Wednesday, May 8th 2013

AMD's Answer to GeForce GTX 700 Series: Volcanic Islands

GPU buyers can breathe a huge sigh of relief that AMD isn't fixated with next-generation game consoles, and that its late-2013 launch of its next GPU generation is with good reason. The company is building a new GPU micro-architecture from the ground up. Codenamed "Volcanic Islands," with members codenamed after famous islands along the Pacific Ring of Fire, the new GPU family sees AMD rearranging component-hierarchy within the GPU, in a big way.

Over the past three GPU generations that used VLIW5, VLIW4, and Graphics CoreNext SIMD architectures, the component hierarchy was essentially untouched. According to an early block-diagram of one of the GPUs in the series, codenamed "Hawaii," AMD will designate parallel and serial computing units. Serial cores based on either of the two architectures AMD is licensed to use (x86 and ARM), could handle part of the graphics processing load. The stream processors of today make up the GPU's parallel processing machinery.

We can't make out text in the rather blurry block-diagram, but are rather convinced that if it's authentic, then AMD is making some big changes. Another reason for AMD's delay could be silicon fab process. "Tahiti" as implemented on Radeon HD 7970 GHz Edition, already poses high thermal envelope. AMD doesn't want the 28 nm process to restrict its next-generation architecture development, and is holding out till the 20 nm process is in place at TSMC. The fab set Q4 as its tentative bulk manufacturing date for the process.

The source that leaked the block-diagram also posted specifications of the chip that's codenamed "Hawaii," which appears to be the flagship part.
  • 20 nm silicon fab process
  • 4096 stream processors
  • 16 serial processor cores
  • 4 geometry engines
  • 256 TMUs
  • 64 ROPs
  • 512-bit GDDR5 memory interface
Source: ChipHell. Many Thanks to SIGSEGV for the tip!
Add your own comment

145 Comments on AMD's Answer to GeForce GTX 700 Series: Volcanic Islands

#1
theoneandonlymrk
Wow compute cored gfx is around the corner nice , whilst the pic may be a hoax it does all make perfect sense , imho those serial cores will either be a new special Dp shader sub unit aimed at serial compute or something much more interesting like Jaguer cores perhaps;) but they are not at all like an smx unit btw ( cant remember who said that) they will be separate and special , I think the bus/fabric that binds it all will be very very interesting too.
Posted on Reply
#2
Mathragh
by: W1zzard

this slide might be a pipedream of some chinese forum user who's laughing his ass off now
Aye, all that's new with this pic is the fact that its a negative of the original apparently :p
Posted on Reply
#3
Casecutter
With word on the street of Apple looking at A7 chip production at TSMC, perhaps this is why AMD a GloFlo might be worth reflecting on. Plus demand 20Nm parts where to be high at least what TSMC and others where saying. While we know Nvidia signed with TSMC last December for 20Nm with Maxwell in mind. Then when you search "AMD TSMC 20nm" the information trail is really cold?
Posted on Reply
#4
FrustratedGarrett
by: theoneandonlymrk
Wow compute cored gfx is around the corner nice , whilst the pic may be a hoax it does all make perfect sense , imho those serial cores will either be a new special Dp shader sub unit aimed at serial compute or something much more interesting like Jaguer cores perhaps;) but they are not at all like an smx unit btw ( cant remember who said that) they will be separate and special , I think the bus/fabric that binds it all will be very very interesting too.
The small custom Arm CPUs might be used for scheduling complex workloads with thousands of threads, and perform some heavy duty in-order or serial processing.

The specs do make sense, but like Wizzard said, it could be a fabricated news. I hope it's not, cuz BF4 is coming out later this year, and I want to upgrade to a worthy card that can max BF4 out on a 120Hz 2560x1440 res. display.

on another note: Are we getting any 8 bit, backlight-strobed displays this year with 120Hz refresh rate? I mean with those 8 bit panels that are rated at 6ms pixel transition times, it's perfectly possible with a strobing backlight to achieve perfectly smooth and clear animation that surpasses that of 120Hz TN displays with 2-3ms pixel response ratings.
Posted on Reply
#5
TheHunter
by: Casecutter
With word on the street of Apple looking at A7 chip production at TSMC, perhaps this is why AMD a GloFlo might be worth reflecting on. Plus demand 20Nm parts where to be high at least what TSMC and others where saying. While we know Nvidia signed with TSMC last December for 20Nm with Maxwell in mind. Then when you search "AMD TSMC 20nm" the information trail is really cold?
Bring on 2014: Volcanic Islands

After Sea Islands reaches the end of the line, AMD will introduce perhaps the most important GPU family of them all - Volcanic Islands. Volcanic Islands (VI) will go head to head against NVIDIA Maxwell and second-gen Xeon Phi architecture.

Manufactured at 20nm Gate-Last process, this will be the first GPU family which AMD should be able to manufacture in Common Platform Alliance as well as its long-standing foundry partner, TSMC. Thus, AMD will have the choice between TSMC GigaFab Hsinchu/Taichung, IBM East Fishkill, GlobalFoundries in New York and Dresden or Samsung in Austin. The manufacturing flexibility will be of paramount importance, for Volcanic Islands GPU architecture will represent the pinnacle of system integration between the CPU and GPU.
http://vr-zone.com/articles/amd-next-generation-codenames-revealed-2013-2014-2015-gpus-get-names/17154.html
Posted on Reply
#6
15th Warlock
by: W1zzard
you can buy titan today, and it's tested to be fast and stable with decent power consumption, low fan noise and no coil noise.

this slide might be a pipedream of some chinese forum user who's laughing his ass off now
Exactly, it seems like Titan is getting a bad rap from a lot of ppl here lately, calling it a crappy, high leakage GPU when it's anything but. Yes the price is outrageous, but the performance and low power usage are there.

Anyways, this thread is not even about Titan, so moving on.... it seems like AMD is shooting for a GPU with more general purpose focus, if these cores are ARM cores they will use a cut down version of this GPU for mobile devices, depending on power consumption, this might actually be the breakthrough AMD needs to get their designs into more mobile devices :rockout:

Let's not forget that the green team is also an ARM licensee, and they already have parts that couple A15 cores with Kepler cores in the pipeline, GPU designs take over aprox 24 mos. to reach the market, so they may be working on a similar design using Maxwell cores as well, but at this point this is all speculation :)
Posted on Reply
#7
AlB80
It's not GPU pic!!! It's APU pic!!!
Posted on Reply
#8
theoneandonlymrk
by: 15th Warlock
Exactly, it seems like Titan is getting a bad rap from a lot of ppl here lately, calling it a crappy, high leakage GPU when it's everything but. Yes the price is outrageous, but the performance and low power usage are there.

Anyways, this thread is not even about Titan, so moving on.... it seems like AMD is shooting for a GPU with more general purpose focus, if these cores are ARM cores they will use a cut down version of this GPU for mobile devices, depending on power consumption, this might actually be the breakthrough AMD needs to get their designs into more mobile devices :rockout:

Let's not forget that the green team is also an ARM licensee, and they already have parts that couple A15 cores with Kepler cores in the pipeline, GPU designs take over aprox 24 mos. to reach the market, so they may be working on a similar design using Maxwell cores as well, but at this point this is all speculation :)
I think thats what nv are working towards with project denver (arm 64/32bit cpu) and project Parker ( denver with maxwell gpu) as jhen hsung himself stated but they're aimed optimistically at 2015 with that and it also unifies all system memory.
Amd are beating many to the ball with this tech and due to their apu , gpu and soc achievements they are certainly one to bet on imho.
Posted on Reply
#9
HumanSmoke
by: Slacker
Getting off topic here, Hawaii isn't part of the Pacific Ring of Fire. It ain't a composite volcano sow why name it that? Amd needa hire some people that really know their stuff.
With marketing ? Why start now ?
These are the same people who name a Southern Islands GPU after a Northern hemisphere island archipelago (Cape Verde)

AMD ain't care.

by: Casecutter
This is more interesting than I thought it would get, if this is true it's going to be a stagnate 6 months. And I might be wrong! Nvidia might... might wise-up and drop all their GTX7XX prices to grab sales before this information really gets traction.
If this information gets traction, How many 7970/7990/7950's do you think AMD are going to sell ? You really think that a purported AMD slide (and lets face it, don't you find it just a little bit suspicious that this surfaces just as the GTX 700 hype machine gains momentum?) is going to headshot Nvidia's sales but leave AMD untouched ?
You do remember AMD Osborning themselves with the Trinity hype that resulted in the company taking a write down on Llano inventory recently ?
by: Casecutter
Just wondering could this be on GlobalFoundries 20nm?
AFAIA, GloFo's 20nm HKMG so far is of the low power (LPM) variety, 20nm TSV isn't slated for volume production until 2014.

Still, the "late-2013" launch, could mean just about anything (and I'm presuming the original time frame originates from James Prior's addition at the end of this piece) - tape out, risk wafer debug/validation, volume wafer starts laid down, shipping for revenue, or even ....retail availability...but if it's on 20nm then the signs are that it would be TSMC.
Posted on Reply
#10
D007
by: amdftw
Sorry but AMD's single 7970 crad has better frametimes than NV's GTX680...
http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-Dissected-Full-Details-Capture-based-Graphics-Performance-Testin
by: MxPhenom 216
I might be blind but I'm not seeing it and i just checked every graph. they are nearly identical.
Not to mention it's a GHZ edition he's clocking against.. All that graph shows is just how terrible Crossfire is.. Look at what you get when you sli 2 GTX 680 vs Crossfire..

Also if you are going to compare an ATI superclocked edition in a vs comparison, compare it to an NVIDIA superclocked edition.
Not an NVIDIA standard model..
Posted on Reply
#11
erocker
by: HumanSmoke
Osborning
I haven't heard that term before. I like it!
Posted on Reply
#13
theoneandonlymrk
by: HumanSmoke
With marketing ? Why start now ?
These are the same people who name a Southern Islands GPU after a Northern hemisphere island archipelago (Cape Verde)

AMD ain't care.s


If this information gets traction, How many 7970/7990/7950's do you think AMD are going to sell ? You really think that a purported AMD slide (and lets face it, don't you find it just a little bit suspicious that this surfaces just as the GTX 700 hype machine gains momentum?) is going to headshot Nvidia's sales but leave AMD untouched ?
You do remember AMD Osborning themselves with the Trinity hype that resulted in the company taking a write down on Llano inventory recently ?

AFAIA, GloFo's 20nm HKMG so far is of the low power (LPM) variety, 20nm TSV isn't slated for volume production until 2014.

Still, the "late-2013" launch, could mean just about anything (and I'm presuming the original time frame originates from James Prior's addition at the end of this piece) - tape out, risk wafer debug/validation, volume wafer starts laid down, shipping for revenue, or even ....retail availability...but if it's on 20nm then the signs are that it would be TSMC.
Tsv @ any size is just chip stacking and id say maxwell/ volta will get to use that first, I doubt Amd are daft enough todo a complete refresh of architecture inc bus changes and major Ip insertion and drop process size as well and then add several more (first time out Tsv) complications, all Imho :).
Posted on Reply
#14
Vinska
by: Mathragh
This might not be so new btw, someone at tweakers.net forums showed that this picture has actually been floating around the webs quite a while now:

http://www.flickr.com/photos/8091090@N02/8006861851/in/photostream/

So it could either actually be really old news, or a fake.
IF this thing is not fake, it resembles much more of a SoC than a GPU.
Especially so because those serial processing modules have an uncanny resemblance to bulldozer/piledriver cores...

EDIT: OH WAIT this thread had a second page *derp*

EDIT #2: OH TEH WAIT AGAIN this much more resembles a ramped-up version of the PS4 :|
Posted on Reply
#15
cdawall
where the hell are my stars
by: drdeathx
That's ludicrous untill you see the performance IMO. Looks like this may blow Nvidea out of the water.
I would hope the next generation cards perform better than the current generation of cards. :laugh:

As for manufactured news I don't care bumped the stock up another $.25 :)
Posted on Reply
#16
HumanSmoke
by: theoneandonlymrk
I doubt Amd are daft enough todo a complete refresh of architecture inc bus changes and major Ip insertion and drop process size as well and then add several more (first time out Tsv) complications, all Imho :).
Well, Southern Islands accomplished most of that.
Bus change...256-bit bus width to 384
IP insertion...VLIW4 to GCN
Process node...40nm to 28nm
TSV....I was under the impression that TSMC had already demonstrated viability under 28nm (CoWoS) over a year ago. TSV looks to be an extrapolation of the process.
Posted on Reply
#17
D4S4
looks like soon enough gfx cards will be gaming pc's all on their own

edit - also, take a look at the new play station hardware, kinda similar? i know little about programming but i imagine porting to a system that has cpu and system memory separated by a comparatively slow, high latency pcie bus could be a hassle. now if you had 8gb of memory and some cpu-like cores on the card?
Posted on Reply
#18
Casecutter
by: TheHunter

Wasn't it Southern Island 7XXX, then Sea Island, while Volcanic Islands would via Maxwell? Well, it looks like Sea Island is gone and the road map has move-up.

:toast:
Posted on Reply
#19
theoneandonlymrk
by: HumanSmoke
Well, Southern Islands accomplished most of that.
Bus change...256-bit bus width to 384
IP insertion...VLIW4 to GCN
Process node...40nm to 28nm
TSV....I was under the impression that TSMC had already demonstrated viability under 28nm (CoWoS) over a year ago. TSV looks to be an extrapolation of the process.
Tsv was first done by Ibm 1-2 years ago but it is just a chip stacking tech which also conveniantly enables odd stacks like 20nm on (45-90nm) but isn't directly tied to lithographic size and by bus I didn't mean adding more memory controllers I ment internal inter Ip buses which will have also increased a vast amount id imagine.
Im sorry though i should have said, I do agree with your original point that Amd will be waiting on 20nm finfet bulk process or Soi gate last I forget the actual process names.
Posted on Reply
#20
drdeathx
by: cdawall
I would hope the next generation cards perform better than the current generation of cards. :laugh:

As for manufactured news I don't care bumped the stock up another $.25 :)
Agreed CD:toast:
Posted on Reply
#21
esrever
by: Casecutter
Wasn't it Southern Island 7XXX, then Sea Island, while Volcanic Islands would via Maxwell? Well, it looks like Sea Island is gone and the road map has move-up.

:toast:
sea island is refresh of southern island. It also includes some new chips like bonaire and mars. There may be more cards coming by end of the year, especially for mobiles. CI isn't gone, its just going to be delayed a bit a merged with existing products. VI will be more than a year away from the looks of it.
Posted on Reply
#22
Casecutter
by: HumanSmoke
don't you find it just a little bit suspicious that this surfaces just as the GTX 700 hype machine gains momentum?)
We know this is Spy-Vs.-Spy! Nvidia has held-off on the 7XX news as long as they could and have worked down inventories without letting prices tumble. AMD has been holding something extremely "tight-fisted" ever since that "no new card revelation" back in Feb. I take all rumors with “pinch of salt” even saying in my first post on the topic…
by: Casecutter
If this is the course AMD is on...
Sure it's released (by whoever) to rain on Nvidia that we know… AMD will still sell the cards they've figured based on them releasing this news, or debunk it if it not truly going to help their sales. Even if totally false they wouldn't necessarily care unless it hurts their sales figures. I think Nvidia has to take a hard look at this decide, is it true, then either say take sales on lower profit, or hope it doesn’t go optimistically, could goes south for AMD, or not true... and not show any or much change in current price points.

If AMD did really "out'd-this" info this early they played it wrong! I would've waited the month when Nvidia started releasing card like the GTX780/GTX770 for review and had set MSRP's, then fired off with "technical white paper" on what they will have, that would've been the right timing.

Heck for all we know Nvidia out'd this information and some parts of it carry some truth. Better for them to get it out now than letting AMD pull it out and rain on their release. We just never know who behind the curtain and what the string actually mean.
Posted on Reply
#25
Casecutter
by: esrever
CI isn't gone, its just going to be delayed a bit a merged with existing products. VI will be more than a year away from the looks of it.
While I might say some re-spins and/or re-badging of Southern Island parts is probably still viable into 2014. Although, I'm not sure why you're saying VI is "more than a year away" the article is projecting, speculation, alleging for the 4th Qtr of this year. I’m not saying your opinion can’t be valid still, but working from this AMD may have adjusted it's evolution up.
Posted on Reply
Add your own comment