|[News] Transcript of Alan Redhouses speech at AmiGBG 2003||ANN.lu|
|Posted on 31-Mar-2003 19:40 GMT by Peter Gordon||54 comments|
I have transcribed Alans AmiGBG speech. It can be viewed here.
|Transcript of Alan Redhouses speech at AmiGBG 2003 : Comment 51 of 54||ANN.lu|
|Posted by IanG on 03-Apr-2003 14:14 GMT|
|In reply to Comment 50 (Anonymous):|
> What about a petition for Alan to get us rid of the expensive useless L3
> cache then?
As I interpreted what was said in the article, Eyetech will not be adding an L3 cache if all you get is 3%.
I suppose it is too optimistic to hope the A1 price comes down if the L3 is left out? ;-)
|Transcript of Alan Redhouses speech at AmiGBG 2003 : Comment 52 of 54||ANN.lu|
|Posted by Sigbjørn Skjæret on 03-Apr-2003 17:38 GMT|
|In reply to Comment 49 (Ole-Egil):|
Well, you know, these things are only as useless as the code using them, like AltiVec you won't see any real performance gains until you code (or have a compiler that knows about it and optimizes) with these things in mind...
Once you have code/data that fits snuggly into pipelines and caches you can start to see some pretty dramatic improvements in speed, after all, this is why the engineers added these extras .. not to crank up prices of their hardware...
So instead of blowing off L3 cache and saying it's useless because nothing fits in it, and that it only slows down memory access, why don't you look at ways to improve your code with caches in mind? ;)
|Transcript of Alan Redhouses speech at AmiGBG 2003 : Comment 53 of 54||ANN.lu|
|Posted by Darth_X on 03-Apr-2003 18:48 GMT|
|Could someone *please* do a transcript with all of the british cursing & swearing that was [muffled]? ;)|
|Transcript of Alan Redhouses speech at AmiGBG 2003 : Comment 54 of 54||ANN.lu|
|Posted by Jim Forbes-Ritte (AGAfaster) on 04-Apr-2003 05:20 GMT|
|In reply to Comment 53 (Darth_X):|
hehe - or put in the Curses in belgian ... 'Verdummer.Verdummer Verdummer Verdummer Verdummer.' ;-)
|Anonymous, there are 54 items in your selection (but only 4 shown due to limitation) [1 - 50] [51 - 54] ||