After saying the drivers would be the same he now backtracks saying, sure you can optimize it, but you don't need a special driver. Nobody said you need a special driver.
How is it not relevant?
290x = 113.63 MB/sec per SP
390x = 156.625 MB/sec per SP
390x has HBM @ 1.25Ghz
After saying the drivers would be the same he now backtracks saying, sure you can optimize it, but you don't need a special driver. Nobody said you need a special driver.
Optimizations is a completely different thing. And that you have to show it with 2 completely GPUs as the NV1x and NV20 series was shows it(Programmable shaders etc). Nobody says that a driver cant optimize the GPUs performance. But saying you need a special driver just because someone placed HBM modules on a card instead of GDDR is rubbish.
People should at least claim it is due to a new uarch to fuel the speculation.
And bandwidth wise there may not even be any relevant difference in terms of bandwidth/SP ratio.
290X=113MB/sec per SP.
390X=125MB/sec per SP.
I dont see what HBM and drivers have in common. For the drivers its just plain VRAM.
390X ready for a long time and just waiting for drivers is a big pile of BS.
It doesnt matter. Its all handled by hardware.
No, you said that the delay was rubbish and that the GDDR5 and HBM was so similar that AMD didnt have to work on the drivers regarding memory management.
The hardware itself deals with the changes according to you.
Which is a very simplistic and wrong way to look at it like mentioned by people here who work with software which have experience with drivers, unlike you I bet.
Even I understand that there is software behind every hardware to make sure the hardware works properly. Which is extremely crucial to test 1000x before pushing out a graphic card with a new type of memory to ensure stability. Especially for AMD which is in a difficult position market share wise and can`t afford to mess up like Nvidia did with GTX 970 for example.
HBM is a brand new technology. I bet you that making drivers for those and making sure its 100% stable, aint an easy task. Probably new ways to move data in and out the memory stacks compared to traditional GDDR.
What difference is there from the drivers view in relation to GDDR and HBM? It was you claiming that HBM was the reason for driver delay because it was a new thing.
Let me remind you of your own statement.
Optimizations is a completely different thing. And that you have to show it with 2 completely GPUs as the NV1x and NV20 series was shows it(Programmable shaders etc). Nobody says that a driver cant optimize the GPUs performance. But saying you need a special driver just because someone placed HBM modules on a card instead of GDDR is rubbish.
People should at least claim it is due to a new uarch to fuel the speculation.
And bandwidth wise there may not even be any relevant difference in terms of bandwidth/SP ratio.
290X=113MB/sec per SP.
390X=125MB/sec per SP.
Yes but it's likely full Tonga or at least full Tonga with only the memory controller not fully used due to power efficiency concerns.
What difference is there from the drivers view in relation to GDDR and HBM? It was you claiming that HBM was the reason for driver delay because it was a new thing.
Let me remind you of your own statement.
if it was ready, they wouldn't be waiting months to launch.....I call bs on this
Seems like AMD's PR machine is in panic mode and trying to present whatever they can to get people to delay purchasing 980s or Titan X. If it was ready, they'd push it and recoup lost market share. Besides, this thing is built on 28 nm, I think people are expecting way too much out of AMD, HBM or not. The hype may end up just making it an underwhelming launch if it barely reaches Titan X performance or even beats it by 5%. They'd need at least a clear 15%+ advantage across the board over Titan X to really live up to the hype and being 3+ months late. Those that go Crossfire may be in for a world of pain given AMD's track record with driver and crossfire support.
Edit: All of you should also keep in mind that AMD isn't primarily in the business of making graphics, so you should temper your expectations.
Seems like AMD's PR machine is in panic mode
Seems like AMD's PR machine is in panic mode and trying to present whatever they can to get people to delay purchasing 980s or Titan X. If it was ready, they'd push it and recoup lost market share. Besides, this thing is built on 28 nm, I think people are expecting way too much out of AMD, HBM or not. The hype may end up just making it an underwhelming launch if it barely reaches Titan X performance or even beats it by 5%. They'd need at least a clear 15%+ advantage across the board over Titan X to really live up to the hype and being 3+ months late. Those that go Crossfire may be in for a world of pain given AMD's track record with driver and crossfire support.
Edit: All of you should also keep in mind that AMD isn't primarily in the business of making graphics, so you should temper your expectations.
So much speculation, so much debate over rumors and so many posts that are making claims without any sort of official release or reviews. How about we all take a deep breath and wait for the official announcement?
And do what in the meantime?
I Know (*_*) that on R390X non OC'ed Crysis 3 MSAAx2 will have constant 120 fps in 4k -> the Single GPU !!!!
Waiting for that card, and i will have it 100%
The HBM (AMD and HyniX tech) is So powerfull.
And new GCN arch. is astonishing.
Also there will be new ATI CATALYST Omega for them Brand new tech in drivers are coming. (they still working on optimisation for Phenom x4 x6 and FX line + APU )
Considering the Titan X can only hit 30 fps at 4K with Crysis 3, it would be a miracle if 390X could even hit 35 fps.
To do that it would have to be over 70% faster than a 290X.