R520 launch delayed?

[Hellbinder]

Nope.. the INQ is wrong.

The R520 has always been a second half 2005 product. They have been assuming it was a Spring product and have been wrong the entire time. ATI has never clearly stated when the product was being released. I personally have *suspected* for lack of a better term that they were targeting a September release for several months now.

Secondly the Whole SLI thing is stupid. That has nothing to do with it whatsoever. The Price of 2 Cards is so much higher than the Price of One Complete card that a Few FPS here and there wont ammount to a hill of beans. Thats not even considering the superior features that the 520 has over the current crop of Nvidia hardware.

 

[R300King!]

Well, lets spill that hill of beans Hellbinder!

What definates do you know about the R520?

 

[trinibwoy]

I think the real bind, to borrow your word, is that ATI needs an SM 3.0 card now (yesterday)

I disagree, I still don't see a big need/use for SM 3.0 yet.

True but tell that to all those folks answering "get the GT, it has SM3.0 for free!" in a lot of "Help! 6800GT vs X800XL" threads. Apparently most people are in the market for AGP still and here's where SM3.0 support is beating ATI when it comes to that particular debate. SM3.0 on its own isn't pushing much but when you're getting the same SM2.0 performance and SM3.0 as a freebie that's when it becomes a deciding factor.

Thats not even considering the superior features that the 520 has over the current crop of Nvidia hardware.

You're not gonna just drop that on us and not tell us what these superior features are.....are you?

 

[DemoCoder]

He doesn't know. He's been caught lying before in the forums.

 

[overclocked_enthusiasm]

Add these as possible reasons:

1. x800 XL is selling well and margins are better than R520 will be. ATI needs an EPS surprise that only gross margins can acheive at this time.
2. They want a few months to build up inventory so they can actually launch with product in the channel.
3. They want AMR to have a few months to get established so that when R520 does launch the demand will be that much more. In the mean time, they can sell more R4xx cards.
4. Inventory levels are just too high with R4xx cards and they want to deplete this inventory before launching next generation.
5. They will have to lower the ASP on the R4xx cards when R520 is launched to avoid a surplus inventory of R4xx. This will hurt gross margins and based on the last 2 quarter's ATI reported they are in NO position to do so.
6. They want to ride the Xbox 360 wave with no distractions until the fall.

x800 Xl is too new, selling too well and gross margins are too fragile to introduce a new generation right now. AMR will buy them several months at the least to get their financial house in order, get some positive PR and try to get inventory established for a successful R520 launch with cards in the channel. The risk is obvious...Nvidia will be seen as the leader AGAIn if they release G70 before R520. ATi will be seen as playing catch up again.

 

[Tahir2]

I think everyone assumed this part was coming out in June/July, perhaps it has been pushed back a little further. Who knows...yawns.

 

[Geo]

Add these as possible reasons:

1. x800 XL is selling well and margins are better than R520 will be. ATI needs an EPS surprise that only gross margins can acheive at this time.

I have been considering the bump up to 512mb on the xl as proof positive that it still has legs in ati's product range. Said another way --they were planning to keep it around anyway. Probably until earlyish next year as the midrange on the R520 line gets out in volume.

The ati road map we all poured over does not support a September release for high-end R520. July, maybe, but not September. If we're moving to September, that's a change and no question --the question is why, not whether.

Assuming the inq is accurate, and as someone pointed out upstream, 50%ish is their accuracy range --but I suspect that breaking a significant product delay story on a high-profile part gets held to a little higher scrutiny over at the inq than, say, some product feature pieces. Certainly I'd hold such a story to a higher standard if I was editor.

 

[kemosabe]

Another source referring to this as a delay, so obviously there was widespread expectation of a late spring/early summer release.

 

[digitalwanderer]

Another source referring to this as a delay, so obviously there was widespread expectation of a late spring/early summer release.

THG doesn't really qualify as a reliable source though, they could easily be basing their blurb on the Inq article.

 

[overclocked_enthusiasm]

I have been considering the bump up to 512mb on the xl as proof positive that it still has legs in ati's product range. Said another way --they were planning to keep it around anyway. Probably until earlyish next year as the midrange on the R520 line gets out in volume.

What is going to be increasingly difficult for ATI is the defense of the "no SM 3.0" with x800 xl. Against the 6600GT the x800 xl rocks. Against a midrange G70 the x800 xl will be dead meat...just like the 9600 was versus the 6600 GT. I see ATI only able to ride this SKU until a midrange G70 is launched.

To further complicate things, ATI will obviously want to hype SM 3.0+ on R520 and that might make the x800 family (xl included) the ugly red haired step child unintentionally.

 

[Geo]

However, R520 - reasonably big chip, change of architecture, new process, mmmmmm, may be a little time needed to understand all the variables?

My success at interpreting Wavey-isms is somewhat less than the Inq's, but that strikes me as "man not surprised", and possibly even "man with a little visibility".

Remember that assuming we are right on the 16-pipes bit, R520 was never going to be a speed-demon in non-shader-limited apps anyway, at least by, say, X800 vs 9700 standards. So tweaking it and the driver settings for the max it has to give was probably a necessity at launch to avoid disappointed reviews, rather than doing a "aw, hell, we'll give it a nice driver bump 3 months from now". Particularly if the optimizing is required (as it probably is) in the new bits --it is hard to make a marketing case for why you went away from "more pipes, More Pipes, MORE PIPES" Conventional Wisdom if the stuff you went for instead isn't looking at least a little sexy in initial benchmarks.

 

[Geo]

It does suddenly occur to me to wonder if Xbox plays into this. Isn't Beta 2, which is supposed to be very close to the final product, due to ship in a month or so? It wouldn't be too far a stretch to wonder if as they got visibility that they needed more time/resources to optimize R520 that they also realized that signficant members of the "A Team" to do it wouldn't be available to help until summer. . .

 

[thatdude90210]

that they needed more time/resources to optimize R520 that they also realized that signficant members of the "A Team" to do it wouldn't be available to help until summer. . .

What does Mr T have to do with this.

 

[caboosemoose]

Perhaps R520 is being dumped in favour of R580...

 

[trinibwoy]

Another source referring to this as a delay, so obviously there was widespread expectation of a late spring/early summer release.

THG doesn't really qualify as a reliable source though, they could easily be basing their blurb on the Inq article.

I don't see how the info in the THG article could be inferred from the Inq's. Either way all of this won't amount to a hill o' beans. Soon enough we'll have R520, AMR and all the goodies and soon after that it won't be good enough and we'll want more. A fall release is fine with me - at least my GT won't be outdated as fast.

 

[ANova]

I think the real bind, to borrow your word, is that ATI needs an SM 3.0 card now (yesterday)

I disagree, I still don't see a big need/use for SM 3.0 yet.

True but tell that to all those folks answering "get the GT, it has SM3.0 for free!" in a lot of "Help! 6800GT vs X800XL" threads. Apparently most people are in the market for AGP still and here's where SM3.0 support is beating ATI when it comes to that particular debate. SM3.0 on its own isn't pushing much but when you're getting the same SM2.0 performance and SM3.0 as a freebie that's when it becomes a deciding factor.

This is not the case however. SM3 isn't a freebie, as in most cases the 6800 GT is still at least $50 more then the X800 XL. So, I guess most people have to decide whether or not SM3 is worth an extra $50 or more.

 

[BRiT]

Perhaps R520 is being dumped in favour of R580...

You took the words right out of my mouth...

 

[trinibwoy]

Well we've been over this a bazillion times. I'm not sure which cases you're referring to but the AGP GT is currently cheaper than the AGP XL and most people are still buying AGP parts. For those going with PCIe the obvious choice is the XL but this is no where near the majority.

 

[Pete]

Well, Orton still promised the financial analysts a doubling of revenue in the July-Aug quarter.

Maybe he was expecting revenue from Xbox 360 ramping production?

This is not the case however. SM3 isn't a freebie, as in most cases the 6800 GT is still at least $50 more then the X800 XL.

The PCIe 6600GT costs about as much as the PCIe X700P, and the AGP 6800GT costs as much as the PCIe X800XL. So, the native parts cost the same, at least to the consumer. Bridge chips add cost, but it doesn't look like SM3 is costing nV or their AIBs too much extra.

He doesn't know. He's been caught lying before in the forums.

Surely you can come up with a better example than an April Fools thread?

 

[karlotta]

Well we've been over this a bazillion times. I'm not sure which cases you're referring to but the AGP GT is currently cheaper than the AGP XL and most people are still buying AGP parts. For those going with PCIe the obvious choice is the XL but this is no where near the majority.

Well at Bestbuy a XL is 50$ cheeper..... AGP to AGP and the PCIe XL is 299$