He's talking about Hardware.fr's guess :Quote:
Originally Posted by boxleitnerb
http://translate.googleusercontent.c...ZomZXpNuPPoJLQ
Printable View
He's talking about Hardware.fr's guess :Quote:
Originally Posted by boxleitnerb
http://translate.googleusercontent.c...ZomZXpNuPPoJLQ
Lawl.Quote:
Originally Posted by Olivon
http://beyond3d.com/showpost.php?p=1...&postcount=446
http://www.3dcenter.org/news/erste-s...aii-grafikchip
?? :DQuote:
AMD Hawaii
28nm-Fertigung bei TSMC
~430mm? Chipfl?che (+18% gegen?ber R1000/Tahiti)
GCN 2.0 Architektur (R1000/Tahiti: GCN 1.0)
DirectX 11.2 (wie R1000/Tahiti)
4 Raster-Engines (+100% gegen?ber R1000/Tahiti)
maximal 2816 Shader-Einheiten aka 44 Shader-Cluster ? in jedem Fall mehr als die 2304 Shader-Einheiten der GeForce GTX 780 (+37,5% gegen?ber R1000/Tahiti)
Chiptakt (Hawaii XT): etwas mehr als 900 MHz (?hnlich der Radeon HD 7970, aber niedriger als bei der Radeon HD 7970 "GHz Edition")
Temperatur-kontrollierter Boost-Takt ? die Karte wird ihre volle Performance wegen Limits f?r Temperatur & Leistungsaufnahme regul?r nicht ausspielen k?nnen
wahrscheinlich nur ein 384 Bit DDR Speicherinterface (wie R1000/Tahiti)
Vorstellung: 25. September 2013
Verkaufsstart: wahrscheinlich Mitte/Ende Oktober 2013
Verkaufsnamen: wahrscheinlich Radeon R9-290 (Hawaii Pro) & Radeon R9-290X (Hawaii XT)
i spoke about 44 CUs + 4x ACE engines before month in this thread, it is nothing new here.
A lot of people mentioned 44CUs, I said something about it sometime in July and again early Aug.Quote:
Originally Posted by PedantOne
44CUs was suggested when there was a codename or driver list that got posted mentioning a "40PRO" SKU.
I spoke about Oreo cookies a few days ago so there's pretty much no reason to ever bring it up again ok?Quote:
Originally Posted by PedantOne
Wow! 37.5% SP increase and 100% RP increase with only 18% more area , great IF true. I am really looking forward to that 100% RP increase in Pro and LE versions too
PS: IS there a secret XS competition going on 'Predict and Win' :p: ?
Tao I think what it comes down to is, everyone is hoping and praying that this is going to be bigger than it might be. A lot of people feel they have lost Amd as a competetive entity in the enthusiast CPU market, and with only Nvidia and Amd in the discrete graphics market, enthusiast are anxious for some good competition between the two here. Amd needs to knock one out of the park here
Am I the only one who sees this going down just like Cayman did? Everyone gets their hopes up, finds out the specs were only 2/3rd what the specs claimed (because the die shrink wasn't ready for release), then we see the part holding those specs as the next gen (the 7970 was literally the exact specs we were hearing for the 6970)?
I'm not saying that is what is going to happen, but just simply taking a look at the situation it's pretty possible. It's pretty similar in build-up, that much is for sure.
You heard Cayman had 2048SPs and a 384bit bus?Quote:
Originally Posted by DilTech
Ah, well I'll be... Could've sworn I remembered the rumor saying 2048. They said 1920, so I was a bit off, but it's still a VERY similar situation we see going on right now. Guess it's true, once the kid is born your memory is the first thing to go!Quote:
Originally Posted by LordEC911
As for memory bus, I don't recall that aspect being spoken about much during the build up to cayman.
There was one rumor that mentioned the possibility of a 384-bit bus for Cayman.
If you know me, you know I'm a great lurker and I try to have the more informations I can get before buy.Quote:
Originally Posted by BababooeyHTJ
I just didn't see any good feedback on this card especially with the OC.
I got a friend who have a reference one and OC is not so good, especially past 1.25v (more juice give nothing, even with a good WB on it).
You think you're the center of rumors world ?Quote:
Originally Posted by LordEC911
As I already said to you, Damien don't give a shee about what you're sayin'
True, but that was also leaks mixed in with a canned higher performance ASIC.Quote:
Originally Posted by DilTech
I was honestly asking because I really only remember the 1920 rumor.
Like Dil, I don't recall that. I might have to go back through the Cayman threads.Quote:
Originally Posted by iMacmatician
Nope, I was just suggesting that Damien isn't the only person who mentioned that transistor count.Quote:
Originally Posted by Olivon
It is just amazing that a couple of days after I post something, he posts something strangely similar that hasn't been mentioned anywhere else.
Just like back in the good old days of the Inq, they were ripping off some of my speculation in the RV7x0 days.
Not implying he is doing that but it is amusing to me.
This is the Tahiti die.
http://www.abload.de/img/nfqygvru4k.jpg
All the blue stuff won't get any bigger. That's 37% of the die or about 135 sqmm. That leaves 230sqmm to get bigger.
From what we know (official sources) Hawaii die is ~430sqmm. I'm going to substract the blue stuff for the sake of it. That leaves us with 295sqmm.
295/230 = 1.28. This gives about 30% more space to play with. That's atleast enough for +25% SP and 50% ROPs. Add to that increases in process efficiency (less spare transistors), architecture improvements => those specs are extremely possible.
At least 2560 SP and 48ROPs. That's the min you can expect. Since 7970 is ROP starved and OCing mem doesn't do much it won't get any slower than 7970+30%.
Whatever the specs it better have a titan type cooler on it
Just a question. Will increasing the raster engine which I guess is the front end increase the size of the GCN cores or something else?
Will it be similar to cayman, where the cores themselves take up more space than before and as a result, no increase in cores actually occurs? Just more efficient use of the current cores?
Lets hope the GTX 780's price drops.
That and they release less cut down version of gtx and titan and give both cards a little more volts and a higher TDP limit.Quote:
Originally Posted by HelixPC
Why?Quote:
Originally Posted by HelixPC
If AMD release a card strong enough to force a price drop it means both companies are competing at the top end again, which in turn puts more high end cards in users hands. The knock on effect is that these high end card sales get noticed by developers (mostly via steam surveys) and they raise the target specification of their PC games, as there will be a larger target audience able to enjoy them.Quote:
Originally Posted by DefStar
Just guessing, but if the raster engine is a front end to the cores then it probably lives in the command processor and not the compute unit cores. A change there would change the size of the white part in the pic above.Quote:
Originally Posted by tajoh111
http://videocardz.com/45623/amd-hawa...r-kepler-gk110
If its been posted sorry!
Wow that is a bad article.Quote:
Originally Posted by Jethro
http://gyazo.com/5277601545751be08aad643a7d044fae.png
Im thinking about BF4 with eyefinity..
So, would people be happy with a card from AMD that has a performance level in the neighborhood of a GTX780/Titan? Sure it would be cool to get lower prices, but at the same time it's worrying. This is AMD's big upgrade, their new gen GPU, not just a GPU that is clocked a bit higher with some optimizations. Nvidias next gen will be out in, what, about half a year after AMD releases their new cards? And if Maxwell delivers even half of what is promised, AMD's next gen GPU is toast.
If Maxwell is on 20nm then AMD will not compete with Hawaii but rather, a new 20nm card obviously.
People need to realize that all performance expectations and predictions on Maxwell are based on it being at 20nm. There will be nowhere near that level of improvement if it is made in the same 28nm process, everyone knows the challenges of improving performance on the same node yet a lot of people expect a miracle 100% improvement of performance on Maxwell at 28nm. Cumon...
Besides that, AMD and Nvidia do not have identical manufacturing schedules, just because one releases a product today, doesn't mean the other can and should release a new product equally competitive in the same period.
I think AMD are trying to make GPGPU much more important this generation, the 680 and 780 are weak in this area while the 79*0's are strong. For example I said a few months back, that GPU compute was going to become very important:Quote:
Originally Posted by nossy23
http://www.youtube.com/watch?v=vF1zjDSqPoo
After seeing this, if Nvidia are going to make a Maxwell with most of it's GPU compute removed it's going to be very vulnerable next year, especially with AMD's developer ties. So it doesn't matter if the benchmarks next month say it's only on par with Titan, what really matters is have AMD got the compute balance right to make all the PS4 and XBOne ports run like silk. Engines built to take advantage of GPU compute might become the future, and for once AMD are holding the reins.
Why? Hawaii is meant to compete with GK110...Quote:
Originally Posted by nossy23
i am sure, if will have gaming SKU 2816 SPs, can beat Titan in performance. I hope for 512-bit mem bus, but not for width but 4 GB RAM.
No worries from the Red Team :)Quote:
Originally Posted by nossy23
My understanding: Volcanic Islands is more of a optimized *tweener* card still in the architectural range of Southern Islands, but a major preparation for the die-shrink to 20nm.
Bigger chip, expanded power-states, higher frequencies (if Bonaire, and Cypress ---> Cayman are any guide) My guess: 5.2 billion transistors, 2688 cores and 440mm2
Not quite a Titan, but Jen-Hsun Huang will be sweating bullets on those GTX780 price drops :D
You have to understand that cutting down GPGPU is a choice made by Nvidia. GK110 is better at GPGPU than anything AMD has to offer right now; the thing is, Nvidia makes a lot of money on professional GPU segment (unlike AMD), so they cripple gaming GPUs on purpose to get people who do need GPGPU performance to spend more.Quote:
Originally Posted by Iconyu
Once there is sufficient demand for gaming GPGPU performance, I am sure they will step up.
This is very unfortunate, but there's absolutely nothing we can do about it. It's their business model.
Not happening, would increase die size and PCB cost too much. Not really feasible.Quote:
Originally Posted by PedantOne
It's called FirePro.Quote:
Originally Posted by Elmy
If anyone didn't know they got a killer promotion for 50% off the W7000 & W5000 till Oct 31.
Aye, Nvidia's philosophy is to build the best card they can for today's games, tomorrows games will need another card.Quote:
Originally Posted by zalbard
Edit: @LordEC911: Aye that's pretty much it, when Nvidia were in Crytek's ear Crysis 2 went tessellation crazy, AMD now pretty much has every console game developer for the next 5 years under their sway, even the WiiU can exploit these features. If these particular strengths are exploitable in a way that actually makes games look better then it would be stupid to ignore them. But I do imagine that most developers won't allow it to get too out of hand and literally build games that cripple a 680.
Think more with transistor count. I estimate, depending on the exact die size but assuming around 440mm2, around 5.5b-6b.Quote:
Originally Posted by AbortRetryFail?
I think, though you did mention it, he was specifically talking about gaming compute performance, possibly OpenCL?Quote:
Originally Posted by zalbard
They didn't remove any of that for GK110 dies and they still are getting beat by Tahiti.
http://www.anandtech.com/show/6973/n...-780-review/18
Possibly driver related?
512b is possible by using a more area efficient PHY like in Bonaire/Pitcairn but lower speed would easily give +300Gbps of bandwidth while decreasing the area of the PHY by a third or about 66% the size of Tahiti's total PHY. There is also the comment that Stilt made about a rectangular die which suggests they are looking to maximize the perimeter.
We already know that Hawaii has a more complex PCB than Tahiti, more layers, so that could hint at either faster memory speeds(7Gbps) or a larger bus(512b) or just more power.
Geforce vs quadro, on the 680's they disabled dp, the titan though has this option as a gaming card (oddly enough).
Then you also have the usual diff's between quadro and geforce, a loss of a tiny bit of detail when drawing distances with mip mapping and some speed differences.
All in all, the 680 chip still has that stuff..., but on a geforce pcb it won't.
You got the 2880 shader k6000 (pro, dp), the 2668 shader titan (dp, gaming), the 2304 shader 780 (sp, gaming), and the 1536 shader 680 (sp, gaming).
(Edit: oh and who can forget the monster of a pro card the k2, 3072 shaders and dp, equ to the gaming card the 690 which only does sp)
The 2 that stick out in terms of dp ability are the titan and the 680, the 680 was originally a k5000 pro chip.
As for the 780, maybe it has the ability, maybe there was a quadro that was the exact equ, but to my knowledge it doesn't.
The 780 was a step down in costs compared to the titan.
Both really aren't the full fledged quadro equ.
I got the 680, now as far as gaming is concerned I wouldn't mind an upgrade, but once you go quadro, you wouldn't wanna jump ship so quickly, and that's how I feel.
I'll stick with it until I can find another exact gaming vs pro equ chip.
That and I don't have the money to jump onto another vga card so soon ;).
But I like to see prices drop, and many people don't mind amd/ati at all, I don't think dp is a big issue right now.
I think it's more or less a quality and feature thing then a performance thing.
I think the main issue this gen and the next is vram, and texture performance, because eventually we're gonna have to get an upgrade in the tmu department (without needed a giant expensive die that is...).
We ain't gonna be pushing the same amount of textures at the same old resolution for ever...
Edit:
In a few more gens from now, they'll have to focus on 3d monitor perf, they demo'ed last year some nice non-glasses 3d screens with viewing angles for up to 20 so on people at a time.
Sooner or later those will hit mainstream.
And performance right now in that department is a little meh, it could be better is what I mean.
But then again we have some games only using one cpu core :\ ...
It is?Quote:
Originally Posted by zalbard
GK110 seems to go back and forth between Tahiti in GPGPU benchmarks, and that's at a significant die size advantage. GCN seems like it's at least as good at general purpose tasks architecturally (a theoretical Tahiti scaled up to GK110 proportions would likely thrash GK110 in the majority of GPGPU benchmarks), though Nvidia obviously has a software advantage with CUDA.
wonder how the cost compares of 4g 512bit vs 6g 384bit if its close i think i would rather the extra memory
as it has the potential to make the most noticeable difference but only if games are made to make use of it
although 7970 has ample memory bandwidth and increasing memory clock speeds from 5600-7600 makes very little difference that may change if what amd has for us is significantly faster
does titan see much benefit from overclocking its memory?
The drivers haven't been made for quadro k6000 yet so I don't think we can judge it yet.Quote:
Originally Posted by hurleybird
As far as a compute card, tahiti is just so so for a next gen card. IT performs well for consumer purposes but not so well in the professional space. The firepro w9000 cards had tahiti and quadro 6000 fermi basically trading blows in the reviews available online(tomshardware and hothardware). Considering the computer power of tahiti it should simply decimate fermi but it doesn't.
http://www.tomshardware.com/reviews/...mark,3265.html
http://hothardware.com/Reviews/AMDs-...vidias-Quadro/
If AMD puts too much of a gaming emphasis on Hawaii, it's going to screw up the gains it made with tahiti. And particularly in the professional market, AMD needs to make bigger improvements, even more so than the gaming market for its top end card.
How well AMD big chips perform in professional spaces is what's going to make them money not so much its consumer performance(at least not for the next 3 years).
Those aren't GPGPU benchmarks, and yes AMD needs to improve their Firepro drivers.Quote:
Originally Posted by tajoh111
Not sure why AMD would bring out their highest end GPU with just 3 GB of ram. Sure it's more than enough today. But with next gen consoles that have a lot of memory that can be used for graphical things, and taking in to account on PC we use (higher) AA etc, I can see the memory becoming a bottleneck.
The things Tahiti lacks, IMO, that Hawaii will make up for is specifically in the compute department.Quote:
Originally Posted by tajoh111
OpenCL is ignored to get more people to use CUDA. Whether that's wise or not, time will tell.Quote:
Originally Posted by LordEC911
Pretty stupid, in my opinion, since some people will keep using OpenCL as long as CUDA only runs on Nvidia hardware.
Proprietary stuff rarely wins in the long run.
Not going to happen before OpenCL development is as easy and convenient as CUDA's, though...
GK110 has a lot of GPGPU features that AMD offers lack. It's not just performance.Quote:
Originally Posted by hurleybird
Wow so we get 4 days warning of a solid release date? Had I known this a week ago I would have held off buying my 780 Lightning, I am sure it will do the job and I know I can always sell it but I would have liked to see how these go before lashing out.
:)
Care to elaborate?Quote:
Originally Posted by zalbard
Ruby is so hot!
On the 25th AMD will be streaming at 3 EST 12 PST on their Youtube page about their new announcement.
So excited to see what they are bringing to the table!
Would be great to see this happen.Quote:
Originally Posted by LordEC911
You have to remember this generation will be a very short lived one. As soon as 20nm is feasible for high-end we'll see new SKUs from both Nvidia and AMD.Quote:
Originally Posted by nossy23
maybe mid-range market feelers, but I doubt top end SKU's will be "soon"
Dynamic Parallelism and Hyper-Q, for example. AMD GPUs don't support anything like that (yet).Quote:
Originally Posted by hurleybird
CUDA, obviously, and it's more popular / convenient than OpenCL...
Their virtualization features are also very neat, but I haven't worked with them, so cannot comment on it.
Right now it's actually pretty simple. If you really want OpenCL only, you buy AMD. Otherwise, for GPGPU you buy Nvidia... And no, I don't mean gaming cards.
Yes, you mean gaming CARD: Titan, Quadro for best priceQuote:
Originally Posted by zalbard
Attachment 131327
Looks pretty slick :cool:
Haha you guys beat me to it by seconds haha
Pretty impressive if it has 512-bit memory bus. Bandwidth would be insane. Wondering about pricing, though...
I mean what I mean. Don't put words in my mouth.Quote:
Originally Posted by PedantOne
"Haha you guys beat me to it by seconds haha"
Seconds are important...:lastweek:
/jk
I sense the first stirring of unease in the Nvidiasphere :caution:
Size of a Hynix GDDR5 2Gbit IC is 14x12mm ;)
TYQuote:
Originally Posted by The Stilt
for all you crazy kids about to count pixels...
Quick and sloppy, I got ~24.5x18mm for 441mm2
I think a 512bit memory bus is kind of excessive unless AMD architecture is that reliant on it. I think this kills alot of the bigger spec improvements because such a size bus is going to take up a crazy amount of room.
Or be ~66% of the Tahiti PHY.Quote:
Originally Posted by tajoh111
Dang I was hoping to be surprised...almost made it lol
Yep that's 512bit memory, each chip handles 32bits of the crossbar.
crazy ...
512 bits bus back again !!!
with bf4 OMGGGGGGG
Tahiti sized PCB.
8+6pin
6+2 PWM
Looks like 2 DL-DVI + other outputs (yay for when I get my QX/DP2710)
16 memoy ICs all on front, none on back
Has some sort of mod to it, soldered wire.
Also has a switch on it too, maybe it has dual bios...
Ofc it have dual bios.. It wa surely the best idea they have got for 7970 ( and 6970 maybe )... they will not remove it.Quote:
Originally Posted by NEOAethyr
Clean and nice PCB.
You think AMD is going for 512bit memory now, because being able to transfer a lot of graphical data will become more important, as next gen consoles will be using much larger textures etc as compared to current gen consoles?
AMD is going for 512-bit for two reasons:
1. HPC
2. Efficiency (low clocked 512-bit probably consumes less energy than 384-bit with high clocks)
Most likely, they're going 512-bit because:Quote:
Originally Posted by nossy23
-They have a big enough perimeter to do so
-They can fit that 512-bit bus in the same or less area as Tahiti's 384-bit bus, as per Dave B
-4GB of memory is a good sweet spot -- 3GB would be underwhelming, and 6GB would be largely unnecessary
-More bandwidth will help with ultra high resolution gaming and performance in general
-A 512-bit interface with slightly lower clocks is probably more energy efficient than a 384-bit interface with bleeding edge clocks
Looks like a solid design choice overall.
hopefully the cooler fan isn't as loud as the 7970's was.
i can see it now this card beating the titan but having a loud cooler and getting horrible reviews lol
It looks quite promising. Beating Titan seems possible.
I have a Geforce Titan but I really hope this AMD card performs really well. We need the competition to move things forwards. AMD already :banana::banana::banana::banana:ed up the Desktop CPU's :(
This is possible to see titan behind. I think its gonna happen.Quote:
Originally Posted by nossy23
But, If I would be you, I would be angry at nVidia to pay so much, for getting the top of frame for only a few months.
And its not 500$, but 1000$ ...
I really hope that AMD doesn't wait 3 months to release 9950 or whatever they're going to call the cut down Hawaii.
What is the rumored release date at this point?
maybeQuote:
Originally Posted by Olivon
http://wccftech.com/amd-hawaii-r9290...z-clock-speed/
Attachment 131328Quote:
Originally Posted by BababooeyHTJ
HmmmmQuote:
Originally Posted by flopper
http://translate.google.com/translat...%2F539&act=url
Great if true. Nvidia's been charging too much for their chips lately and simply releasing their cards too slowly.Quote:
Originally Posted by m_jones_
Faster than Titan, and more energy efficient. This leak looks legit.
http://udteam.tistory.com/539
With those numbers,if this thing comes in at under $600 then its a clear win for AMD.:up:
Will be an extremely nice early Christmas present if they are able to pull it off.Quote:
Originally Posted by hurleybird
If its under $600 i'll pick one up.Quote:
Originally Posted by xray
At $700 it might take me some convincing...
if its 500 im buying two lol
Any pics of under the hood look at the heatsink yet?
Can't wait to see a dual Hawaii card ;-)
these results are with oced second bios, on default it is below titan, but still great performance.
PS. it seem all my info here was true, my source was great this time, 512-bus, 2560-2816 SPs for gaming SKU, Titans performance :)
Attachment 131329
Sure it is 2560 SP?
it is info from leaker, but still can be 2816 for sure
Where, I don't see it.
If it really does have a 512 bit memory bus, insane, though I wonder how much its going to cost, and I would prefer to see a 8GB version if this is going to beat a Titan.
Hawaii is 44/176/44 (SIMD, TMU, ROP).
4 CUs with 11 SIMDs & ROPs each.
== "Mighty 44" ;)
by the same source, will cost 650 USD
Stilt: You was wrong with die size, maybe you are wrong with this too :)
Why 44Rops ? 48 seems more plausible as whatever they do they have it decoupled from the crossbar. ( 32bits for each IC ) Or it have completely change from Tahiti.Quote:
Originally Posted by The Stilt
44x 64 = 2816 SP ..
So if it is decoupled, why would 48 be more plausible? Since those two values are independent from each other, it can be anything. 40, 44, 48, 64...
The block diagram shows 44, thats why :D
So were you (470-480 sqr mm).Quote:
Originally Posted by PedantOne
I just scaled two very different sized BDs against each other in Photoshop.
by few sites, measured gpu from leaked Picture is aroun 420mm2, by the few days old interview, it is too around 420mm2
great to see things materialise as rumoured, soo soon.