• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Who thinks AMD initial 28nm GPU will be mid-tier VLIW4?

Vesku

Diamond Member
I'm starting to think the ones out this year will be based on VLIW4 perhaps on 28nm LP even. It would be similar to how the 6800s are reworks of VLIW5 and 6900s are VLIW4 (GCN in the case of 28nm).
 
I'm starting to think the ones out this year will be based on VLIW4 perhaps on 28nm LP even. It would be similar to how the 6800s are reworks of VLIW5 and 6900s are VLIW4 (GCN in the case of 28nm).

yeah I that is what AMD are going to do as well
 
I don't think AMD is going to fix anything for a few years, there 28nm products will just be a 6870 with more and/or higher clocked ROPs, shaders, and texture units.
 
They have 4 ways they can go imho, ranked in order of likelihood:

1. Replace 5770/6770. This is the oldest midrange card they have, and replacing it now could/should give them valuabel experience with 28nm. Also, It seems likely that a 7770 would be similar in performance to 6870, giving good value/thermals/etc to the midrange.

2. Replace barts. This would likely meet/beat 69x0 and give AMD more sales momentum as BF/holiday season sales approach, plus BD and possibly even SB-E buyers would look more closely at something like this since it would be the "latest and greatest" available for them as well.

3. Replace something on the lower end, maybe a 7670 or 7570 type of card. Nothing too exciting here, but they still get that very beneficial experience with 28nm to help with potential kinks in the higher end SI offerings.

4. Replace cayman with a new high end offering. Obviously this is going to happen at some point, but I don't think it very likely that they're going to experiment on a new process with their highest-performing parts. Both camps have tried this and failed miserably over the years.
 
Back
Top