anything new on the power vr front ?

[Mulciber]

Does anyone have a VideoLogic Neon 250 they'd like to sell to a collector? I've been on the lookout for one for a while, but can't seem to find any

I think I have 3 from my good'old Graphics Card Reviewer days

ew! I PM'd you

Does anyone remember the Dual Voodoo1 8mb board that came out in 97'? I cannot remember the name of it.

 

[Ailuros]

Didn't PVR recently release their MBX for integrated devices?

Here's a detailed description for it for those interested:

http://www.arm.com/techdocs/5ETJKM/$File/DTO0003B_MBX.pdf?OpenElement

Also here's a recent announcement about a mobile phone architecture carrying MBX Lite:

http://www.imgtec.com/News/PressReleases/SHMobileAndMBXLiteProvideLeadingSolution.asp

 

[Ailuros]

The duck laid some eggs.

Damn, I better watch my step then, don't want to crush them by accident

That was revealing! It will be called Omelette 9900

 

[Fuz]

Dual V1? I remember a 6mb card from Canopus(sp?) that could render at 800x600, but I can't recall a dual card. You are not talking about the dual V2 board from Quantum 3D, the Obsidian X-24 are you?

 

[Mulciber]

Dual V1? I remember a 6mb card from Canopus(sp?) that could render at 800x600, but I can't recall a dual card. You are not talking about the dual V2 board from Quantum 3D, the Obsidian X-24 are you?

Nope, I recall there being a dual voodoo1 8mb. I may be wrong, but I do remember seeing it somewhere

 

[Nick[FM]]

Dual V1? I remember a 6mb card from Canopus(sp?) that could render at 800x600, but I can't recall a dual card. You are not talking about the dual V2 board from Quantum 3D, the Obsidian X-24 are you?

Nope, I recall there being a dual voodoo1 8mb. I may be wrong, but I do remember seeing it somewhere

I know of Voodoo1 SLI, if that's what you mean?

 

[Ozymandis]

Does anyone remember the Dual Voodoo1 8mb board that came out in 97'? I cannot remember the name of it.

Are you talking about the Quantum3D Obsidian? Although I don't recall how much memory it had on it, that was a dual Voodoo1 card.

 

[ram]

Does anyone remember the Dual Voodoo1 8mb board that came out in 97'? I cannot remember the name of it.

Are you talking about the Quantum3D Obsidian? Although I don't recall how much memory it had on it, that was a dual Voodoo1 card.

Yea, and Quantum3D also fabbed Voodoo 2 AGP boards (I still have one of these black parts on the desk). It's broken unfortuantly.

 

[Blitzkrieg]

yes they did release more gpus that bitboys thats why i said nearly, not are.
and also the kyro2 and dreamcast where a long time ago.
What happened? the Dreamcast got owned by Playstation and failed(not pvrs fault but it still happened) and im not aware of the kyro1 or 2 winning any OEM contracts.
To win an OEM contract u need a top to bottom range of cards which they have never had. Also most importantly you need consistency and to show you will have new products and will continue support , neither of which pvr does or has ever done. They are relegated to their MBX core. And even then i think they are eventually going to get overtaken because they lack the size and money to compete with the bigger players.

 

[MatrixShark]

Dual V1 card

The actual name was the Quantum3D Obsidian 100SB. Here's a link to an old benchmark by Tom's. A trip down memory lane...

http://www6.tomshardware.com/graphic/19980121/3dbench-02.html

 

[Ozymandis]

Yea, and Quantum3D also fabbed Voodoo 2 AGP boards (I still have one of these black parts on the desk). It's broken unfortuantly.

Didn't their Voodoo2 AGP boards have two chips running in SLI? I remember that they were damn expensive

 

[Ailuros]

They are relegated to their MBX core. And even then i think they are eventually going to get overtaken because they lack the size and money to compete with the bigger players.

For the time being I'd say that I'd like to see 3D capable PDA/mobile chips from the big players first, before I'd go into any assumptions about future possibilities.

With the rest of your post I agree more or less.

 

[Fuz]

Re: Dual V1 card

The actual name was the Quantum3D Obsidian 100SB. Here's a link to an old benchmark by Tom's. A trip down memory lane...

http://www6.tomshardware.com/graphic/19980121/3dbench-02.html

OMG

I had forgotten why I used to visit Toms site years ago. Did you see all those image quality shots! Man, that site has changed during the years.

 

[Blitzkrieg]

Yes thats true Ailurios
But with the size of pvr i wouldnt be suprised to see them bought out eventually.(If they make a tempting enough product)

 

[Ailuros]

Yes thats true Ailurios
But with the size of pvr i wouldnt be suprised to see them bought out eventually.(If they make a tempting enough product)

You mean Imagination Technologies of which PowerVR is actually a division? If my memory serves me well (apart from the possible patents and technology portofolio), the amount of engineers should exceed a total of 180.

Latest SoC's from all three divisions are gaining licence after licence for a wide palette of products. If you think that ImgTec's income/revenue depends only on PowerVR and exclusively on PC gaming products (or even console designs), then you have only a very small fraction of the whole picture here.

According to the companies own predictions in their recent Interim results the forecasted number of MBX units is supposed to be between 5x to 10x higher than that of a PC graphics part in 2003. Then there are also the digital TV/video playback IP cores that are apparently already under production.

 

[Striker]

Two notes: a) THG apparently was a very good reviewing site back in 1998. Its IQ comparisons are quite fair imo, and its a pity they're not done in 2003.

b) "NVidia's RIVA 128 is not famous for highest image quality and I am sure that you can see the reason why. "

http://www6.tomshardware.com/graphic/19980121/3dbench-28.html

Hehe, I suppose the story continues from RIVA 128 to our days..
(Although the RagePro was an AWFUL renderer back then..)

 

[Entropy]

It's pretty clear by now that for this generation, nVidia concentrated on the DX9 feature set, and not pixel pushing capability. This leaves a window of opportunity for other players, most notably ATI, to grab market- and mind-share.

But ATI are not the only ones to benefit, SIS have actually managed to gain a toe-hold already, and their new design appears to be significantly improved. Their UMC connection ensures fabbing ability at low cost if they were to take off. S3s DeltaChrome, while targeted at mobile application, may prove attractive for low-mid desktop use as well. Both SIS and VIA have good connections with local Taiwanese manufacturers.

There's also an opportunity at the higher end of the performance spectrum. The retail NV30 is a case of too little at too high a cost, and while the NV35 may make an earlier than expected appearance, it is difficult to envision that it could improve all that much on its NV30 foundation given that the NV30 was already at (over?) the limits imposed by process/power/cost. Anandtech, with advance info, expressed some doubts that the NV35 would be able to outperform the R9800.

Thus, the PowerVR design that Kristof felt would be a "performance leader" should be able to perform well comparatively, and would only need to be priced solidly below ATI and nVidia top-of-the-line in order to recieve consideration from a significant part of the market.
The opportunity is there.
Opportunities however are only there for the grabbing for a short while, then they're gone - someone else took it instead, or the window slowly closed.

The Neon 250 and the Kyro 2 were too late to market, the Kyro III never made it at all. The window is open NOW, but for how long? If there had been a product on the way, whoever was going to market it would start hyping it NOW in order to gain mind-share, and keep upgraders hanging on to their money just a bit longer to see how the new offering pans out, and perhaps even start working on winning OEM deals, or at least preparing the ground for future deals.

Instead, we have - nothing.
Just as with the Kyro III.

I've got a bad feeling about this.

Entropy

 

[Ailuros]

Thus, the PowerVR design that Kristof felt would be a "performance leader" should be able to perform well comparatively, and would only need to be priced solidly below ATI and nVidia top-of-the-line in order to recieve consideration from a significant part of the market.
The opportunity is there.
Opportunities however are only there for the grabbing for a short while, then they're gone - someone else took it instead, or the window slowly closed.

I insist on my original question and I'll try to expand it:

estimated time of release and specifications?

All of your considerations are valid rather from a perspective of an equivalent featureset to whatever S3/SiS are planning to release (I don't see either/or making it to shelves prior to H2 2003 anyway). The only other reason why one would guestimate that they (editVR) won't go higher than expected is their recent track record of releases and their according specifications. But does it really represent a guarantee?

At the danger of getting endlessly flamed for it: in a pure hypothetical case scenario of a let's say Q3 or Q4 release while reaching the absolute highest dx9 shader threshold, how would that look as a perspective?

 

[Entropy]

The current market window will definitely close by the time nVidia announces a new line of low-mid-high end cards. If PVR products haven't gotten into shops by then, and for sufficiently long to have achieved some momentum/user base, it's too late.

Say roughly three or four months on the market before the nVidia announcement.

As for when nVidia will do that, I really don't know. However, it seems likely that their market position will erode until they do so, so I'd assume that nVidia will try to move fast. To what extent the arrival of such an updated product line-up could be hampered by lack of availability of improved lithographic processes is another unknown.

My guess is that nVidia will shoot for at least announcements in late November whether they have products available or not, (just as this year) in order to hamstring opposition holiday sales.
Your guess is probably as good as mine though.

Also, ATI is as entrenched as nVidia, so the same applies for them announcing the R400 along with companion parts, which may serve to narrow the window of opportunity further.

A PowerVR part has to show clear advantages in price/performance over ATI and nVidia offerings and generally seem preferable to SIS/S3/Whoever.

All I'm saying is that my suspicion is growing that they have failed to convince anyone of this sufficiently to fund bringing the part to market in a timely manner.
I would be very pleased if this was proven wrong.

Entropy

 

[Reverend]

Thus, the PowerVR design that Kristof felt would be a "performance leader" should be able to perform well comparatively...

Can you point me to where K "felt" this way?