Anyone actually believe this?
I wonder how long it will be before ek releases a water block for it?
I'm looking specifically at 1.4v and the performance of the stock cooler at 1.4v and 1200MHz.
I wasn't looking at the card snapper cooler.
I have used car salesmen phobia. :|
Yeah, I'm not reading the entire thread, but what caught my eye was the double precision was cut to 1/8 on the 290X. Notice how it's slower than the 280X and even the 780 on the benchmarks. Obviously that's not really important in games, but I thought it was interesting.
Interesting that no one really brought this up, wonder if that's artificial or baked into the design. No Firepro version of Hawaii?
Presumably this is how they got the performance density they did out of the card. By stripping back the double float performance they reduced the size of the die making it a great gaming card (similar to how the 680 is stripped back in a similar way to focus on gaming) but poor on compute. That might also explain the heat of the chip, they aren't spreading the heat out over as large an area and the stream processors are getting higher utilisation as the hardware for the double floats isn't sitting idle in the game.
Interesting choice on AMDs part.
but poor on compute.
After years of neglect, AMDs [graphics] workstation group, under the tutorage of Matt Skyner, has the backing and commitment of top management and AMD intends to push into the market aggressively. We have seen them gain market share this year and expect them to gain even more, said Jon Peddie, the head of Jon Peddie Research.
Ah, interesting. 1/8 DP is a bit disappointing, but not very relevant to enthusiasts. I'm also wondering if there is an artificial limit at play.
We've also come to learn that AMD changed the double-precision rate from 1/4 to 1/8 on the R9 290X, yielding a maximum .7 TFLOPS. The FirePro version of this configuration will support full-speed (1/2 rate) DP compute, giving professional users an incentive to spring for Hawaii's professional implementation.
Anandtech R9 290x Review said:A Note On Testing Methodologies & Sustained Performance
We also want to make quick note of our testing methodologies and how they are or are not impacted by temperature based throttling. For years we have done all of our GPU benchmarking by looping gaming benchmarks multiple times, both to combat the inherent run-to-run variation that we see in benchmarking, and more recently to serve as a warm-up activity for cards with temperature based throttling. While these methods have proved sufficient for the Radeon 7000 series, the GeForce 600 series, and even the GeForce 700 series, due to the laws of physics AMD's 95C throttle point takes longer to get to than NVIDIA's 80C throttle point. As a result it's harder to bring the 290X up to its sustained temperatures before the end of our benchmark runs. It will inevitably hit 95C in quiet mode, but not every benchmark runs long enough to reach that before the 3rd or 4th loop.
For the sake of consistency with past results we have not altered our benchmark methodology. However we wanted to be sure to point out this fact before getting to benchmarking, so that there’s no confusion over how we’re handling the matter. Consequently we believe our looping benchmarks run long enough to generally reach sustained performance numbers, but in all likelihood some of our numbers on the shortest benchmarks will skew low. For the next iteration of our benchmark suite we’re most likely going to need to institute a pre-heating phase for all cards to counter AMD’s 95C throttle point.
Presumably this is how they got the performance density they did out of the card. By stripping back the double float performance they reduced the size of the die making it a great gaming card (similar to how the 680 is stripped back in a similar way to focus on gaming) but poor on compute.
Well congratulations to AMD for reaching whatever price point. But I think we all know this costs them in the long run. Brand recognition doesn't come easy, and this is a sure way to deter that from happening. If you're going to be smart (design an excellent GPU) then be smart the whole way. Don't walk up 20 inches from the finish line and cut your own achilles tendons. It's completely not self serving and AMD needs to be self serving if they want the appreciation, not condemnation, of their consumers.
AMD is essentially selling a flawed product. But due to clever marketing of not announcing a baseclock.
The chip is anything but a marvel. Its simply too leaky and even with a 20% OC power consumption rises 200W. But I guess the smaller die was a nice checkbox for the marketing dept.
But thats just another lost oppotunity in AMDs long list of such. And with a year or so to 20nm. Its a long time to carry these blowtorches.
It's most likely deliberate.The 290 has incredible potential. Give credit where it is due. But that cooler makes me just want to line up the AMD engineering staff and smack the dumb looks off their faces. To throw a cooler like that on an engineering feat like the 290? It's criminal. How does something like this get through? What are they thinking?
It's most likely deliberate.
Think about it. AMD has used the same design cooler for how many generations now? It's adequate, cost effective and gets the job done. But it isn't as capable as the aftermarket designs that will eventually follow. AMD can design a 6.2 billion transistor GPU that is the fastest in the world, but their engineers are incapable of coming up with a HSF design to equal or beat aftermarket coolers? I don't buy it.
What I've suspected for awhile now is that AMD deliberately sticks with their blower design because it isn't as good as aftermarket designs. This allows AIBs to come out with their own cooling designs to differentiate themselves. After awhile, the reference cards won't even be available anymore and all you will see in the market will be custom designs.
Remember, AMD doesn't sell cards themselves anymore. They supply the GPUs/reference cards and allow the AIBs to redesign the cooling, the PCBs and the components to suit themselves.
Personally, I think it's all part of their marketing plan.
jumped 30 pages but how many gtx780Ti can you buy now]How many custom cards can you buy now? [/B]None? So reference is the only one we can relate to right now. Its a card running 94C nomatter what mode. Are you saying reference cards shouldnt work because they are reference? Broken products are obviously ok with you.
I saw someone who floated a theory that because Hawaii runs hot AMD didn't hand out carte blanche to AIBs to churn out their cooler designs at will and is perhaps wanting to vet them before allowing them to be used.
It makes some sense because some aftermarket coolers are actually atrocious and perform worse than reference. Some of the truly bad ones ignore proper VRM cooling which could be an RMA disaster waiting to happen with the hot running 290x.
We know there is a DCU2 coming as there is a SKU up on newegg already, a well known competent cooler, but have seen nothing else yet.
show me a full GK not a core cut/crippled , like a full core , 1x cut $1000 ,2nd cut $650.00 give me a break.Obviously also an nVidia hater. Since I first added focus to their problems between baseclock and boost. Since Hawaii is the first chip from AMD where you can actually read the boost clock and not just get no boost or all boost as result.
I find it amazing how diligent some people defend broken products to protect X/Y company.
That's not accurate.
"We've also come to learn that AMD changed the double-precision rate from 1/4 to 1/8 on the R9 290X, yielding a maximum .7 TFLOPS. The FirePro version of this configuration will support full-speed (1/2 rate) DP compute, giving professional users an incentive to spring for Hawaii's professional implementation."
http://www.tomshardware.com/reviews/radeon-r9-290x-hawaii-review,3650.html
This is simply an artificial decision for market segmentation. Rainy beat me to it
In farcry where the difference is, notice fan stopped at 65% anyway. Good or bad news?
Well it means for sure that if you put it at 100% you get every inch of performance from this card in all games.... if your case is well ventilated. Lol.
I saw someone who floated a theory that because Hawaii runs hot AMD didn't hand out carte blanche to AIBs to churn out their cooler designs at will and is perhaps wanting to vet them before allowing them to be used.
It makes some sense because some aftermarket coolers are actually atrocious and perform worse than reference. Some of the truly bad ones ignore proper VRM cooling which could be an RMA disaster waiting to happen with the hot running 290x.
We know there is a DCU2 coming as there is a SKU up on newegg already, a well known competent cooler, but have seen nothing else yet.
It's most likely deliberate.
Think about it. AMD has used the same design cooler for how many generations now? It's adequate, cost effective and gets the job done. But it isn't as capable as the aftermarket designs that will eventually follow. AMD can design a 6.2 billion transistor GPU that is the fastest in the world, but their engineers are incapable of coming up with a HSF design to equal or beat aftermarket coolers? I don't buy it.
What I've suspected for awhile now is that AMD deliberately sticks with their blower design because it isn't as good as aftermarket designs. This allows AIBs to come out with their own cooling designs to differentiate themselves. After awhile, the reference cards won't even be available anymore and all you will see in the market will be custom designs.
Remember, AMD doesn't sell cards themselves anymore. They supply the GPUs/reference cards and allow the AIBs to redesign the cooling, the PCBs and the components to suit themselves.
Personally, I think it's all part of their marketing plan.