, perhaps). It should be able to access system coherent memory via PCIe, and can handle system address translation via ATS/PRI backed by the IOMMUv2. The only thing left is probably whether or not AMD will enable the software for it.
  • pTmdfx - Tuesday, March 17, 2015 - link

    Okay, a correction: It is *likely* able to access system coherent memory via PCIe.
  • R3MF - Tuesday, March 17, 2015 - link

    cheers.

    via IOMMU would be interesting, but where doews this leave people with £500 AMD GPU's and an Intel Platform...

    if AMD made a platform worthy of a £500 GPU (or even a £300 GPU), then this wouldn't be a problem, but they don't.
  • pTmdfx - Tuesday, March 17, 2015 - link

    Perhaps someday. It is not all about AMD's own effort, but Intel's platform has to have the platform features for PCIe peripherals that are required by the whole HSA specification. Well... I know some of the boxes are checked, but I'm not sure if it does work. Still at least discrete GPU supports OpenCL 2.0.

    Frankly, HSA is for APUs in the first place, as HSA is to leverage the advantage of integration. Not a huge problem anyway, as you can probably expect HSA is more of an optimised path, where OpenCL 2 or similar is the baseline target. The feature parity in graphics is also the unaffected.
  • dahippo - Wednesday, March 18, 2015 - link

    My understanding is that IoT will be a big winner using HSA. You can put any accelerator on and use same memory and cache to keep size and power low.
  • Selim Reza - Monday, October 22, 2018 - link

    HSA Foundation engineering works with upstream projects on a set of requirements that are determined by the Technical Steering Committee.

    CatLight is a notification app for developers. It shows the current status of continuous delivery, tasks, and bugs in the project and informs when attention is needed.
    Checkout the page: https://catlight.io/home/sitemap

Log in

Don't have an account? Sign up now