[w0mbat]

What do you expect from the upcoming ATI GPU series? How do you think ATI will try to get the best out of their next (R(V)800?) GPU series? Will they use 45nm or even 40nm process? Will there be a RV7xx refresh or is the next step a complete new desing?

Just post your thoughts here


NordicHardware just postet that they expect the first 40nm GPUs in Q1 2009 which will be RV740 and RV870.
http://www.nordichardware.com/news,7946.html


My thoughts:
RV870 will be a RV770 refresh @ 40nm
~1.6 billion transistors
25 SIMDs
400 5D ALUs (2000SPs)
100 TMUs
16-24 ROPs (8z/clk)
256bit MC @ 5Gbps GDDR5
600-800MHz engineclk

Or less ALUs but a seperated shader clk.

[fellix]

Duh!
And they said "no more big chips"?
/any time-frame about 450mm wafers being scheduled for mass production?/

Anyway, I don't think a 40nm design would be a pure shrink, as 260 sq.mm mark is quite comfortable, even now. Maybe beefing TMUs (I want my single-cycle FP16 back), enlarge some shared-mem buffers (16K>32/64K) and most importantly - yet another clock rate bump at 850~900MHz mark for a reference.
Probably, the economy of scale will be again the main agenda, here.

[Lux_]

Didn't the rumour of 2000-shader core first got posted in this board as a joke ?

[CarstenS]

Originally Posted by Lux_

Didn't the rumour of 2000-shader core first got posted in this board as a joke ?

Indeed I think it was: http://forum.beyond3d.com/showthread...42#post1173642

[Jawed]

Originally Posted by CarstenS

Indeed I think it was: http://forum.beyond3d.com/showthread...42#post1173642

Funny thing is, that was posted at a time when ATI was expected (by me at least) to continue increasing ALU:TEX. RV770 seems to mark the start of a new era, where this ratio holds at 4:1. So it's looking pretty unlikely there'll be 2000 ALU lanes on a single ATI GPU any time soon. Kinda looks like we'll have to wait for 32nm...

Jawed

[ShaidarHaran]

Originally Posted by Jawed

Funny thing is, that was posted at a time when ATI was expected (by me at least) to continue increasing ALU:TEX. RV770 seems to mark the start of a new era, where this ratio holds at 4:1. So it's looking pretty unlikely there'll be 2000 ALU lanes on a single ATI GPU any time soon. Kinda looks like we'll have to wait for 32nm...

Jawed

Hold on now, one generation of products isn't enough to pronounce a trend shift. I don't believe so, anyway. ATi has long held to the notion that compute power should increase with successive generations relative to texture filtering/sampling abilities. Why change now?

I believe R7xx is a "correction" to the mistake that was R6xx and it's horrible lack of texturing/z-fill/and AA sample rates. I'm sure you'd agree with me on this. Now that these mistakes have been corrected, there's no need to do so again. Thus, ATi can return to their preferred design philosophy with the R8xx generation of products if they are in a position to do so (and with the shrink to 40nm I can see no reason why they wouldn't).

[3dilettante]

Originally Posted by fellix

Duh!
And they said "no more big chips"?
/any time-frame about 450mm wafers being scheduled for mass production?/

Best-case, in the eyes of Intel, TSMC, Toshiba, and Samsung, was something in the region of 2012.

In the eyes of the equipment and tool manufacturers, the later the better.
The 300mm wafer transition turned out to be not such a good thing for them.

The likely costs and extremely reduced market size mean that unless the big chip manufacturers start paying some serious cash to finance the effort, it won't happen for quite some time longer, though I don't know enough about the dynamics of the equipment industry to say how many years more.



As for RV870, the rumor said that R870 would have 2000 ALUs, which means RV870 would have 1000.
Given that RV770 already has 800, that's a relatively modest increase despite the jump from 55nm to 40nm.

This might make sense, if the power improvements lag as far behind the density improvements as was reported earlier.

I'm curious as to what's going on at the IHVs.
Did AMD sort of eat into its own future by the more significant redesign of RV770 compared to how much GT200 hewed to G92, or is it that Nvidia is further away from the desired design target for late 2009/early 2010?

[Tim]

Originally Posted by fellix

Duh!
And they said "no more big chips"?
/any time-frame about 450mm wafers being scheduled for mass production?/

Anyway, I don't think a 40nm design would be a pure shrink, as 260 sq.mm mark is quite comfortable, even now. Maybe beefing TMUs (I want my single-cycle FP16 back), enlarge some shared-mem buffers (16K>32/64K) and most importantly - yet another clock rate bump at 850~900MHz mark for a reference.
Probably, the economy of scale will be again the main agenda, here.

1.6 bilion transistors on the tmsc 45nm process would be about the same size as the RV770. I don't think they can make a 2000SPs and 100 TMUs chip with that "few" transistors. My bet would be in the range of 1280-1600SPs and 64-80TMUs, still with 16ROPs and 4z per clock and a clock speed at about 900MHz.

[Jawed]

Originally Posted by Tim

1.6 bilion transistors on the tmsc 45nm process would be about the same size as the RV770. I don't think they can make a 2000SPs and 100 TMUs chip with that "few" transistors. My bet would be in the range of 1280-1600SPs and 64-80TMUs, still with 16ROPs and 4z per clock and a clock speed at about 900MHz.

Looking at the RV770 die shot, the physical I/O stuff along the edges of the die amounts to 24% of the entire die. Presumably this stuff would all end up the same size at 40nm.

So 76% of RV770 is graphics logic. 40% of RV770 is taken by the clusters (ALUs+TUs). So, 36% of RV770 is non-cluster logic. At 40nm that logic could be unchanged in capability (i.e. 16 RBEs, 4x MCs, 1 hub), but presumably would scale.

Scaling from 55nm to 40nm is supposed to be unusual in some respect - I can't remember if the scaling is considerably better or considerably worse than a simple areal evaluation would imply.

Jawed

[PeterAce]

Originally Posted by Jawed

Scaling from 55nm to 40nm is supposed to be unusual in some respect - I can't remember if the scaling is considerably better or considerably worse than a simple areal evaluation would imply.

Jawed

Front Page! :

http://www.beyond3d.com/content/news/608

http://www.beyond3d.com/content/news/568

* Posted to save Arun's teeth from grinding....

[Jawed]

So, about 16% smaller going from 55nm to 40nm?

So a "<=20% bigger" refresh of RV770 seems pretty likely then.

If just RV770's clusters are shrunk to 84%, then naively there's room for 19% more of them 960 ALUs

I'm doubtful a refresh would make any real changes to the MCs, RBEs, L2s, so they would shrink too. In that case that leaves room for the clusters to grow by 36% while retaining a die size of 256mm2.

If, historically, a 256-bit GPU could be as small as ~190mm2, it seems that at about 256mm2 ATI is paying quite a high price for the combination of GDDR5 and CrossFireX Sideport.

Is this the approximate minimum size for all RVx70 GPUs for a few years to come? If so, isn't this GPU going to get progressively more and more expensive with each new node (presuming that each new node has worse yields per mm2)?

Jawed

[fellix]

At 40/32nm nodes, if ATi decides to recycle the RV770 design from a "compact" PoV, is it viable to consider adding an eDRAM array to the core--and simplifying the local memory buffer req's? I mean... with all the GDDR5 expected performance bumps and (don't bash me, here) the possibility of adopting XDR2+ most probably they should sound the "go-ahead" horn, for the gazillions of SPs!

[jamis]

Originally Posted by w0mbat

My thoughts:
RV870 will be a RV770 refresh @ 40nm
.

R8x-series will be DX11

[ShaidarHaran]

Originally Posted by jamis

R8x-series will be DX11

Highly doubtful. R8xx will be out next year. It would be miraculous if DX11 made an appearance prior to 2010.

[jamis]

Originally Posted by ShaidarHaran

Highly doubtful. R8xx will be out next year. It would be miraculous if DX11 made an appearance prior to 2010.

It will make appearance next week, dunno when it will be released though. R800 has always been slated for DX11, the next NV part (g300 or whatever) will propably be DX11 too, the rumours are allready heading that direction.

[ShaidarHaran]

Originally Posted by Dave Baumann

When the ALU/Tex ratios stay the same how is there a "correction" there?

Originally Posted by fellix

The issue here is not the ratio, but the raw capacity of those aspects.

Exactly.

Originally Posted by jamis

It will make appearance next week, dunno when it will be released though. R800 has always been slated for DX11, the next NV part (g300 or whatever) will propably be DX11 too, the rumours are allready heading that direction.

I'm not saying you're wrong, but this is the first I've heard of that. Do you have any links to back this up?

[Wirmish]

SP & TMUs:
RV670 -> 4 SIMD cores, each with 16 SPUs (4D+1 = 80 ALUs) and 4 TMUs -> 320 ALUs + 16 TMUs
RV770 -> 10 SIMD cores, each with 16 SPUs (4D+1 = 80 ALUs) and 4 TMUs -> 800 ALUs + 40 TMUs (+6 SIMD cores vs RV670)
RV870 -> 16 SIMD cores, each with 16 SPUs (4D+1 = 80 ALUs) and 4 TMUs -> 1280 ALUs + 64 TMUs (+6 SIMD cores vs RV770)

ROPs, MC & BUS:
RV670 -> 4 RBEs, each with 4 ROPs x 2z/clk (16 ROPs), 4 MC -> 256-bit bus (72GB/s, 512MB/1GB GDDR4 @ 4.0GHz)
RV770 -> 4 RBEs, each with 4 ROPs x 4z/clk (16 ROPs), 4 MC -> 256-bit bus (115GB/s, 512MB/1GB/2GB GDDR5 @ 3.6GHz)
RV870 -> 4 RBEs, each with 4 ROPs x 4z/clk (16 ROPs), 4 MC -> 256-bit bus (128GB/s, 512MB/1GB/2GB GDDR5 @ 4.0GHz)
RV870 -> 6 RBEs, each with 4 ROPs x 4z/clk (24 ROPs), 6 MC -> 384-bit bus (192GB/s, 768MB/1.5GB GDDR5 @ 4.0GHz)

SIZE:
RV670 -> 192mm²
RV770 -> 260mm²
RV870 -> ~260mm² (~280mm²)

[fellix]

Erroneous info!

>>del post!

[fellix]

The issue here is not the ratio, but the raw capacity of those aspects.

[Dave Baumann]

So, by that logic R420 was a correction on R300

You are saying that prior generations must have been "bad" because they don't offer as much raw performance!

[ShaidarHaran]

Originally Posted by Dave Baumann

So, by that logic R420 was a correction on R300

You are saying that prior generations must have been "bad" because they don't offer as much raw performance!

I think you're over-simplying this, Wavey. R420 did address R300's relative lack of shading compute power, so in a sense it was a "correction". A better example of another correction would be R520->R580 addressing the same lack of shading power. Given the fact that ATi has addressed a shortcoming in shading compute power twice, I can understand why they shifted to a high ALU:TEX philosophy. Unfortunately, having been burned twice they over-corrected, which caused the relative lack of texturing ability in R6xx. R7xx is a correction of the failed R6xx design philosophy. I know you can't admit this as a representative of ATi, but there's no use denying it either.

RV770 is a fantastic GPU, no matter how you slice it.

[no-X]

ShaidarHaran: I can't accept your opinion. RV770 boosted texturing and aritmetic power equally. You can't say that RV770 adressed lack of texturing power, because RV770 boosted nuber of ALUs in the same way and if R600 wasn't lacking of anything, pure aritmetic rate was this thing.

In relative way, RV770 is weaker in texturing than R600, because ALU:TEX remained and TFUs were emasculated.

[Dave Baumann]

Originally Posted by 3dilettante

RV770 has changes that could be construed as a significant rethinking of some of R600's big design decisions, such as the abandonment of the ring bus, changing how the TMUs relate to the ALUs, and a reworking of the cache hierarchy.

The TMU's relation to the ALU's was changed so that we can keep the same ALU/Tex ratio and batch sizes while increasing the entire texture and shader engine. With R600's design increasing the ALU size would mean either adding more SIMD's, but that would retain the same number of textures (thus the ratio would bias more to ALU's), or adding more ALU's per SIMD (and similarly adding more textures to the texture array), but this would result in larger batch sizes carrying other penalties.

R600 was OK in this respect for its generation because it did allow configurable, 2D scaling of numbers of units for the rest of the family, but these all reduced the number of units, not increased. To increase the units the relationship between texture engines and SIMD's had to be changed to allow the same ratios and the same batch sizes.

Caching hierarchy and memory went hand in hand with one another. We put a lot of work into texture cache modeling and changed to a fully tiled memory system that, for the primary task of 3D operation, alleviated the need for the memory channels to be passing data between one another.

[ShaidarHaran]

Originally Posted by no-X

ShaidarHaran: I can't accept your opinion. RV770 boosted texturing and aritmetic power equally. You can't say that RV770 adressed lack of texturing power, because RV770 boosted nuber of ALUs in the same way and if R600 wasn't lacking of anything, pure aritmetic rate was this thing.

In relative way, RV770 is weaker in texturing than R600, because ALU:TEX remained and TFUs were emasculated.

Again, this has nothing to do with the TEX:ALU ratio itself being out-of-balance. Base texturing capability needed to be increased over the previous generation because it was deficient given the existing shading capability. There's no denying this, benchmarks prove my point.

[no-X]

So RV770's texturing power must be even more deficient given it's shading capability. OK?