AMD Southern Islands GPU Codenames Surface

[jamsbong]

Originally Posted by a_ump

where did that chart appear from....and why didn't anyone else comment on the one post that actually pertains perfectly to the OP.

http://wccftech.com/amd-hd7000-south...yst-117/27899/

Based on the info above. It seems that VLIW4 only survived one generation. Now GCN (graphic core next) is the new core design. Considering VLIW5 survived 4 generations (2xxx to 5xxx), ATI is really on the move again.

Seems to me, the ATI team in AMD are the creative ones and the ones bringing on the profits. Their CPU team is very slow in pumping out new stuff. Thats why the focus appears to be more GPU orientated.

I really look forward to see an APU and HD7970 working as one. Things like direct compute or OpenCL can be done on the APU with great efficiency (even more efficient that discrete GPU). Then let the powerful HD7970 do the graphics to produce blazing framerate.

fingers cross that AMD and ATI can pull this off.

[theoneandonlymrk]

Originally Posted by WarraWarra

Any idea if there would be some form of performance increase maybe 8% or more seeing it is a new chip revision of the old 5XXX series ?

not even a revision of the 6xxx series never mind the 5xxx its a new core new shader architecture

[jamsbong]

I have doubts about increase in theoretical performance. But I think practical performance should increase. I expect the new design to have ALU's working more often because it has been well fed with data and have less waiting for other calculation to be finish before executing new ones.

I hope they are able to squeeze in as many ALU as the famous 5870 (1600 cores) and run it at 1+ghz.

[swaaye]

It's pretty likely that focusing more on GPGPU will cause a loss on graphics performance per transistor because the two aren't really directly related. See huge Fermi compared to smaller Cayman. This new AMD GPU architecture is more about going after the NV Tesla market (Intel is after it too with that Knights Corner beast).

I suspect that the chip will be large and use 28nm well but that its performance isn't going to be that big of a leap. But then we haven't really had big leaps in many years.

I find it strange that they moved to VLIW4 at all with this new design coming to replace it.

[theoneandonlymrk]

Originally Posted by swaaye

It's pretty likely that focusing more on GPGPU will cause a loss on graphics performance per transistor because the two aren't really directly related. See huge Fermi compared to smaller Cayman. This new AMD GPU architecture is more about going after the NV Tesla market (Intel is after it too with that Knights Corner beast).

I suspect that the chip will be large and use 28nm well but that its performance isn't going to be that big of a leap. But then we haven't really had big leaps in many years.

i dont see it quite so negatively, this be their once talked of hectadonacles (hundred headed serpent) itll have less shaders then 5870 for deff but they will be twice as useable and if scaled to 28nm they may be able to push the shader/gpu clocks beyond what is poss now by default v excitedbout this

also i saws a doubling in perfprmance going from xfired 3870's to my 1x 5870

and 1x4870 equalled my 2x3870's so why your so negative is beyond me in reallity the 6xxx series was i bit weak an upgrade to the 5xxx series (same core similar shaders) imho and hence no card doubled the 5870's power excepct dual gpu

Originally Posted by swaaye

I find it strange that they moved to VLIW4 at all with this new design coming to replace it.

agreed afaik it was down to die space on 40nm and prob had some relevance to the lano APU design as 400 4xshaders would take up less space in a APU then 400 5x shaders and what with the stars cores taking more room then bulldozer imho it makes some sense as its ALL about size and die area

[[H]@RD5TUFF]

So the 7 series is the 6 series that perhaps requires less power.

[Salsoolo]

*starts saving money *

[swaaye]

Originally Posted by theoneandonlymrk

and 1x4870 equalled my 2x3870's so why your so negative is beyond me in reallity the 6xxx series was i bit weak an upgrade to the 5xxx series (same core similar shaders) imho and hence no card doubled the 5870's power excepct dual gpu

agreed afaik it was down to die space on 40nm and prob had some relevance to the lano APU design as 400 4xshaders would take up less space in a APU then 400 5x shaders and what with the stars cores taking more room then bulldozer imho it makes some sense as its ALL about size and die area

What I said was that GPGPU features do little for gaming performance but they do cost transistors that could have gone towards more game graphics hardware. This is why NVIDIA's chips have gotten so huge and yet don't blow AMD's away. GPGPU stuff we don't really have use for. They are going to blow a big chunk of their 28nm transistor budget on GPGPU and it's going to cost graphics potential. AMD and NV are going after that GPGPU market and we are becoming less of a priority.

Also, 5870 is more like only 30-50% faster than 4870. 3870 was a rather poor GPU and 4870 was a big improvement over it however. 3870 occasionally gets slapped around by even 8800GT for some reason. AMD's R600 and RV670 chips had problems with efficiency.

40nm had nothing to do with the switch to VLIW4. Cayman was originally to be 28nm but the delay to 28nm caused them to go back and set it up for 40nm. The chip is the same thing as it would have been on 28nm though. It takes them several years to design GPUs so they can't make major changes in a short period of time.

[xtremesv]

Don't get me wrong, I really love AMD and Nvidia are still in the game but where are the killer games? I feel PC games industry is stuck somewhere (maybe between obsolete consoles and piracy), comes to my mind Crysis 2, I mean the game supports DX11 through a PATCH!!!! that came out 3 months after game release!!!!

[Horrux]

Originally Posted by xtremesv

Don't get me wrong, I really love AMD and Nvidia are still in the game but where are the killer games? I feel PC games industry is stuck somewhere (maybe between obsolete consoles and piracy), comes to my mind Crysis 2, I mean the game supports DX11 through a PATCH!!!! that came out 3 months after game release!!!!

This is true, we have some very good games, but we've always had really good games in the past before these newer architectures.

[theoneandonlymrk]

Originally Posted by swaaye

Also, 5870 is more like only 30-50% faster than 4870. 3870 was a rather poor GPU and 4870 was a big improvement over it however. 3870 occasionally gets slapped around by even 8800GT for some reason. AMD's R600 and RV670 chips had problems with efficiency.

40nm had nothing to do with the switch to VLIW4. Cayman was originally to be 28nm but the delay to 28nm caused them to go back and set it up for 40nm. The chip is the same thing as it would have been on 28nm though. It takes them several years to design GPUs so they can't make major changes in a short period of time.

5870 is faster by a bigger margin then that + is dx11 thats a big enough improvement right their, 4870 was twice as fast as a 3870 you and i both agree on that and i agree the 3870 wasnt great anyway.

the 7xxx series is what the 6xxx series would have been had 28nm worked, what they released whas a revision of the core from 5xxx series and some regular un risky shaders and i agree it takes years to dev gpu's but their process fell over late in the day and they backtracked fast simples they needed 28nm to fit southern islands plus the old shader type(what went in anyway) but didnt have the room for the new core on 40 nm given the shader count increase needed to counter nvidia

7xxx new not revised southern islands core and new shader array

all im saying is im optomistic AMD will get it right and imho they playing a wise game due to knights corner + nvdia's project denver APU(what they gona call that APU prob NPU)

and i dont think they will choose to harm graphics performance just to better gpu performance they will still need benchmark scores to sell their gfx cards after all

[swaaye]

Originally Posted by xtremesv

Don't get me wrong, I really love AMD and Nvidia are still in the game but where are the killer games? I feel PC games industry is stuck somewhere (maybe between obsolete consoles and piracy), comes to my mind Crysis 2, I mean the game supports DX11 through a PATCH!!!! that came out 3 months after game release!!!!

Ain't it the truth. I don't really know why I bothered buying a Radeon 6950 anymore. I think I just wanted to waste money lol. 8800GTX was running the console ports fine.

[rem82]

Originally Posted by Trackr

No there isn't.

AMD just isn't very good at making CPUs.

ΑΜD is the BEST to make new architectures for CPU !!

A) The AMD A64 Architecture was the first with IMC ,HYPER TRANSPORT and 64bit !
B) The K10 & K10,5 Architecture waw the first with IMC, NATIVE QUAD CORE and L3 CACHE !!
C) The FIRST REAL CHIP with CPU+IGP in ballance computing is Llano.
--------------------------------------------------------------------------------------------------
Intel imitated all these architectures many years later !!!!

Now AMD want to make the first real cp-gpu = TRINITY !!!

For more information, you can see here:
http://www.pcper.com/reviews/Graphic...PUs-and-Beyond
http://techreport.com/discussions.x/21141

[theoneandonlymrk]

Originally Posted by rem82

%u0391ΜD is the BEST to make new architectures for CPU !!

A) The AMD A64 Architecture was the first with IMC ,HYPER TRANSPORT and 64bit !
B) The K10 & K10,5 Architecture waw the first with IMC, NATIVE QUAD CORE and L3 CACHE !!
C) The FIRST REAL CHIP with CPU IGP in ballance computing is Llano.
--------------------------------------------------------------------------------------------------
Intel imitated all these architectures many years later !!!!

part true but your underselling intel their a bit and deff overselling AMD, they are both quite happy to copy each others ideas, amd have been reffering to bulldozer as a competant(moreso then intel) realisation of hyperthreading or was that a truly orig idea by AMD, i think not, but to be fair your right intel dont mind copying either or just using AMD licenced stuff(X64,A64 war 1 0 AMD)


i half thought i was in a BD thread there a min sorry

in these slides
http://wccftech.com/amd-slides-detai...tecture/21471/

it looks like their grouping simd units into fours and calling them compute units so i spose my top question is how MANY 320 512 or 640 are my clear faves making for a poss shader count total of 1280 2048 or 2560 but they also hint at multple primitive paths which imho seems like multiple schedulers etc ie more then one gpu (though probably cut down parts thereof) in a modular array making 64 cu per core x 4 cores my prediction making for 1024 shaders, itd make sense, though 8x cores with 64 CU units each makes more with chip binning in mind( ps4, xboxwoteva)
any news on when 7xxx series is due

[Damn_Smooth]

Originally Posted by theoneandonlymrk

any news on when 7xxx series is due

Rumors point to September. Hopefully they come true

[Aleksander]

Originally Posted by rem82

ΑΜD is the BEST to make new architectures for CPU !!

A) The AMD A64 Architecture was the first with IMC ,HYPER TRANSPORT and 64bit !
B) The K10 & K10,5 Architecture waw the first with IMC, NATIVE QUAD CORE and L3 CACHE !!
C) The FIRST REAL CHIP with CPU+IGP in ballance computing is Llano.
--------------------------------------------------------------------------------------------------
Intel imitated all these architectures many years later !!!!

Now AMD want to make the first real cp-gpu = TRINITY !!!

For more information, you can see here:
http://www.pcper.com/reviews/Graphic...PUs-and-Beyond
http://techreport.com/discussions.x/21141

If you know the truth, is not AMD who makes the architecture, are the people who make them. The more money in, the better architecture. After all, intel is the best for cpu right now. Even though i am an AMD fan. Ideas are relative, people change.

[theoneandonlymrk]

Originally Posted by Aleksander Dishnica

are the people who make them

yeh people working in/for AMD/Intel

FYI that dosnt even makes sense

its getting worse this thread, i cane get wot yo be is saying (.)(.)

[rem82]

Originally Posted by Aleksander Dishnica

After all, intel is the best for cpu right now.

From 2007 up to today, intel has ready the lines of production chip in 32nm and in 45nm the correct moment. For this reason Intel has better products per moment.

[Jonap_1st]

Originally Posted by rem82

ΑΜD is the BEST to make new architectures for CPU !!

A) The AMD A64 Architecture was the first with IMC ,HYPER TRANSPORT and 64bit !
B) The K10 & K10,5 Architecture waw the first with IMC, NATIVE QUAD CORE and L3 CACHE !!
C) The FIRST REAL CHIP with CPU+IGP in ballance computing is Llano.
--------------------------------------------------------------------------------------------------
Intel imitated all these architectures many years later !!!!

Now AMD want to make the first real cp-gpu = TRINITY !!!

For more information, you can see here:
http://www.pcper.com/reviews/Graphic...PUs-and-Beyond
http://techreport.com/discussions.x/21141

meh... new archictecture cant justified every point if it still slower than the competitors.

every company has their own fields, from my point of view AMD is still the best when focus on balancing performance between CPU and GPU, when intel is specialize focusing on CPU, and nvidia are good when focusing on GPU,,

eventhough i had build majority of my cpu's with AMD, i gotta to admit that the fastest CPU is still on Intel's hand..

[rem82]

Originally Posted by Jonap_1st

meh... new archictecture cant justified every point if it still slower than the competitors.

every company has their own fields, from my point of view AMD is still the best when focus on balancing performance between CPU and GPU, when intel is specialize focusing on CPU, and nvidia are good when focusing on GPU,,

eventhough i had build majority of my cpu's with AMD, i gotta to admit that the fastest CPU is still on Intel's hand..

that`s why ....

From 2007 up to today, intel has ready the lines of production chip in 32nm and in 45nm the correct moment. For this reason Intel has better products per moment.

[de.das.dude]

anyone else thinking AMD is really blooming everywhere?

[entropy13]

3DCenter.org with complete codenames list here.