Dreamcast-Talk.com

@CruSega. The Saturn was originally "Model 1" based early in its development. There's a popular myth that the chipset was switched because of "Playstation". This is untrue because the PSX was not known publicly until 1994. Both Nintendo & SEGA knew nothing about Sony's plans until early 1994.

What stu is claiming about Model 3 is false. Model 3 came MUCH later & used a completeley different 3D Engine. Model 3 also used an IBM CPU as oppossed to an Hitachi CPU, it is stand alone hardware & not related to Model 1 & 2.

Now let me explain what happened & how the Saturn came to be: In 1991-1992 SEGA obtained some new 3D polygon techology from Lockheed Martin. From this techology came 2 chipsets both 32-bit. One very primatitive capable of only Polygon Rendering & Texture Mapping. The other, alot more powerful with Translucency, Flat Shading, Mirco Textiring, Diffuse Refection & Much smoother true 3D features.

In early 1993, after scrapping the System 32 based 2D only hybrid "GigaDrive" SEGA set to work on a completeley different project that used parts of System 32 with tacked on Model 1 they called "Aurora" in which they had planned on
dubbingh "Jupiter".




Now here's what happened & here's the real truth behind the switch. At JAMMA 1993, SEGA demostrated what they thought was the wave of the future by displaying its Model 1 arcade hardware. Needless to say, Japanese
crowds weren't impressed with the advent of basic 3D & Clamored for better
looking 3D. So naturally SEGA panicked & made key emergency design changes in Aurora. First, SEGA had to trash the Model 1 based VDP. Realizing that incorporating Model 2 hardware would have been a bit expensive & slightly time consuming since Model 2 used a state of the art high resolution single core VDP, SEGA decided that the solution was to replace it with 2 low res VDPs. The actual so called "Off The Self Conponets" are two VDPs that split the Model 2's power in half. But when combined together, run full Model 2 which is superior to PSX despite lack of key 3D Engine features. But as you said, Model 2 had plenty of Loopholes around that.


The finish touch design for Aurora was that as a cost cutting measure, SEGA finally scrapped the cartridge based Jupiter and went for CD ROM. When all the changes in Aurora were finalized, the chipset was officially named "Saturn".

The Model 1 board was a commercial and critical failure. In fact it bombed so bad that SEGA retired it in less then a year. And here's a FACT: Most of the early Model 2 games like Daytona USA,VF2,SEGA Rally,Virtua.Cop were originally developed for Model 1.

MrSega wrote:@CruSega. The Saturn was originally "Model 1" based early in its development. There's a popular myth that the chipset was switched because of "Playstation". This is untrue because the PSX was not known publicly until 1994. Both Nintendo & SEGA knew nothing about Sony's plans until early 1994.

What stu is claiming about Model 3 is false. Model 3 came MUCH later & used a completeley different 3D Engine. Model 3 also used an IBM CPU as oppossed to an Hitachi CPU, it is stand alone hardware & not related to Model 1 & 2.

Now let me explain what happened & how the Saturn came to be: In 1991-1992 SEGA obtained some new 3D polygon techology from Lockheed Martin. From this techology came 2 chipsets both 32-bit. One very primatitive capable of only Polygon Rendering & Texture Mapping. The other, alot more powerful with Translucency, Flat Shading, Mirco Textiring, Diffuse Refection & Much smoother true 3D features.

In early 1993, after scrapping the System 32 based 2D only hybrid "GigaDrive" SEGA set to work on a completeley different project that used parts of System 32 with tacked on Model 1 they called "Aurora" in which they had planned on
dubbingh "Jupiter".




Now here's what happened & here's the real truth behind the switch. At JAMMA 1993, SEGA demostrated what they thought was the wave of the future by displaying its Model 1 arcade hardware. Needless to say, Japanese
crowds weren't impressed with the advent of basic 3D & Clamored for better
looking 3D. So naturally SEGA panicked & made key emergency design changes in Aurora. First, SEGA had to trash the Model 1 based VDP. Realizing that incorporating Model 2 hardware would have been a bit expensive & slightly time consuming since Model 2 used a state of the art high resolution single core VDP, SEGA decided that the solution was to replace it with 2 low res VDPs. The actual so called "Off The Self Conponets" are two VDPs that split the Model 2's power in half. But when combined together, run full Model 2 which is superior to PSX despite lack of key 3D Engine features. But as you said, Model 2 had plenty of Loopholes around that.


The finish touch design for Aurora was that as a cost cutting measure, SEGA finally scrapped the cartridge based Jupiter and went for CD ROM. When all the changes in Aurora were finalized, the chipset was officially named "Saturn".

The Model 1 board was a commercial and critical failure. In fact it bombed so bad that SEGA retired it in less then a year. And here's a FACT: Most of the early Model 2 games like Daytona USA,VF2,SEGA Rally,Virtua.Cop were originally developed for Model 1.

Well, although PS1 wasn't a household name anywhere until its launch, legend (and by legend, I mean Next Gen Magazine) has it that word got out to Sega concerning the more powerful Sony console from a mutual 3rd party developer, Sega realized their single SH2 based Saturn could not come close to the PS1 and its CPU with built-in geometry co-processor. And that was when Sega went back to the drawing board.

This story has been documented for years:

http://www.giantbomb.com/saturn/60-42/

Rumour has it that the Saturn was originally designed to have a single main processor, but after learning the capabilities of the new Playstation, the Sega Team added a new processor to increase performance. Part of the reason for the system's failure was the complex graphics hardware, which was often difficult to work with. Many ports that came to the Saturn were downgraded and completely rewritten. As a result, third-party support was very thin for the Sega Saturn.

Whether this legend is fact or fiction, it is puzzling why Sega would choose a non-parallel dual SH2 design, when even Yu Suzuki went on record as stating that only 1-2 percent of programmers could ever harness the power of both CPUs simultaneously while even average coders could achieve astonishing results on the Playstation.

Why would Sega knowingly put themselves in that unenviable position on the eve of their biggest challenge, post-genesis?

“ One very fast central processor would be preferable. I don't think all programmers have the ability to program two CPUs—most can only get about one-and-a-half times the speed you can get from one SH-2. I think that only 1 in 100 programmers are good enough to get this kind of speed [nearly double] out of the Saturn. ”

—Yu Suzuki reflecting upon Saturn Virtua Fighter development.

@CruSega. While this legend is widly popluar, there's really no proof that its true at all. Since Saturn's development was top secret, & patents for it remain protected, I'd say this is just an old wives' tale.

In 1993, Sony selling a CD based system was the furthest thing from SEGA's mind. The contenders they knew were Nintendo,3DO & NEC. Since these were 2D known competitor, Saturn was originally designed for mostly 2D. Trust me, if they had found about PSX early on, Sega would have gone through the trouble of scrapping the chipset altogether.

CruSega wrote:“ One very fast central processor would be preferable. I don't think all programmers have the ability to program two CPUs—most can only get about one-and-a-half times the speed you can get from one SH-2. I think that only 1 in 100 programmers are good enough to get this kind of speed [nearly double] out of the Saturn. ”

—Yu Suzuki reflecting upon Saturn Virtua Fighter development.

Suzuki is further explaining the complex design of Saturn's 32-bit Super H CPUs.

The 2 CPUs while 32-bit at heart, are designed to run System 32 hardware on their own were also programmed to code Both Model 1-2 Engines. The VDPs could run Model 1 Engine Graphics with much smoother,better preformance by themselves seperately hergo "Low Resolution Model 2". But for full Model 2, both were needed to run. The reason why Model 2 games were so hard to code on Saturn because the system itself only used half of its Internal Memory using 2MBs instead of 4.

SEGA didn't discover the solution to these problems until Later in the Saturn's life. Not only with the CRX Arcade machine, but that in order to give the VDPs more power an expansion pac was needed to help add extra memory & enhancements to the Engines.

Most Saturn games could have been sold with a backup RAM cartridge ulitlized with converter chips. This would have been cheap & very easy to do. The chips which would have used CRX converters which would have given the
proccessors more features.

Bottom Line: SEGA should have used Model 2 RIGHT from the start. Had they done so, the system would have been much easier to program & Saturn's fate overseas would have turned out different.





Here's the entire Model 1-2 series.





http://www.system16.com/hardware.php?id=712 As you can clearly, see Model 1 was very primatitive. No wonder why SEGA scrapped it.


http://www.system16.com/hardware.php?id=713 Model 2.
Main CPU : Intel i960-KB @ 25 MHz 32bits RISC
Graphics Co-Processor : Fujitsu TGP MB86234 FPU 32bits 16M flops
Co-Processor Abilities : Floating decimal point operation function, Axis rotation operation function, 3D matrix operation function
Sound CPU : 16bits 68000 @ 10Mhz
Sound chip : 2 x Custom 28 channel PCM chips, 1 for Music and 1 for Effects (Can access up to 8meg sample rom *per chip*)
Sound Timing Chip : YM3834 @ 8MHz (only used for its timers)
Audio RAM : 540 Kilobytes (4 megabits)
Video resolution : 496x384 in 65536 colors
Geometry : 300,000 polygons/s. 900,000 vectors/s
Rendering : 1,200,000 pixels/s
Video : Shading Flat Shading, Perspective Texture, Micro Texture, Multi Window, Diffuse Reflection, Specula Reflection.

Much more High Res and State of The Art than Saturn's design.


http://www.system16.com/hardware.php?id=711 Saturn and its specs:

Main CPUs : 2 x Hitachi SH-2 @ 28.6364 MHz
Sound CPU : MC68000 @ 11.45456 MHz
Sound chip : SCSP/YMF292-F (315-5687)/"LAKE" @ 11.3MHz, 32 PCM (Pulse Code Modulation) Channels, 44.1 khz Sampling Rate
Secondary CPUs;
SCU DSP : fixed point maths coprocessor, up to 4 parallel instructions.
VDP 1 : 32-bit video display processor : sprite and polygon, dual 256KB frame buffers for rotation and scaling effects, Texture Mapping, Goraud Shading, 512KB cache for textures
VDP 2 : 32-bit background and scroll plane video display processor, 2 Windows for special calculations, transparency, shadowing, background engine, 5 simulataneous scrolling backgrounds, 2 simultaneous rotating playfields, up to 60 frames per second animation
Main RAM : 2 Megabytes (16 megabits)
VRAM : 1.54 Megabytes (12 megabits)
Audio RAM : 512 Kilobytes (4 megabits)
Rendering Speed : 200,000 Texture Mapped Polygons/Second, 500,000 Flat Shaded Polygons/Second
Colours : 24-bit true color graphics, 16. Million Available Colors
Resolution : 320x224, 640x224, and 720x576 horizontal and 240, 448, and 480 vertical
Board composition : Main board + Game cartridge (one or more in the same time)
Game Cartridge Size : Max size = 384mbit (48meg)
Other infos : This board is Jamma and uses a standard resolution monitor

Sega's console developments have really never been top secret. The design process of all of their consoles is pretty well documented; even for systems that weren't released, like the competing Dreamcast boards or the all-in-one 32X.

Hmm they are pretty much the SAME specs as I posted, MrSega and it shows that the Saturn and Model 2 are COMPLETELY DIFFERENT in their design. Oh btw, those "Saturn Specs" you posted are actually for the arcade version of the Saturn, known as the ST-V. Going back to what you falsely claimed about my previous post, I suggest you read it again, I never mentioned Model 3, those specs I posted WERE for Model2.

As for the legend that CruSega referred to, I would of thought that if that story wasn't true then at some point in the last 20 years someone would of corrected it. Speaking personally I'm more inclined to beleive what Edge/Next Gen wrote than what you are claiming, unless you want people to believe that Sega purposely designed the Saturn to be an overly complicated, badly designed mess.
If that were true then it would make the Sega management back then look even more incompetent than if they had simply ordered a quick and poor redesign of the original single CPU and VDP spec of the Saturn, having found out that the Playstation eclipsed their original Saturn spec.

When it comes down to it, the Saturn had lots of potential, but was completely let down by Sega's incompetence in it's design and component integration and dev tools/documentation, also the fact that Sony came up with a far better, more streamlined design for the PS1, Sony also had better dev tools/documentation and actually gave better support to the devs working on Playstation games *shock horror*. This meant that that the Playstation games ended up looking better, even though the Saturn was possibly more capable, however hardly any developers was capable of working on such a complicated mess of a console, like the Saturn is.

Comrade Snarky wrote:Sega's console developments have really never been top secret. The design process of all of their consoles is pretty well documented; even for systems that weren't released, like the competing Dreamcast boards or the all-in-one 32X.

That's actually not true at all. Both MEGA CD and Saturn were developed under miltary grade top secrecy. No magazine article, report or confirmed information regarding what went on during MEGA CD's development are known to exist. In fact there's was no aknowledgement by the gaming press & public alike that Sega even had a CD attactment in existance until mid 1991. The actual time period of when its development started is uncertain, but most accounts point to 1990. Based on its hardware design, it is believed that the original concept was retooled.


As for for Saturn, Kalinske's CES 1993 statement was the first to confirm the existance of a Genesis/Mega Drive successor. But since information was so vague, nobody knew what its codename was or its bit power. Until its public unveiling in May 1994, there were no reports, no rumors or info of any kind on Saturn.

Instead the public was fixated on "MARS".

To once again clarify any confusion between Model 2 & Saturn, I'll make it simple:

The difference between Model 2 & Saturn is only in design & configuration, not in power.

Model 2 uses a single core proccessor based off Hitachi Super H SH2. There's no difference between this 32-bit Intel proccessor & Saturn. Except Model 2 uses more memory and RAM.

Model 2 has one single Lockheed Martin processor. Which is eqquiped with High Res 3D polygons & and is programmed to run several instructions. Saturn's VDPs split that power by following multiple instructions. Its a fairly complicated mechanism called M.I.P.S. "Million Instructions Per Second".

BTW, I have finally found the "32X" trademark doc. Kabshiki Kaisha Sega is NOT listed as the applicant. Definative proof that "MARS/32X" was an AMERICAN only project not Japanese.

Specs aside, Saturn Daytona USA was horrendous, ran at below 25 fps and had a draw distance of about 2 feet in front of the driver. Even the immediate road ahead would pop up sometimes up. Although Model 2 Daytona had a degree of pop-up it was nowhere close to as bad as the home version and ran at a brisk 60 frames per second. Why is that, Mr. Sega? If the Saturn VDPs were truly capable of performing the same chores as the Model 2 chips, shouldn't this translation (and it was one of the most important Saturn launch titles) have been closer to the arcade version?

Meanwhile the PS1 version of Ridge Racer was nearly flawless which made Sega Saturn look even less appealing to early adopters.