Memory access pattern: Difference between revisions

Revision as of 15:56, 13 June 2016

Revision as of 15:57, 13 June 2016

Line 40:

==== sequential reads with scatter ====

The scatter approach may place less load on a cache hierarchy since a [[processing element]] may dispatch writes in a 'fire and forget' manner (bypassing a cache altogether), whilst using predicatble prefetching (or even using DMA) for it's source data. However, it may be harder to parallelise since there is no guarantee the writes do not interact.

In the past, [[forward texture mapping]] attempted to handle the randomness with 'writes', whilst sequentially reading source texture information.

The [[playstation 2]] console used conventional inverse texture mapping, but handled any randomnessscatter/gather processing 'on-chip' using EDRAM, whilst 3D model (and a lot of texture data) from main memory was fed sequenatially by DMA. This is why it lacked of support for indexed primitives, and sometimes needed to manage textures 'up front' in the [[display list]].

==== gather with sequential writes ====