<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Thanks for the additional in Graphics</title>
    <link>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086582#M87412</link>
    <description>&lt;P&gt;Thanks for the additional information. I have a ticket filed and am getting the driver team to investigate.&amp;nbsp;&lt;/P&gt;</description>
    <pubDate>Tue, 17 May 2016 14:37:37 GMT</pubDate>
    <dc:creator>Michael_C_Intel2</dc:creator>
    <dc:date>2016-05-17T14:37:37Z</dc:date>
    <item>
      <title>Possible Performance Regression</title>
      <link>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086579#M87409</link>
      <description>&lt;P&gt;System Used: ASUSTek Z170-WS&lt;BR /&gt;
	CPU SKU: Intel Core i7-6700K CPU @ 4.00Ghz&lt;BR /&gt;
	GPU SKU: Intel HD Graphics 530&lt;BR /&gt;
	Processor Line:&lt;BR /&gt;
	System BIOS Version: American megatrends Inc. 0207&lt;BR /&gt;
	CMOS settings: default&lt;BR /&gt;
	Graphics Driver Version: 20.19.15.4352 (also tried earlier *.4300 and later *.4380 Beta)&lt;BR /&gt;
	GOP/VBIOS Version: 1024.0&lt;BR /&gt;
	Operating System: Windows 7 64-bit Professional&lt;BR /&gt;
	OS Version: 6.1.7601&lt;BR /&gt;
	API: DirectX9 / DXVA&lt;BR /&gt;
	Occurs on non-Intel GPUs?: No&lt;/P&gt;

&lt;P&gt;Hello,&lt;/P&gt;

&lt;P&gt;We've got a fairly complicated video application which used to run great on Haswell/HD4600, but after upgrading to Skylake/HD530 exhibits frequent frame drops and frame times in excess of 30 ms under high load.&amp;nbsp; The application ultimately displays upwards of 24 independently sized video windows.&amp;nbsp; There are at most 8 video sources, however, each fed to the application via shared memory at exactly 60Hz.&amp;nbsp; To minimize bandwidth, we upload each video to the GPU (using Lock/Unlock()) only once, then use DXVA to convert it.&amp;nbsp; From there, we blit the transformed data to each window using either DXVA or StretchRect().&amp;nbsp; The only difference there is the quality/speed tradeoff.&amp;nbsp; Each video source is handled on a separate thread (including unique D3D and DXVA devices).&amp;nbsp; As mentioned, this used to work great, with solid 60Hz performance even under high load (&amp;gt;24 windows).&amp;nbsp; Excepting the upgrade to Skylake (new motherboard, cpu/gpu) the hardware and software is largely the same.&amp;nbsp; Not to draw comparisons, but for what it's worth, other vendors also work well.&lt;/P&gt;

&lt;P&gt;We're not sure if this is a performance regression in the drivers or just what the hardware can do.&amp;nbsp; Everything we've read, however, suggests that Skylake beats out Haswell in nearly every case.&amp;nbsp; Any ideas would be appreciated.&amp;nbsp; Thanks!&lt;/P&gt;</description>
      <pubDate>Wed, 24 Feb 2016 18:11:18 GMT</pubDate>
      <guid>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086579#M87409</guid>
      <dc:creator>Sean_C_</dc:creator>
      <dc:date>2016-02-24T18:11:18Z</dc:date>
    </item>
    <item>
      <title>Hi Sean,</title>
      <link>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086580#M87410</link>
      <description>&lt;P&gt;Hi Sean,&lt;/P&gt;

&lt;P&gt;Thanks system setup information, can you also fill in the rest of the template (below) so we can reproduce the issue? If you have a sample file we can use to reproduce that would help too.&amp;nbsp;&lt;/P&gt;

&lt;P&gt;Thanks,&lt;/P&gt;

&lt;P&gt;-Michael&lt;/P&gt;

&lt;P style="word-wrap: break-word; font-size: 12px;"&gt;Steps to Reproduce:&lt;BR /&gt;
	-------------------------------&lt;BR /&gt;
	1.&lt;BR /&gt;
	2.&lt;BR /&gt;
	3.&lt;BR /&gt;
	4.&lt;BR /&gt;
	5.&lt;BR /&gt;
	6.&lt;BR /&gt;
	7.&lt;BR /&gt;
	8.&lt;BR /&gt;
	9.&lt;BR /&gt;
	10.&lt;/P&gt;

&lt;P style="word-wrap: break-word; font-size: 12px;"&gt;Expected Results:&lt;BR /&gt;
	-------------------------------&lt;/P&gt;

&lt;P style="word-wrap: break-word; font-size: 12px;"&gt;Actual Results:&lt;BR /&gt;
	-------------------------------&lt;/P&gt;

&lt;P style="word-wrap: break-word; font-size: 12px;"&gt;Additional Information:&lt;BR /&gt;
	-------------------------------&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Wed, 24 Feb 2016 21:14:59 GMT</pubDate>
      <guid>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086580#M87410</guid>
      <dc:creator>Michael_C_Intel2</dc:creator>
      <dc:date>2016-02-24T21:14:59Z</dc:date>
    </item>
    <item>
      <title>We've produced a sample which</title>
      <link>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086581#M87411</link>
      <description>&lt;P&gt;&lt;SPAN style="font-size: 1em;"&gt;We've produced a sample which demonstrates the regression in DXVA2 performance on Skylake. When executed, the sample writes the FPS into the window title bar.&lt;/SPAN&gt;&lt;/P&gt;

&lt;P&gt;&lt;SPAN style="font-size: 1em;"&gt;On Ivy Bridge i7-3700 @ 3.4GHz, this produces 59.8 to 60.0 FPS. Haswell produces similar results.&lt;/SPAN&gt;&lt;/P&gt;

&lt;P&gt;On Skylake i7-6700 @ 3.4GHz, the sample produces an astonishing regression: 32-42 FPS. Obviously we expected numbers at least as good as Ivy Bridge.&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;

&lt;P&gt;Source, which compiles with Visual Studio 2015 Update 2, is available at:&amp;nbsp;&lt;A href="https://www.dropbox.com/s/96m11xejt9xkhan/VideoBLT.zip?dl=0"&gt;https://www.dropbox.com/s/96m11xejt9xkhan/VideoBLT.zip?dl=0&lt;/A&gt;&lt;/P&gt;

&lt;P&gt;Static-linked executable is available at:&amp;nbsp;&lt;A href="https://www.dropbox.com/s/ul9j01mfynfaf0u/VideoBLT.exe?dl=0"&gt;https://www.dropbox.com/s/ul9j01mfynfaf0u/VideoBLT.exe?dl=0&lt;/A&gt;&lt;/P&gt;

&lt;P&gt;-&lt;/P&gt;

&lt;P&gt;Colin Jensen&lt;/P&gt;

&lt;P&gt;Staff Engineer, Abekas Inc.&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Mon, 09 May 2016 19:59:42 GMT</pubDate>
      <guid>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086581#M87411</guid>
      <dc:creator>Colin_J_</dc:creator>
      <dc:date>2016-05-09T19:59:42Z</dc:date>
    </item>
    <item>
      <title>Thanks for the additional</title>
      <link>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086582#M87412</link>
      <description>&lt;P&gt;Thanks for the additional information. I have a ticket filed and am getting the driver team to investigate.&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Tue, 17 May 2016 14:37:37 GMT</pubDate>
      <guid>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086582#M87412</guid>
      <dc:creator>Michael_C_Intel2</dc:creator>
      <dc:date>2016-05-17T14:37:37Z</dc:date>
    </item>
    <item>
      <title>Our investigation shows that</title>
      <link>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086583#M87413</link>
      <description>&lt;P&gt;Our investigation shows that your&amp;nbsp;&lt;SPAN style="font-size: 1em; line-height: 1.5;"&gt;app doesn't appear to be aware that the images are being scaled. We&lt;/SPAN&gt;&lt;SPAN style="font-size: 1em; line-height: 1.5;"&gt;&amp;nbsp;can make HD 520 pass by changing the DPI settings in the OS. &amp;nbsp;We&amp;nbsp;&lt;/SPAN&gt;&lt;SPAN style="font-size: 1em; line-height: 1.5;"&gt;can also make HD 4600 fail by changing the DPI from 100 to 200. Have you added&amp;nbsp;&lt;/SPAN&gt;&lt;SPAN style="font-size: 1em; line-height: 1.5;"&gt;DPI-awareness into your app as descibed here:&amp;nbsp;&lt;/SPAN&gt;&lt;SPAN style="font-size: 1em; line-height: 1.5;"&gt;&lt;A href="https://msdn.microsoft.com/en-us/library/ms701681(v=vs.85).aspx"&gt;https://msdn.microsoft.com/en-us/library/ms701681(v=vs.85).aspx&lt;/A&gt;?&amp;nbsp;&lt;/SPAN&gt;&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Tue, 07 Jun 2016 16:06:25 GMT</pubDate>
      <guid>https://community.intel.com/t5/Graphics/Possible-Performance-Regression/m-p/1086583#M87413</guid>
      <dc:creator>Michael_C_Intel2</dc:creator>
      <dc:date>2016-06-07T16:06:25Z</dc:date>
    </item>
  </channel>
</rss>

