I wonder if anyone could fill me in on this, because I am constantly perplexed. If you don't know, all AMD's (and I think Nvidia is in a similar situation, though I don't know for sure) low-end cards are VLIW4, as in, the 40nm graphics architecture that preceded the 29nm Graphics Card Next that we see in the 7xxx series. Why is this? The thing is, these cards find a market in people who need low profile/low power/low noise descrete cards, and GCN is so much more efficient than VLIW4 that AMD could really dominate the market by releasing passively-cooled cards that perform as good as VLIW4 dual-slot solutions. I for one would definitely buy one of these for small projects!
And while I'm at it, why don't they release an AGP card for legacy machines, but using GCN? I am aware AGP's bandwidth wouldn't handle a tenth of the raw gigaflops of a HD 7850, but they could easily saturate the AGP bandwidth with a low profile, passively cooled card that would also be the fastest card released for that slot. I refuse to believe there are not customers for this!
And while I'm at it, why don't they release an AGP card for legacy machines, but using GCN? I am aware AGP's bandwidth wouldn't handle a tenth of the raw gigaflops of a HD 7850, but they could easily saturate the AGP bandwidth with a low profile, passively cooled card that would also be the fastest card released for that slot. I refuse to believe there are not customers for this!