Steve Dave Part Deux
Newcomer
Legion, my only reasonable guess would be that Hitachi plans to take Motorola's place next to IBM.
MS teaming up with ATI for the next xbox?
- Thread starter bbot
- Start date
If the choice is between ATI or Nvidia, I would have to go with Nvidia being the most likely choice for GPU design. Although I think ATI would have a good shot on getting in on the action.
The biggest problem is fabbing the chips for the X-Box 2. Intel probably has the best fabs that Microsoft could use. I think a lot of possibilities are out there. One possibility is that Intel could merge with ATI (this is wild speculation) or Nvidia and then fab both the CPU and GPU for the X-Box. I'm sure there will be some intresting twists and turns along the road to the X-Box 2.
The biggest problem is fabbing the chips for the X-Box 2. Intel probably has the best fabs that Microsoft could use. I think a lot of possibilities are out there. One possibility is that Intel could merge with ATI (this is wild speculation) or Nvidia and then fab both the CPU and GPU for the X-Box. I'm sure there will be some intresting twists and turns along the road to the X-Box 2.
heh, I guess I should choose my words more carefully.
All I meant was that I don't think AMD is in a position where they are looking or wanting to be bought out by Nvidia (or some other company), nor do I think Nvidia is currently in a position to be taking such a (potential) risk. Perhaps if Hammer doesn't perform as well in the market as they are hoping (which, given market conditions, is possible), but I don't see it right now. Either way it is purely speculation as to their future.
All I meant was that I don't think AMD is in a position where they are looking or wanting to be bought out by Nvidia (or some other company), nor do I think Nvidia is currently in a position to be taking such a (potential) risk. Perhaps if Hammer doesn't perform as well in the market as they are hoping (which, given market conditions, is possible), but I don't see it right now. Either way it is purely speculation as to their future.
Ooh-videogames said:
Bzzt! Wrong
The 3D part of Saturn's GPU that was such a hell to code (quadratic primitives) was made by none other than the side subject of this thread - yes, nVidia.
What was it, NV3 right? Or was it NV2... I don't remember offhand, look it up in the History of nVidia article at FiringSquad... The 2D part of the core, which is probably what you're talking about, was a designer's dream, from what I understand of it
Tag,
Nvidia didn't have anything to do with the sega saturn hardware. Saturn also didn't support quadratic primitives, only triangles. The two main chips in the Sega Saturn were hitachi designed processors. Nvidia worked out a deal with sega to convert some of their games to ship with a PC card called Diamond edge 3D (If I remember the name correctly).
You could essentially call it the NV1 since it was the first consumer card nvidia produced.
Nvidia didn't have anything to do with the sega saturn hardware. Saturn also didn't support quadratic primitives, only triangles. The two main chips in the Sega Saturn were hitachi designed processors. Nvidia worked out a deal with sega to convert some of their games to ship with a PC card called Diamond edge 3D (If I remember the name correctly).
You could essentially call it the NV1 since it was the first consumer card nvidia produced.
Ooh-videogames said:
Hitachi didn't design the Saturn though. SEGA designed it. It was an ok design until they threw in that second SH2 at the last minute when they found out about the polygon throughput of of PSX.
Hitachi's cpu's are very well designed and it had nothing to do with the relativey poor design of the Saturn.
Model 3 used the LHM chip.
Also the arcade version of the Saturn was the ST-V Titan.
SEGA has had a long history in using different cpu and gpu manufacturers. Their arcade boards have used cpus from Hitachi, Intel, NEC, Motorola, IBM and gpus from Fujitsu, Lockheed Martin, PowerVR, and of course thier own in house designs.
BTW the Model 1 board was the first real true 3D system from SEGA.
Also the arcade version of the Saturn was the ST-V Titan.
SEGA has had a long history in using different cpu and gpu manufacturers. Their arcade boards have used cpus from Hitachi, Intel, NEC, Motorola, IBM and gpus from Fujitsu, Lockheed Martin, PowerVR, and of course thier own in house designs.
BTW the Model 1 board was the first real true 3D system from SEGA.
That's what Yu-Suzuki wanted Sega to do, instead, they went with their internal R&D team to put together Saturn, if it wasn't for PSX, Saturn would be quite lower spec too, probably 32X level.
I also read somewhere, that Sega executives were quite shock to see PSX 360,000 polygons/s figure is higher than their high-end board at the time, Model 2 at 300,000 polygons/s. I find that funny if that was true
Saturn 3D aren't that bad, I prefer some of the Saturn version of games compare to PSX version, though I think both are quite horrid. Last generation, they should still stick with 2D games.
A little more...
4 if you include the SH-DSP in the SCU (System Control Unit). Throw in VDP1, VDP2, the FH1 DSP and M68EC000 of the SCSP, and the SMPC (a little 4-bit Hitachi microcontroller) and you're looking at managing 9 processors...
Well the quirky thing about the system was that it's CPU config wasn't SMP, rather a master/slave setup. That along with the distributed cache pool setup (much like the PS2) and you weren't really contending for data on the main bus. The processor configuration did make it a pain at times to maximize utilization though.
While you're correct about the Nvidia part, you're wrong about the primitave support of VDP1.
VDP1's primitives were called 'parts', and it drew textured parts and non-textured parts. Textured parts consisted of a normal sprite, scaled sprites, and distorted sprites. Distorted sprites were like normal sprites except they could be rotated and 'distorted' by specifying the coordinates of its' four corner points and it would map the bitmapped image within those points.
Non-textured parts were straight forward, very much like distorted sprites except you applied an RGB value or a single palette color. You had a polygon (a four-point primitive (which could also be gouraud shaded), a polyline which was like a polygon except you didn't fill it, you only coloured it's edges. And you had a basic line.
4 if you include the SH-DSP in the SCU (System Control Unit). Throw in VDP1, VDP2, the FH1 DSP and M68EC000 of the SCSP, and the SMPC (a little 4-bit Hitachi microcontroller) and you're looking at managing 9 processors...
Well the quirky thing about the system was that it's CPU config wasn't SMP, rather a master/slave setup. That along with the distributed cache pool setup (much like the PS2) and you weren't really contending for data on the main bus. The processor configuration did make it a pain at times to maximize utilization though.
While you're correct about the Nvidia part, you're wrong about the primitave support of VDP1.
VDP1's primitives were called 'parts', and it drew textured parts and non-textured parts. Textured parts consisted of a normal sprite, scaled sprites, and distorted sprites. Distorted sprites were like normal sprites except they could be rotated and 'distorted' by specifying the coordinates of its' four corner points and it would map the bitmapped image within those points.
Non-textured parts were straight forward, very much like distorted sprites except you applied an RGB value or a single palette color. You had a polygon (a four-point primitive (which could also be gouraud shaded), a polyline which was like a polygon except you didn't fill it, you only coloured it's edges. And you had a basic line.
Definitely quadratics...
And if there's no nVidia of any kind within Saturn, why did FiringSquad do a big investigation into it as a follow-up to their History of nVidia article, titled Whatever Happened to NV2 or something like that...? I'd look myself right now, and quote, and stuff like that, but my school's proxy blocks FS
I know they talked about SEGA grabbing nVidia's then-latest chip at the last possible moment as a response to the 'amazing' 3D capabilities of PlayStation X... was all of their research wrong?
And if there's no nVidia of any kind within Saturn, why did FiringSquad do a big investigation into it as a follow-up to their History of nVidia article, titled Whatever Happened to NV2 or something like that...? I'd look myself right now, and quote, and stuff like that, but my school's proxy blocks FS
I know they talked about SEGA grabbing nVidia's then-latest chip at the last possible moment as a response to the 'amazing' 3D capabilities of PlayStation X... was all of their research wrong?