Kyro 3 Effectively Dead!
- Thread starter Crusher
- Start date
Because we haven't seen a T&L part out from them yet and it is reasonable to expect that a DX7 part was being worked on for some time now but due to problems it hasn't been released.
In case anyone is reading into what I said earlier, I was merely voicing my presumptions. I have no clue what PVR is/has been up to.
In case anyone is reading into what I said earlier, I was merely voicing my presumptions. I have no clue what PVR is/has been up to.
To avoid misunderstandings I'd like to see next time around at least a half way dx9 product from them. Full dx9 compliance would be ideal of course. That's what I actually meant.
Kristof,
I figure you could at least answer that question: is the VPE to be found on upcoming MBX PRO dx8/8.1 compliant or not? (Or haven't the full specs for it been announced yet?)
Kristof,
I figure you could at least answer that question: is the VPE to be found on upcoming MBX PRO dx8/8.1 compliant or not? (Or haven't the full specs for it been announced yet?)
ActionNews
Newcomer
Reverend said:
As i know PowerVR has a Vertex Shader in their PowerVR MBX Chip and i think David Harold said it is full DX8 kompatible
! Kristof correct me if i'm wrong !IAnd i know the PowerVR chpis for Naomi2 from SEGA anno 1999 had a hardware T&L-engine too ! So a DX8-chip should be no problem for PowerVR
!CU ActionNews
ActionNews
Newcomer
Ailuros said:
Hmm...there is a rumor from www.hardware.fr wich says that PowerVR Series 5 will be full DX9 compliant, but i think Kristof will not comment on this
! IMHO should be ready more or less 1. half of next year !CU ActionNews
Dave B(TotalVR)
Regular
Sorry maybe Mike being a nice guy was a little subjective rather than fact.
The Inquirer makes fun of itself so thats how it should be viewed I think. Also The Inquirer was 100% accurate when they revealed that Palomino would have SSE and SSE2 would be on Hammer line of chips only. This was before launch.
The Inquirer is mostly full of crap and you'd be a fool to trust it but also a fool to ignore every rumour too. Maybe... anyway I dont know why I am defending Mike.. maybe it is cos I felt sorry for him the way The Rogister treated him. I love to root for the underdog even when he is a complete rascal.
The Inquirer makes fun of itself so thats how it should be viewed I think. Also The Inquirer was 100% accurate when they revealed that Palomino would have SSE and SSE2 would be on Hammer line of chips only. This was before launch.
The Inquirer is mostly full of crap and you'd be a fool to trust it but also a fool to ignore every rumour too. Maybe... anyway I dont know why I am defending Mike.. maybe it is cos I felt sorry for him the way The Rogister treated him. I love to root for the underdog even when he is a complete rascal.
misae said:
Mike Magee is a hack. But by all appearances he is a GOOD hack. Read the articles at the Inquirer, and you see news written by someone who has dug for information, rather than simply rewrites official PR. If the material he bases his conclusions on is on the web, he posts the links, and you can dig through it yourself, if you are interested. The same goes for Paul Hales.
Mike Magee founded both the Register and the Inquirer, and the service they provide is easy to snub your nose at, but it is useful. And if you don't think so, don't go there.
I even appreciate them enough that I try hard to forgive Mike that he allows Fuad to write. I guess that in order to appreciate quality hacks, Mike figures you need to have something to compare to.
Entropy
i find the inquirer one of the most valuable news website.
as some others stated, they don't report news that they read on other sites, or rephrase PR.. (a lot of news websites do nothing more than re-re-reporting news)
they report new information that they gathered themselves, and happen to be right a lot of times.
as some others stated, they don't report news that they read on other sites, or rephrase PR.. (a lot of news websites do nothing more than re-re-reporting news)
they report new information that they gathered themselves, and happen to be right a lot of times.
Dave B(TotalVR)
Regular
I sure hope Series 5 is a DX 9 part and I sure hope that it hasn'tbeen held up by the KYRO 3 delay, hopefully we will see it this year (hah, call me an optimist) as I dont think VIA are gonna want to hang around waiting for the graphics tech to become out of date. S5 really does need to be got out of the door soon though because theproblem powervr has at the moment is lack of feature support so getting developer support will be consisting of persuading devs to allow for the missing hardware as apposed to allowing for the great advantages of PowerVR. IMO compliance is the most important thing to PVR at the moment as developer support (which has grown enourmously but is still lacking) comes with this IMGTEC need to push it. When the KYRO came out it had all the features and was quite quick, thats why the KYRO ended up getting the support it did.
Dave
Dave
Dave B(TotalVR) said:
Hmm.
While your analysis seems reasonable, I see a couple of problems with the VIA-Series 5 scenario.
First, VIA had the Columbia under development since way before ST withdrew from graphics, and it is still the part that VIA refers to as their next higher end gfx product. This would seem to clash with the notion that VIA would try to get a Series 5 out the door in a similar timeframe.
Secondly, unless I misremember, ST had a liscense for the Series 5, and if this would require VIA to deal not only with IMG, but also to try to make some kind of deal with ST, I'd say that this would make the series 5 less appetizing for VIA. Too much money and time spent on lawyers and liscenses in order to offer a part which competes with what they have already developed in-house.
Do you have any indication that VIA actually wants to bring a Series 5 part to market?
Entropy
relativity
Newcomer
Indeed, STM did sign an agreement for Series 5. However, as with all
technology licenced from IMG, its use is not exclusive to STM. So, AFAIR,
IMG is free to licence the technology to others. However, STM owns the
KYRO trademark. It also owns all right to any silicon produced or the
PG tapes. This is the effect with KYRO III. STM can not without IMG
agreement sale the tapes to any third party. It can sell them to IMG, but
I understand, this may not have occured.
STM has screwed up big. Their actions has had effect on IMG and other
parties. I am sure IMG is a strong enough company, that even in this
difficult market down turn, it will come through stronger; with a partner
which understands graphics much better than STM deed.
Entropy,
Where's smoke, there's also fire (or some sparks if you prefer); at least that's what they say. If "in house" tech would be sufficient for what VIA intends to do within the PC graphics market, they wouldn't have been researching for alternatives.
I'm not so sure if I would place my bets that easily on VIA for the future (I really have no idea what is going on now in the background), after all they seem to have a track record in breaking major deals in the past
As things look like any negotiations that took place between the 3 interested parties concerning Series4 sounds more like the responsibility of ST Micro in the end.
Quite a few details were known about Series4; compare it to Zoetrope. What makes you think that the analogy will not be the same in say Series5 vs. Columbia? That's the only logical explanation as to why VIA would be researching for alternatives.
Where's smoke, there's also fire (or some sparks if you prefer); at least that's what they say. If "in house" tech would be sufficient for what VIA intends to do within the PC graphics market, they wouldn't have been researching for alternatives.
I'm not so sure if I would place my bets that easily on VIA for the future (I really have no idea what is going on now in the background), after all they seem to have a track record in breaking major deals in the past
As things look like any negotiations that took place between the 3 interested parties concerning Series4 sounds more like the responsibility of ST Micro in the end.
Quite a few details were known about Series4; compare it to Zoetrope. What makes you think that the analogy will not be the same in say Series5 vs. Columbia? That's the only logical explanation as to why VIA would be researching for alternatives.
One area that would seem suitable for IMG tech is UMA graphics for low cost motherboard chipsets. There I can see an excellent fit for VIA. Of course, the liscensing costs can't be too high, that would defeat the entire purpose of integrated graphics. And if VIA has a track record of anything, it is one of not wanting to pay excessive liscensing fees.
But I hope a deal can be made for those kinds of products.
I'm less convinced as far as standalone graphics cards go. Regarding the comparison between the Zeotrope and Series 4, I'd say that the Zeotrope wins on features, and the Series 4 would win on effective fillrate. Columbia vs Series 5 would be complete speculation of course. The Series 5 would be expensive for VIA, as they already carry the costs for the in-house development of the Columbia, plus the would have to pay liscensing fees and silicon realization costs of the Series 5. They then have to produce and sell in sufficient numbers and with sufficient profit-margin to pay for it all. Carrying such excess cost and still make a profit seems like a stiff task in todays gfx business.
Taking a broader view, it doesn't make much sense to me for VIA to drive in-house graphics development and then turn around and liscense their gfx tech from IMG. In that case they should scrap their in-house development completely, cutting costs. If they go with IMG, they can either liscense tech, or buy the whole outfit similar to how they bought S3.
So I can possibly see VIA liscensing IMG tech for UMA motherboard graphics, where it offers compelling advantages. I'd guess that is where the smoke is coming from, particularly since Harold referred to chipset gfx specifically as promising. http://www.powervr.org.uk/ I have a much harder time making sense of VIA liscensing and producing a Series 5 gfx chip for standalone gfx-cards. That would require the Series 5 to offer roughly R300/NV30 performance and features at a much lower cost, that is, offer something VIA doesn't already have (paid for) in the Columbia.
Entropy
But I hope a deal can be made for those kinds of products.
I'm less convinced as far as standalone graphics cards go. Regarding the comparison between the Zeotrope and Series 4, I'd say that the Zeotrope wins on features, and the Series 4 would win on effective fillrate. Columbia vs Series 5 would be complete speculation of course. The Series 5 would be expensive for VIA, as they already carry the costs for the in-house development of the Columbia, plus the would have to pay liscensing fees and silicon realization costs of the Series 5. They then have to produce and sell in sufficient numbers and with sufficient profit-margin to pay for it all. Carrying such excess cost and still make a profit seems like a stiff task in todays gfx business.
Taking a broader view, it doesn't make much sense to me for VIA to drive in-house graphics development and then turn around and liscense their gfx tech from IMG. In that case they should scrap their in-house development completely, cutting costs. If they go with IMG, they can either liscense tech, or buy the whole outfit similar to how they bought S3.
So I can possibly see VIA liscensing IMG tech for UMA motherboard graphics, where it offers compelling advantages. I'd guess that is where the smoke is coming from, particularly since Harold referred to chipset gfx specifically as promising. http://www.powervr.org.uk/ I have a much harder time making sense of VIA liscensing and producing a Series 5 gfx chip for standalone gfx-cards. That would require the Series 5 to offer roughly R300/NV30 performance and features at a much lower cost, that is, offer something VIA doesn't already have (paid for) in the Columbia.
Entropy
I read in a French magazine Joystick that Hercules is working on the... 3D Prophet 4800 based on Kyro II SE (DX7 card) now.
It seems that Hercules was too busy concentrating on launching their Radeon series cards and now that that is done, they will release the Kyro II SE.
Don't know if it's true or not, but heh, who knows?
It seems that Hercules was too busy concentrating on launching their Radeon series cards and now that that is done, they will release the Kyro II SE.
Don't know if it's true or not, but heh, who knows?
RaolinDarksbane
Newcomer
Why does people have problems with Kyro next card being DX7 when Nvidia and ATI are still producing DX7 cards(GF4MX and 7000/7500)??
Similar threads
