ATI Radeon HD 4670 and HD 4650 launched by AMD

Wednesday 10th September 2008, 05:21:00 PM, written by Rys

AMD have launched a couple more Radeon HD SKUs today, using a new chip called RV730.  A relative of RV770, the chip packs the same number of shading ALUs as RV670, in a die about 75% of the size.

Although AMD have increased area efficiency in their RV7 GPUs, there are a couple of minor areas where RV730 doesn't quite match up to RV670, which complicates figuring out how much more efficient it became.

For starters, there are half the ROP units, and therefore a half-width memory bus compared to RV670, the chip that restored Radeon fortunes before RV770 delivered its hammer blow.  The design team saw fit to double the number of samplers, though, although like RV770 there's not quite the interpolation rate to feed each of the 32 units a unique texcoord per clock.

Core clock for the best RV730 SKUs is a healthy 750MHz, giving you just shy of half a teraflop of FP32 compute per second, and while the ROP count means it's a little anaemic in terms of pure pixel rates, the resolution target and price mean you're getting good value for money.

For 70 EUR, you're getting the top Radeon HD 4670 SKU.  Graphics transistors, and there's 514M of them in RV730, have rarely been so cheap.  4670 pairs the chip with 512MiB of speedy GDDR3, and the product seems set to match the outgoing RV670-powered HD 3850 products in terms of rough performance.

AMD have tweaked things here and there when it comes to making the best use of the chip's internal memory pools, but in general it's just a nicely cut down RV770.  Crossfire is supported, you get two 10-bit display pipes, UVD, DisplayPort support and all the other display mod cons you'd expect from a graphics product in the latter half of 2008.

NVIDIA have seen fit to release some new products to swing a counter punch, too.  9500 GT seems like a bit of a weak offering to the consumer in retaliation, and it's likely that NVIDIA will use G92 at the 4670 price point at some point soon.  That won't help anyone's margins, but it keeps competition healthy.

Indeed, 9600 GSO is only 20 EUR more costly today, and is powered by that very chip.  Not a bad price for what was recently a resolutely high-end piece of silicon, and it's what you'll find in 9800 GTX and friends higher up the stack.

In short, RV730 seems competitive where AMD would like it to be, and while it's not going to set any pixel pushing records, back of a napkin scribblings here at B3D HQ to work out price/perf/area show it doing great.  AMD will sell as many as they can convince TSMC to make, in HD 4670 and HD 4650 guise.  Beyond3D readers are well advised to not bother saving the 10 bucks for 4650, though.  Get 4670 if < $99 is your price range.

Our friends at the Tech Report aren't quite there yet with their analysis, but Damien has something ready in his native tongue, and his graphs and analysis are always worth a read, even if you don't parlez Francais.

RV710 debuts shortly.

Discuss on the forums

Tagging

ati ± rv730, radeon, hd, 4670, 4650, area, efficient


Latest Thread Comments (92 total)
Posted by Lukfi on Saturday, 13-Sep-08 18:22:53 UTC
Quoting Mintmaster
The problem was that the ROPs, ALUs, and TMUs were so damn big that they couldn't fit many on the chip.
Perhaps they originally thought they'd manufacture R600 at 65nm?
Quote
I'd be surprised if they didn't have enough time.
They did not realize there were those bugs in the design until they got the first chips back from manufacturing, perhaps.

Posted by ShaidarHaran on Saturday, 13-Sep-08 19:26:35 UTC
Quoting Mintmaster
I still think that when you're given an engineering task that it should be relatively optimal, especially for ASICs.
Depends on what you're optimizing for. It's not fair to say R600 wasn't fast enough without considering its design goals.

That being said, R600 did not perform as a 512-bit enthusiast part would be expected to, IMHO.

Posted by Dave Baumann on Saturday, 13-Sep-08 19:44:38 UTC
Quoting Mintmaster
I still think that when you're given an engineering task that it should be relatively optimal, especially for ASICs.
Quote
It really said something to me when I saw slides saying perf/mm2 was a design goal with RV770. Duh! WTF were the hw architects smoking when this wasn't a design goal for R600?
Of course. But its also to make it as optimal as you can in the schedule. RV770 is not magically different, it just had a lot more time to make it optimal.

Posted by no-X on Saturday, 13-Sep-08 20:48:23 UTC
Mintmaster: Are you sure, that failure of R600 originate from bad design decisions and not from design bugs? You have mentioned R200 and I was told, that R600 was the most bugged product since the R200...

Posted by Lukfi on Saturday, 13-Sep-08 21:52:46 UTC
=>*no-X*: Do you mean the major bugs or the minor ones that are always worked around somehow (but the may slow the thing down a little bit?) Nevertheless I think that the design decisions contributed to the failure more than bugs.

Posted by Wesker on Sunday, 14-Sep-08 02:44:18 UTC
The argument of "not having enough time" doesn't bode too well, considering that Nvidia pushed out G80 (which I may remind you was a 90nm chip) well before R600. Yes, G80 wasn't exactly a smooth ride either (from a software point of view), but at least it delivered record breaking performance.The question on my mind is; why couldn't the R600 design team make something akin to RV730, but at 80nm? Obviously at 80nm the chip would be bigger, and you could also arm the GPU with a wider bus (RV730 is clearly bandwidth starved).

Posted by fellix on Sunday, 14-Sep-08 07:48:57 UTC
AFAIK, Nvidia spent four years in R&D for G80, anyway. At the same time, ATi had other extra predicaments to deal with, outside the pure engineering cycle.
Quote
why couldn't the R600 design team make something akin to RV730, but at 80nm?
Well, R600 at its present form was already stretching the limits of the available 80nm tech. And, as someone previously reminded, the R600's TMUs were much fatter than ones in R700 -- a consequence of misguided design decisions at the time, so putting more was off limits, somewhat. Major architecture redesign was needed, for that, as we see it now.

Posted by CarstenS on Sunday, 14-Sep-08 11:24:44 UTC
Is it just me or is there a kind of queer connonation with the fact, that all they really had to do from ground up were the ROPs and those were a major player in R600 performing sub-expectaions?

I mean, at launch Ati went on about R6xx being their second generation 'USA' with which they would profit from their experience in Xenos. But the Xenos GPU itself did not have --- ROPs (call it RBEs if you like). Those were residing in a separate Die connected to Xenos with a 32 GB/sec link IIRC. The ROPs themselves had EDRAM access at a rate of 128 GB/sec (which, incidentally, was just the same memory bandwidth, the GDDR4-version of R600 would sport).

Plus, Atis engineering team(s) seemed quite a bit stretched with so many products to design for in rapid succession while (AFAIK) being sigificantly smaller than Nvidias. The latter had the luxury of doing almost no work since G70 except from shrinking exisiting designs and did not have to design a revolutionary console GPU to boot.

Looking back, my impression that R600 was an unfortunate attempt at glueing together some architectural parts that weren't able to fit properly at the time togehter with some decisions that are rather questionable (Full-FP16-everyhting - even all textures were apparently promoted to FP16), did not change.

Reverting some of their decisions in RV670 (FP16 everywhere) and then some more in RV7x0 (ringbus, traditional ROPs) really made the underlying architecture shine recently, though.

--
On another thread of thoughts:
The whole FP16-Texturing-thing boggles my mind. I mean, they designed Xenos themselves, optimizing it AFAIK for RGBA1010102 (as also support in R5x first) and must have known that this format would be used far and wide by every Xbox-360-port ever coming to the PC. So why was it even possible that they so firmly believed in FP16?

Posted by no-X on Sunday, 14-Sep-08 12:32:02 UTC
It could be stupid idea, but nVidia supported HW FP16 filtering/blending since the NV40. It's known, that nVidia have significant influence on developers. Maybe ATi assumed that due to nVidias influence, FP16 filtering will be common thing in 2007?

Posted by fellix on Sunday, 14-Sep-08 13:38:17 UTC
OK, here are some quotes from SirEric, over the matter:
Quote
The samplers were designed to be 64b samplers from nearly the beginning, and matching that to BW and keeping the 4:1 ratio on ALU:Tex was the design choice made. In the latest games, where ALU:TEX ratios hit 15~20, this really shines.
Quote
I remember when we decided to focus on FP16 as "the basic" unit for most things on the R600. At the time, HDR rendering was just starting, but it certainly seemed the way of the future. Not only for floating point €œimage€ textures, but also for all types of floating point textures, such as normal maps and other more complex floating point data sets. In retrospect, the industry is probably moving a little slower than we expected, though I would say that games like Oblivion in 2006, and a few of the upcoming titles are really changing the tides, at least at the avant-garde of games. I'd like to think that our chips are very forward looking.


Add your comment in the forums

Related ati News

ATI Radeon HD 4890 launched at $250 with improved GPU
ATI Mobility Radeon HD 4860 and 4830 appear on 40nm
ATI Catalyst 8.10 driver released
ATI Radeon HD 4870 X2 at Rage3D
Life w/Playstation Released via Folding@Home Update
CATALYST 8.9 shows up with early OpenGL 3.0 support
RV730 and RV710 board designs pictured at HEXUS
Catalyst 8.8 Hits the Street
ATI Radeon HD 4870 X2 released and reviewed
AMD 790GX review at The Tech Report