One last thing I can reveal about NV30...

[Qroach]

I'd like to hear about your info from AMD too.

 

[Tahir2]

Right I will tell you guys, next year sometime.

Just kidding. I'll post it in about 30 minutes. I hope you guys like what information I managed to gain. 8)

Like I said it's a new thread.. but before I post it I need to ask DaveBaumann/Kristoff etc - I don't want them to get into trouble.

 

[Captain-Caverne]

AFAIK (and i know only this so don't ask for more ), the announcement date was delayed one month from the original one...

IIRC the original announcement date was supposed to be in August, implying that we should see something in the next 2 or 3 weeks?

Well i've heard that the original date was around mid-August... do the math by yourself then

 

[elimc]

That's two more.. only one more person needed and All Shall Be Known (TM)

This thread is being hijacked! AHHHHHHhhh . . . I would like to know, too.

 

[Tahir2]

Yep I hijacked it.
Soon I will have my own thread and you will all Respect Me (TM)

You can hijack that thread as well if you really want to.

Update: Sorry guys you will have to wait a little longer for the information. Real life just got in the way. Doomtrooper and Qroach, my AMD source in fact had the most to say.. Thanks for waiting with Baited Breath (TM), but soon All Shall Be Revealed (TM)

 

[Dio]

Did SiS and Trident help you come up with those trademarks?

 

[pascal]

OhMyGod(TM) this VapourThread (TM) was hijacked!!!
MyReliableSource (TM) says YouMustBeJocking(TM)
And ICantTalkAboutItBecauseIAmUnderNDA(TM)

 

[cellarboy]

By announced I mean that Nvidia will announce NV30, like ATI announced R300.

What? You mean Nvidia are going to announce the NV30 along with working samples and have it available for purchase 30 days later.

Maybe if they announce it in November, then they will do it "like ATI announced R300." Otherwise it'll be the biggest paper launch since I tripped over the dog whilst carrying my paper recyling to the car.

 

[Tahir2]

Did SiS and Trident help you come up with those trademarks?

Well yes that was my school of thought. They say that Imitation is the Sincerest Form of Flattery (TM) ... pokes fun at pascal

I have always wondered who 'They' were.

 

[Nagorak]

I still think the NV30 is going to be just barely better than the R300.

There's no reason to assume otherwise. Radeon and Geforce 2 were basically on-par. Radeon 8500 and Geforce 3 Ti500 are basically on par (actually if you look now the 8500 can keep up with the Ti4200). The idea that the NV30 is just going to blow the R9700 away, after it already blew the Ti4600 away is just a dream.

I'm sure Nvidia will hype it up (as usual), but it's not going to be anymore earthshattering than their previous "revolutionary" cards.

 

[Tahir2]

Anyone interested should head on over here:

http://www.beyond3d.com/forum/viewtopic.php?p=33755#33755

Remember kids, you didn't hear it from me.

 

[Sage]

ill tell you why NV30 will blow away R300. because it is Rampage. if you dont know (may god help you, jk) Rampage is ancient stuff. The Rampage project was to originally produce the sucessor to the original Voodoo Graphics, but due to problem after problem it was continuously delayed and around the time the Voodoo3 came out, the design had to be scrapped because of the impending GeForce. Nevertheless, the Rampage project continued to work on a chip that would change everything. The architecture is really revolutionary in more than just the specs we all see, it has been totally redesigned from the ground up, and i mean totally. Do not confuse this with simply another generation, it is another world. Just look at the 3Dfx bastard chip, the Napalm (half Rampage half Voodoo). Yes, it was inferior in many ways to the competition but only because of its legacy Voodoo tech. However, if you forget about the TnL the VSA-100 / 101 become very superior to all competing products. The Voodoo5 5500 scales with CPUs past 2GHz even. It really was quite incredible for what and when it was.

Alas, i believe there to be a curse on the Rampage and anyone who attempts to create it. Look what it did to 3Dfx. Now, look at all of the problems nVidia is having getting it to work, and look at where the competition is. nVidia could be behind an entire cycle before the Eclipse hits the market in mass. Also look at nVidias main North American card producer- VisionTech. VisionTech also made the high end Elsa cards used for CAD and DCC. it gets worse. the SEC is now investigating nVidia. also the price of nVidia stocks has sunk unbelievably. For those of you not in the US, the SEC has been investigating a lot of companys for fraudulent revenue reports and has totally destroyed a few companys. SEC investigation is very bad, even if they are found innocent.

I'd like to say one more thing: welcome back, Trident.

 

[multigl2]

I hope your kidding.

 

[Richthofen]

"
Maybe if they announce it in November, then they will do it "like ATI announced R300." Otherwise it'll be the biggest paper launch since I tripped over the dog whilst carrying my paper recyling to the car.
"

well who cares?
The majority of people definetly not
Launch is launch no matter if it is on paper or not.
If they launch it on paper it will hold people back from buying a R300 card.
Well that's just fine with me

 

[alexsok]

Just a question there Alexsok, if you know so much about Nv30 why are you over at cgshaders.org begging for information ??

Your source should be able to answer those questions

If you look closely, I asked there questions about various features of the architecture, not NV30.

 

[kid_crisis]

Alas, i believe there to be a curse on the Rampage and anyone who attempts to create it. Look what it did to 3Dfx.

.....the following is an intercepted email message concerning NV30 development:

First all layer tapeout happened, came back DOA. Attempts to resuscitate chip with metal layer changes at the gate level not successful. FIBs? Are you crazy? Didn't you hear me say chip was DOA? Problems traced back to immaturity of .13 tools that appear to give wrong results during backend layout. Timing and load analysis during routing turned out to be cr*p.

All prayed that the all metal layer change could be done in 3 months, but alas, the cursed .13 tools were just no d*mn good and they didn't work. Full resynthesis from verilog with fixed versions of tools are happening and everybody is still praying they will work this time, they are days, make that weeks from tapeout and their fingers are crossed.

Murphy's law ran rampant throughout this product cycle and those 120M transistors turned out to be beyond what those cursed .13 tools are able to handle. Did we mention the .13 tools are cursed? Hardly a day passed that the backend guys did not curse the guy who made the call to go to .13, but he already cashed out his 8M and is living somewhere warm safely out of the earthquake zone. Come to think of it, where are the talented engineers with the intellect the size of planets and stock options that will feed a small nation? Who are sending all these postcards from Maui and Bahamas with pictures of islands and white sandy beaches in the lunch room.

But time heals all cursed tools and finally! it appears the new tools are a lot less cursed and there is light at the end of the tunnel. We will see if lady luck will shine on us this time.

 

[Dio]

ill tell you why NV30 will blow away R300. because it is Rampage. <snip> Nevertheless, the Rampage project continued to work on a chip that would change everything.

I don't understand this. Fundamentally there sounds like there is virtually no difference in technology between nv30 and R9700. Nobody has come up with anything new. They are just both standard Z-buffered rasterisation engines with Z-features for bandwidth reduction & perf improvement and DX9 support. Both have floating-point; both are fully capable of cinematic rendering.

It would not surprise me in the slightest if R9700 is slightly superior to nv30. Neither would it vice-versa. I expect the difference between the two to be small. Nobody has any reason to believe that either should be faster than the other because one of them DOESN'T EXIST YET!

Everybody reading this, ask yourselves one simple question. Do you believe the hype?

 

[Joe DeFuria]

I'll tell you what I believe:

The performance difference between NV30 and R-300 will more than likely come down to the difference in core clock and physical bandwidth. Nothing new here.

 

[kid_crisis]

But time heals all cursed tools and finally! it appears the new tools are a lot less cursed and there is light at the end of the tunnel. We will see if lady luck will shine on us this time.

BTW, I believe that the email about 0.13um and cursed tools is a joke. I found it over on silicon investor.

At least I hope its a joke.

 

[Pete]

Otherwise it'll be the biggest paper launch since I tripped over the dog whilst carrying my paper recyling to the car.

ROFL!