7 Replies Latest reply on Feb 2, 2012 11:50 AM by Meteorhead

    Double  Precision Support for AMD A8 APU?

    3dmashup
      Does the AMD A8 3850 APU Support Double Precision Floating Point?

      Does the AMD A8 3850 APU support Double Precision Floating Point arithmetic for OpenCL?

      Can any one provide a link to an "official' AMD document  with its OpenCL

      Device Profile details?  This APU comparison table  does not give enough information

      http://www.amd.com/us/products/desktop/processors/a-series/Pages/a-series-model-number-comparison.aspx

       

      The Radeon 6550 spec page  isn't detailed enough

       

      http://www.amd.com/us/products/desktop/graphics/amd-radeon-hd-6000a/hd-6550a/Pages/amd-radeon-hd-6550a.aspx#2

      IMHO, it would be a  Big Miss if the first A Series APU deson't support DP!

       

       

       

       

       

        • Double  Precision Support for AMD A8 APU?

          None of the graphics parts of the APU support double precision at this time.  The APU part is for the consumer market which typcially doesn't a have high demand for double precision workloads.  May I inquire what type of workload you have that requires double precision on APU?

          The OpenCL programming guide has some information on double precision.  The graphics cards that support DP are the Radeon XT 6970  and HD 5870.  In that doc it only lists the 5870 (I've alerted folks that it needs to be updated to include the 6970).

          Edit:  The 5850 and 5830 also support double precision

          1 of 1 people found this helpful