{ "numMessagesInTopic": 2, "nextInTime": 497, "senderId": "i9iY2N0tpsqF6I9iFF0y_XDynwQWZPMKmLsxRPJC8jsOGPWrYpIZ_HnQsrABErgEwUVmp1cnCnYZDa6fIioh-cBqkVsFaz4EEsdK6CJjGg", "systemMessage": false, "subject": "Re: Sluggishness of PCI Quadro FX600", "from": ""fluppeteer" <yahoo@...>", "authorName": "fluppeteer", "msgSnippet": "... solution, ... transferred ... *Nod*. The greying out is, as far as I know, done with the CPU - meaning the display needs locking, and transferred over the", "msgId": 496, "profile": "fluppeteer", "topicId": 495, "spamInfo": { "reason": "0", "isSpam": false }, "replyTo": "LIST", "userId": 192443393, "messageBody": "
> responsiveness. The easiest way to see this bottleneck is when yousolution,
\n> hit the "Turnoff Computer" button in WinXP. As you all know, the
\n> screen slowly desaturates to monochrome, and on a PCI T221
\n
> this takes FOREVER -- each desaturation pass takes about 0.5 - 1transferred
\n> second, whereas it's much faster on a AGP solution.
\n> I guess that this is one case where the data needs to be
\n
> en masse from the CPU to the GPU.*Nod*. The greying out is, as far as I know, done with the CPU -
\n
\n
> The point is, moving to a ultra-expensive exotic solution like thepeculiarity
\n> Parhelia HR256 @ 41Hz refresh will probably not fix these problems
\n> if it is directly due to the PCI bottleneck and not some
\n
> of drivers or the particular card.Ironically, 3D stuff will probably be better (at least under
\n
\n