Dan
17-08-2002, 02:11
Fonte: xbitlabs (http://www.xbitlabs.com/news/story.html?id=1029533341)
"(...)First of all, they reported that ATI hasn’t yet given up the idea about developing dual-chip graphics cards. RADEON 9700 chips proved capable of working in pairs on a single PCB, which should theoretically allow creating 3D accelerators with tremendous performance. At the same time an ATI representative said that at present they had no plans concerning the launching of these monstrous graphics cards and the only argument that could push them to consider this possibility could be the market. If the market wants high-performance solutions like that supporting 4-monito configurations, then why not give it a try."
Tambem informaçoes sobre a Radeon9500:
"Speaking about more down-to-earth things we should also touch upon RADEON 9500. It turns out that the light version of ATI RADEON 9700 will be not just a graphics card with lower working frequencies. ATI is going to reduce its performance tangibly by removing half of the rendering pipelines (so that there are 4 of them used) and reducing the memory bus width by 50% (down to 128bit). (...)"
"(...)First of all, they reported that ATI hasn’t yet given up the idea about developing dual-chip graphics cards. RADEON 9700 chips proved capable of working in pairs on a single PCB, which should theoretically allow creating 3D accelerators with tremendous performance. At the same time an ATI representative said that at present they had no plans concerning the launching of these monstrous graphics cards and the only argument that could push them to consider this possibility could be the market. If the market wants high-performance solutions like that supporting 4-monito configurations, then why not give it a try."
Tambem informaçoes sobre a Radeon9500:
"Speaking about more down-to-earth things we should also touch upon RADEON 9500. It turns out that the light version of ATI RADEON 9700 will be not just a graphics card with lower working frequencies. ATI is going to reduce its performance tangibly by removing half of the rendering pipelines (so that there are 4 of them used) and reducing the memory bus width by 50% (down to 128bit). (...)"