videocardzFirst AMD Radeon R9 290X 1080p performance review

[itsmydamnation]

Tahiti wasn't compute heavy per say. it was the GDRR5 PHY that was massive. Tahiti is effectively "unbalanced", it had more ALU and memory bandwidth then setup and ROP's. If you look at all the other GCN chips Hawaii returns to a linear scaling of setup/alu/mtu/rop/bandwdith.

 

[Will Robinson]

On topic, where's the 290? It'd be nice to see how it is performing relative to the 290x as it will likely be the more popular of the two.

The 290 has been locked away with a division of storm troopers guarding it till launch day.:sneaky:
Its going to be a hummer.
$450?

 

[raghu78]

Tahiti wasn't compute heavy per say. it was the GDRR5 PHY that was massive. Tahiti is effectively "unbalanced", it had more ALU and memory bandwidth then setup and ROP's. If you look at all the other GCN chips Hawaii returns to a linear scaling of setup/alu/mtu/rop/bandwdith.

true. Tahiti's biggest problem was the front end which was same as Pitcairn event though it had 60% more sps. 2 ACE, 2 geometry engines and 2 rasterizers. the memory bandwidth was there but the front end to drive those sp was not. Hawaii is a perfect doubling of all resources except sp count which is 2.2x. given how HD 7870 is 1.9x the perf of HD 7770 perf, R9 290X should easily land at 2x the perf of HD 7870 which it seems to according to leaked benchmarks.

 

[boxleitnerb]

That wasn't tapeout, that was samples returning from TSMC.

As far as the performance/mm^2 thing, considering Tahiti was already almost at Titan levels for FP64, Hawaii is adding SP's and getting 1:2 DP performance. It's going to demolish it there. However compute will never be that comparable, since a lot of it depends on architectural features that are better for certain tasks. Hawaii will dominate the FLOPS contest though, but that is no real world indicator of performance.

Hm, K20 does 1.17 TFLOPs, S9000 does 806 GFLOPs, both at 225W TDP. Not quite the same.

When was Hawaii tapeout incidentally? In any case, Hawaii is much newer than GK110 and possibly on 28HPM as well (any hints? ). The comparison between so different GPUs regarding perf/mm2 is just nonsense imo.

You should notice already that it's the very same excuse that AMD loyalists sow at Kepler launch.

2012 AMD zealot: "But Tahiti's die is bigger because it has compute capabilities!"
2012 Nv zealot: "OMG what's with that horrible gaming performance per square mm?"

2013 Nv zealot: "But GK110's die is bigger because it has dedicated compute capabilities!"
2013 AMD zealot: "OMG what's with that horrible gaming performance per square mm?"

It certainly is a fact and a valid point that some things make GPUs larger but don't affect performance in all workloads. I never had a problem with that and I had hoped that there wouldn't be double standards regarding this - I was mistaken. Perf/mm2 and perf/W alone don't say much - you can trade the one for the other to a certain extent. Add the other factors that I've brought up and it should be clear that the comment made by RS before is just pure nonsense.

 

[Rvenger]

When was Hawaii tapeout incidentally? In any case, Hawaii is much newer than GK110 and possibly on 28HPM as well (any hints? ). The comparison between so different GPUs regarding perf/mm2 is just nonsense imo.

Well then you know more than AMD because from what I heard, its based on the same architecture and same 28nm process. Nothing newer, just what they should have done to start.

 

[Skurge]

Well then you know more than AMD because from what I heard, its based on the same architecture and same 28nm process. Nothing newer, just what they should have done to start.

Yep, Lessons learned from pitcairn. The just doubled that up.