Ati following in nvidia's footsteps ?

[dizietsma]

Although reports of ATi's demise have been greatly exagerated and,at the very least the r520 is going to be the match of G70, it does seem that what happend to nvidia in the past does seem to be hitting ATI this time around. What did Henry Ford say about history ? Maybe he was wrong.

Of course the nv3x was not massively successful due to a better competitor, something that is probably not the case this time around, but even so r520's late launch does bring back memories of this time in nvidia's history.

Then nvidia had the slower card, it's IQ was worse and it cost the same. How things have changed, now they have a faster card and it's IQ, with transparency supersampling, is still good enough to see on a Friday night after 6 beers . We're not talking about having your eyes 3 inches away, peering and getting a headache at a mouseover here, no, it's so good expect all TWIMTBP games to have at least 5 chainlink fences per level.

And they are shoving them out the door. On the day of release they had 200 cards at dabs.com at one point, now they have about 50 on order, so it must be selling well.

It's fast, it's got great IQ, its available in quantities. The nv30 had none of these, the nv40 had some and the nv, sorry, G70 has all. Maybe G stands for "gee!"

So Ati has some catching up to do on all fronts and maybe this catchup has been sidetracked by Xenos and Crossfire to some extent.

 

[trinibwoy]

expect all TWIMTBP games to have at least 5 chainlink fences per level.

Haha, that's funny. But probably true.

 

[dizietsma]

expect all TWIMTBP games to have at least 5 chainlink fences per level.

Haha, that's funny. But probably true.

Far from me to be the one to spread conspirency theories but the game with most chainlink fences is HL2. Now you know the real reason why it was delayed a year, not because the code was stolen, but Gabe got a nice nvidia backhander to start putting in all those fences

Digital Wanderer was right, nvidia are damn sneaky....

 

[wireframe]

What did Henry Ford say about history?

"You can interepret history in any way you like as long as it's black"?

But yeah...it's almost spooky how the tables seem to have turned. ATI seems to be doing so many strange things. First they say that multi-GPU solutions are pointless, then they make their own, then they paper launch it at a time when everyone is expecting their next gen cards. This, to me, says that not only have they got problems with execution, but their vision is faltering. If they had stuck to their guns about "no SLI" then I could at least respect them for that and had some expectations about how they would go about tackling the performance curve.

 

[Bouncing Zabaglione Bros.]

Let's wait to see what R520 looks like and when it arrives. There's no point comparing G70 to R420 and then proclaiming that Nvidia is way out in front. You should be comparing G70 to R520. Then we can tell who did the best job.

If R520 launches end of July with availabilty end of August (ie "late summer"), the two month head start for G70 will be negligable. All the people buying now rather than waiting a couple of months either can't wait because they need to build systems now, or would be buying Nvidia regardless of whether they launch a couple of months earlier or later.

 

[dizietsma]

Let's wait to see what R520 looks like and when it arrives. There's no point comparing G70 to R420 and then proclaiming that Nvidia is way out in front. You should be comparing G70 to R520. Then we can tell who did the best job.

Yes totally agree, I tried to make that point in the above part, that is why the analogy is not 100% because r520 is likely to be a lot better than nv30 was to r300 and therefore the halo effect of the top card should run down the line.

Also, if r520 goes AGP and nvidia do not then that will be an interesting twist and 180 degree turn in respect to r420 v nv40. I know people say AGP has had it's day but where you do not need bi-directional and where AGP still has enough bandwidth I do rather plump on just a 7800 or r520 AGP than a PCIe which means I do not have to get a PCie board and also a 24 pin PSU as well. When you have a very good 20 pin Antec 550w and MSI Neo2 Amd motherboard working fine that seems a waste and a hassle selling on....

That money saved on the PSU and motherboard would be better spent on a new monitor which does 1600x1200 native rather than 1280x0124 so I would not be as cpu limited.

 

[Richard]

Let's wait to see what R520 looks like and when it arrives. There's no point comparing G70 to R420 and then proclaiming that Nvidia is way out in front. You should be comparing G70 to R520. Then we can tell who did the best job.

If R520 launches end of July with availabilty end of August (ie "late summer"), the two month head start for G70 will be negligable.

Heh, you know, this is starting to remind me of the 9700 comparisons with the GF4 and how we should all wait for the NV30 to clear any doubt.

Btw, I agree with you, I'm just saying.

 

[Bouncing Zabaglione Bros.]

Heh, you know, this is starting to remind me of the 9700 comparisons with the GF4 and how we should all wait for the NV30 to clear any doubt.

Btw, I agree with you, I'm just saying.

Well the proviso is that R520 has to arrive in a timely manner. If people can't buy it some time in August, then ATI will be losing sales to people buying G70 because that's the best they can get when they want to upgrade. People with money burning in their pockets don't want to wait long when there is something good like G70 tempting them.

 

[Geo]

Heh, you know, this is starting to remind me of the 9700 comparisons with the GF4 and how we should all wait for the NV30 to clear any doubt.

Btw, I agree with you, I'm just saying.

NV30 was a quadruple-whammy: late seven months to its competition *AND* a mutt *AND* they lied about it after release *AND* they engaged in . . .err, "overly aggressive" (or whatever formulation they ended up owning-up to) "optimizations".

So, y'know, I totally understand a bit of a queasy feeling about ATI and R520 right now (because I got it too ). . .but there is more than a little ways to go before we should be working ourselves into full-scale "NV30 scale disaster" mode.

 

[Richard]

NV30 was a quadruple-whammy: late seven months to its competition *AND* a mutt *AND* they lied about it after release *AND* they engaged in . . .err, "overly aggressive" (or whatever formulation they ended up owning-up to) "optimizations".

And don't forget *AND* discontinued shortly after.

So, y'know, I totally understand a bit of a queasy feeling about ATI and R520 right now (because I got it too ). . .but there is more than a little ways to go before we should be working ourselves into full-scale "NV30 scale disaster" mode.

I was actually commenting on the whole "these comparisons are useless and let's wait for x part before deciding" (which I agree the same way I agreed the 9700 should not be compared against the GF4, or at least, many conclusions drawn until the NV30 arrived) not on making any sort of parallels between NV30 and the R520.

 

[digitalwanderer]

Methinks all this talk about the sunset of ATi is a wee bit premature, I'd at least hold off until we see the R520.

My personal theory is that they did have some drastic changes of their roadmap sometime early this year which we still don't know about, but that's purely theory.

Then again, you know what they say:

Digital Wanderer was right

 

[no-X]

            NV30      R300
pixel p.     4         8   
vertex p.    3         4
mem. bus    128bit   256bit

NV30 (compared to R300) had only 50% of pixel pipelines, 75% of vertex pipelines and 50% narrower memory bus

            G70     R520
pixel p.    24       12       ( =G70 - 50%)
vertex p.    8        6       ( =G70 - 25%)
mem. bus   256bit   128bit    ( =G70 - 50%)

Somebody thinks than R520 will be a 128bit 12/6 pipeline graphic card? No? So where is the parallel with NV30? (except its 2-3 months delay)

 

[DemoCoder]

Was it really only 4 pipelines, or was it 4 ROPs. Big difference. NV3x would have been a great DX8 card had the R300 not been so good. I'd hesitate to call it a mutt. Just the R300 was so much better. NV3x looked like DX9 was "tacked on" later.

 

[Geo]

I was actually commenting on the whole "these comparisons are useless and let's wait for x part before deciding" (which I agree the same way I agreed the 9700 should not be compared against the GF4, or at least, many conclusions drawn until the NV30 arrived) not on making any sort of parallels between NV30 and the R520.

Yeah, you were selected with the honor of allowing me to reply to the whole thread topic/tone.

 

[no-X]

Was it really only 4 pipelines, or was it 4 ROPs. Big difference. NV3x would have been a great DX8 card had the R300 not been so good. I'd hesitate to call it a mutt. Just the R300 was so much better. NV3x looked like DX9 was "tacked on" later.

Sorry, but 3rd generation of (just) DX8/SM1 cards? It doesn't make much sense. Or did nVidia saw the future of 3D in well-tuned 1.x shaders?

Ok, ROPs. So G70 is 16 ROPs graphic card and R520 will be (50%) 8 ROPs core? (still the same nonsense and still don't see the parallel between NV30 and R520).

 

[Geo]

Was it really only 4 pipelines, or was it 4 ROPs. Big difference. NV3x would have been a great DX8 card had the R300 not been so good. I'd hesitate to call it a mutt. Just the R300 was so much better. NV3x looked like DX9 was "tacked on" later.

"Mutt" is always relative to something tho, isn't it? The sainted TNT2 is a mutt if you compare it to G70.

Tho, generally, I agree with you in the larger point and have said so many times. . .NV30 wasn't as bad as R300 was really just exceptional; "discontinuity" exceptional even, IMHO. Tho that statement is liable to summon Uttar to explain why NV30 really was a disappointment on its own terms too.

 

[dizietsma]

no-x , I have said at least 3 times above that in pure performance that r520 is unlikely to be as nv30 was to r300, what I am trying to say is are Ati following in nvidia's footsteps in regards to going to a new process and suffering delays because of this !

Everyone seems quite happy with nvidia's 7800 approach. Ok, it is not mind blowing but it adds a bit of speed there and it adds some IQ here and you can buy it in the shops on the day it was released. None of those will make the punters unhappy. Compare that to nv30.

Now Ati have a new bar to reach, not only on performance but availability as well . For the first time in a while they might have to catch up on the IQ as well. It seems they have delayed since spring also.

So what I am saying is they are delayed and playing catchup, something nvidia is used to, hence my thread title.

 

[Geo]

So what I am saying is they are delayed and playing catchup, something nvidia is used to, hence my thread title.

Well, that's about an 18 month history there. I don't think NV ever got "used to" it. As bad as NV30 was competitively, NV40 is a damn fine piece of work, and in a relatively short period of time considering the improvement made.

Tho I still think there is a pretty good chance that when all the cards are counted they won't have a clean kill this round again on performance, and thus will still be grinding their teeth --in their hearts, they *know* they are King and ATI is the Pretender.

 

[Jawed]

I think the new memory "system" in R520 is ATI's secret weapon. What if it turns out to perform like a 512-bit bus, but is in fact only a 256-bit bus?

And what if R520 runs at 700MHz core?

And what if we get R580, not R520?...

Has anyone ran Ruby 2 on 7800GTX, by the way? What's the performance like?

Jawed

 

[ANova]

This comparison that has surfaced between the R520 and the NV30 is so rediculous that I forgot to laugh.

delay != bad performance