GMA500 vs. G3 software render

Al Dunsmuir al.dunsmuir at sympatico.ca
Thu May 3 17:07:10 UTC 2012


On Thursday, May 3, 2012, 12:52:13 PM, Adam wrote:
> On Thu, 2012-05-03 at 10:05 -0400, Adam Jackson wrote:
>> So if we want to blacklist low performers, okay, that's a thing we can
>> do I suppose.  Where do we draw the line?

> 'anything Atom is slow as hell' would be an obvious win, I suspect. If
> your baseline of acceptable performance is 'Core and up', I think we can
> be pretty confident that Atom is not going to make it.

I personally believe that blacklisting based on CPU capacity is wrong.

Blacklisting  might  have the side effect of blocking future Atom CPUs
(perhaps  in  a  multi-CPU  configuration)  with  a better performance
story.  It  certainly  eliminates  any pressure to further improve the
code.

Perhaps  a  given  setup  it  may  be  slow,  but  blacklisting  is  a
non-configurable  and  arbitrary way to prevent users from making that
choice on their own.

I  am  personally  using  some  32-bit  Pentium  4 (desktop and laptop
systems. My hardware.  My choice.

Someone  who  has  a  reason  to use a given configuration may be more
motivated to be patient than a developer who is used to the latest and
greatest.





More information about the test mailing list