SEGA JUPITER AND OTHER PROTOTYPES

Hi people,

I´m trying to get the most confident information about the sega console prototypes between Genesis and the Saturn final version that was released at the end of 1994.

As you know, there´re plenty of online staff about it and it´s a matter of discussion in forums. I´ve found lots and lots of possible descriptions for the "32 bit generation" Sega console. So in order to start the research, I´ve looking for references about Jupiter, Mars, GigaDrive in magazines.

I´ve been lucky with GigaDrive(mentioned in EGM 16) but not with Jupiter. The only thing I have is a text which is supposed to come from (or be based on) issues 5 and 6 of EDGE:

SEGA JUPITER SPECIFICATIONS
---------------------------

The specifications of Jupiter are virtually the same as Saturn, but
with around half the internal RAM

Format: Cartridge
Memory: 16 Megabits RAM - the internal memory allocation has
not yet been confirmed
Price: Yen 30,000
Released: November 1994 [Japan]

SATURN & JUPITER (SEGA'S BRAVE NEW WORLD)
-----------------------------------------

The fever pitch surrounding Sega's new 32bit system has just tapped
another dimension. Following detailed reports in Edge five and six,
we can now reveal that not one, but two consoles will be released by
Sega in November this year.

The Saturn project was expected to be a CD-ROM based system with the
possibility of a cartridge port. Not only have Sega decided to
include the port, but a second, cartridge only, machine codenamed
`Jupiter' will also appear.

Only in the last weeks have Sega released details about Saturn to the
Japanese press, but at the time of writing, project Jupiter has still
not been offically announced. Complete compatibility between the two
systems will be possible with the release of a low-cost add-on for
Jupiter which will provide the same double-speed CD ROM drive, MPEG
chips and extra RAM.

But why have Sega gone for this approach instead of the fully fledged
CD machine? A beefed-up spec could be one reason - Sega are reputed
to have increased the polygon rendering abilities of Saturn to try
and match the power of Sony's more powerful PS-X.

A cartridge based system allows them to bring out a more affordable
machine - yen 30,000 ( 185) as opposed to Saturn's price of yen
50,000 ( 310) - while still retaining the same power. Both systems
will rely on the same seven-processor architecture, including two
ultra-fast Hitachi SH2s plus 24bit digital signal processing.

Rumours of an American launch just before Christmas have already
started circulating in the US, although that will mostly likely depend
on the strength of the Genesis market and the quality and number of
32bit games ready for release. Also, Microsoft's involvement with
Saturn's operating system has delayed the arrival of non-Japanese
development kits for another two months.


I dont have access now to the EDGE issues, but does this information suit your knwledge about Jupiter?

Thanks

<Anthaemia.> on my forums has been researching material for a book on this era through the Saturn's production life, he has a lot of ex-industry contacts and masses of documentation, may be worth getting in touch.

(Yes I'm sorry for plugging my own forum again)

http://www.segasaturn.co.uk/

Quote:

---

Originally Posted by mick_aka

<Anthaemia.> on my forums has been researching material for a book on this era through the Saturn's production life, he has a lot of ex-industry contacts and masses of documentation, may be worth getting in touch.

(Yes I'm sorry for plugging my own forum again)

http://www.segasaturn.co.uk/

---

Thanks for the information! I wil take it on account and also will check your forum.

There´re lots of speculations about the design of Saturn. The clearest facts in my opinion are as follows:

-The final Saturn version that was released differs a lot from the early designs, in direct response to PSX capabilities. Few months before Saturn launch, several changes were introduced in order to compete with Sony´s 32 bit console.

- Fron 1990 to 1993, Sega projected a 32 bit cartridge-based console. But it had different incarnations as probably Sega Japan and Sega America were working in their own prototypes:

GIGA DRIVE: based on System 32. First mentioned in 1991

JUPITER: maybe it was a Genesis update or a cartridge-only Saturn version.
There are also some reports about talks between SGI and Sega America

¿When Jupiter or GigaDrive were cancelled in favour of the Saturn we know? ¿When the twin hitachi processors were added?

Probably Sega focused first on making a poweful 2D console and changed to ·3D because of PSX.

There are some rumours about that Jupiter was also based on arcade Model 1 (and even Model 2) but I´ve not found any report about it

I will go on searching.

You've unlikely to get 100% concrete info on any of this unless anyone actually has recordings of conversations or copies of emails they obtained the original information from.

That's why I'm not going to go into this myself.

Sam Pettus's article makes a mention of it here:
http://www.sega-16.com/feature_page....%20a%20Failure

Quote:

---

It has been intimated over the passing years that Project Mars shared similarities in design with the system that would later be publicized as Project Jupiter, the abandoned nextgen 32-bit cart-only console supposedly conceived by Sega of Japan alone and carefully kept under wraps from Sega of America until it was eventually abandoned. This would have meant that carts designed for use with Mars would have also worked in a Jupiter console, and there was also talk about a Jupiter-inspired cart adaptor for Sega's other 32-bit console, which was even then already in development. Back-compatability with currently owned games is the dream of many home videogame players, and several reports about Project Jupiter were somehow leaked to the public. While Sega did hint at times that a 16-bit path to its up-and-coming 32-bit technology was possible, they never actually committed themselves to this idea except for the 32X. As it turned out, what was reported by many gaming magazines at the time as Project Jupiter was nothing more than the original conception for Project Mars - a dedicated 32-bit cartridge console based on Genesis technology.. Also, as we now know, Sega of Japan had by this point already decided to make a total break from carts and was doing so with Project Saturn - the direct result of its experimentation with Sega CD.

---

However, I'm not really sure that make sense and particularly conflicts with the idea that the "Jupiter" was directly related to and compatible with the Saturn hardware.
And the fact that we have next to no idea of what SoJ's actual Mars design was in January of 1994, there's even less to go by.

Another interesting point is that apparently the names of several projects hadn't actually been dubbed with planet names aside from Mars and Saturn and several other projects were later publicized with such names (Jupiter, Venus, Neptune) tack-ed on to tie into the myth started by media rumors, at least according to Pettus's article here:
http://www.eidolons-inn.net/tiki-ind...egaBase+Saturn

Quote:

---

Even the console logos for both East and West pay subtle tribute to the planet. Whatever the reason, the fact that they already had a console coming to market named after one planet (Project Mars, aka the 32X) and now were going to release another (Project Saturn) led many industry reporters at the time to conclude that Sega was naming all of its new systems after the planets in the Solar System. Thus was born the myth of the "planet projects," although it must be said in all fairness that Sega went ahead and played along with it. They knew their reputation with both the industry and gaming public was not what it once had been, so the decision was made to accept any free publicity that came along. If that meant perpetuating an unfounded myth, then fine - and with that, other consoles in the "planet series" (Venus, Jupiter, Neptune, Pluto) also made their public appearance in like fashion.

---

We've discussed a hypothetical "jupiter" on the premise of a direct, cut-down Saturn derivative, but that doesn't necessarily have any ties to the historical project publicized as "Jupiter."

Also, that 30k Yen figure would equate to approximately $300 US at the time. (vs approximately $450 for the Saturn launched in Japan at 44,000 yen and 40,000 Yen for the PSX -just over $400)


In terms of a direct Saturn derivative with some RAM and the CD unit cut-out and a planned add-on to supply full upgrade to Saturn spec: I can't see 16 Mbit (2 MB) as particularly feasible for cutting the Saturn back ans still facilitating such an add-on efficiently. It would make the most sense to leave all the Audio/video subsystems alone, thus avoiding any need to add those buses (and many more pins) to such an expansion port, so: what would make the most sense is cutting the CD unit and associated RAM/processors out entirely including the SH1 and 512 kB SDRAM block tied to it and leave the full 512+512+512+2x256 audio and video banks of DRAM and SDRAM (totaling 2 MB/16 kBit), so the remainder would be main RAM and leaving that as the only segment of RAM being added to the expansion unit would make sense (meaning only the main bus would need to be connected to the expansion port, plus analog mixing lines for CDDA). You'd likely have 512 kB of main RAM at the very least, if not a bit more, possible the full 1 MB SDRAM bank (leaving off the slow DRAM), but 512 kB of SDRAM might be reasonable, especially in terms of cost cutting, the dedicated RAM in the subsystems, and the possibility of streaming data -possibly compressed- from ROM directly. (the latter has the primary disadvantage of the cart bus being slow, slower than using the slow DRAM bank in the Saturn, but similar to ROM/RAM carts in the Saturn, so you'd definitely want to optimize use of such -and compression saves space and can even speed things up by being a compressed data stream, but has the disadvantage of eating up CPU resource for decompression, so not a good option in certain cases but preferable in others -a bunch of trade-offs there)

So anyway, in the context of such a system, it would seem likely that you'd have 20 Mbit of onboard RAM at the very least. (you could make cuts to the subsystems, but that would make the expansion interface more complex and expensive as you'd have to have multiple 32-bit buses on it, and you almost certainly wouldn't want to cut back on sound RAM given how onboard synthesis becomes so much more important for a cart based console)

^ I fucking hate Sam Pettus' article. So much.

Quote:

---

Originally Posted by Mafarka

I´m trying to get the most confident information about the sega console prototypes between Genesis and the Saturn final version that was released at the end of 1994.

---

Great! If you're really serious about this, try to get in contact with Hideki Sato (one of Sega of Japan's hardware designers at that time) and Joe Miller (father of the 32X at Sega of America).

These guys should know exactly which hardware variants were discussed at that time.

Why do you hate Pettus's articles, Christuserloeser

It's biased and inaccurate, but the main problem is that it's quoted, referenced and linked to way too often.

Well if you have something better I would love to check it out.

It was written in like 2000. It has inaccuracies, but I think it was the best he could do at the time.

I have a book somewhere called 'developer's marketplace' or something like that written early in '98 that has loads on the development during this period, time to raid the attic perhaps...

Quote:

---

Originally Posted by Christuserloeser

Great! If you're really serious about this, try to get in contact with Hideki Sato (one of Sega of Japan's hardware designers at that time) and Joe Miller (father of the 32X at Sega of America).

These guys should know exactly which hardware variants were discussed at that time.

---

Unfortunately it seems unlikely that either of those things would be possible: Melf has already been trying to do so for quite some time with total lack of interest from Japanese Sega engineers/staff and Joe Miller as well.
In the latter case, Joe Miller agreed to an interview that excluded the 32x as he's so tired of going over that and rehashing details he's already been asked for many times before. (in that case, such information from miller should be out there, but perhaps all such interviews were too dumbed down to be useful -ie no actual nitty gritty tech details on the Mars/32x design -Apparently Sam Pettus interviewed Miller on the subject, but his quotes and paraphrasing leave much to be desired in terms of detail)

I wonder if Pettus still had that interview archived with more details. (unless the original interview lacked any focus on such details)

In particular I really wonder if the Mars design was as limited as Joe Miller's quote makes it out to be as in Pettus's Project Mars article:

Quote:

---

The system that would be known as Project Mars was given birth on 8 January 1994, the night before the opening of the 1994 Winter CES in Las Vegas, Nevada, in a hotel room during a conference among top-level Sega executives from both Japan and America. Those present at this meeting included Sega CEO Hayao Nakayama, Sega of America president Tom Kalinske, his special assistant Joe Miller, Sega of Japan's Hideki Sato, Sega of America's Paul Rioux, and a couple of other Sega of Japan personnel. Surprisingly enough, Nakayama was the one who first broached the subject at this meeting. As such, it is he and not Sega of America's Joe Miller who should be given credit as being "the father of the 32X." Miller remembers this meeting well.

Quite simply, Nakayama-san had directed the company to design and produce a cartridge-based 32-bit platform and bring it to market before the Christmas selling season of 1994. This was a lengthy, somewhat heated meeting - but in the end there was no question that Sega of Japan (in the form of a classic Nakayama mandate) had determined that this was what we were going to do. It was [now] up to the senior team to figure out and go execute. The difference, this time, was that Sega of Japan was actually inviting Sega of America into the process - instead of creating new platforms in a vacuum and throwing them over the ocean at us when it was too late to have meaningful input ....Sega of Japan was completely committed and was [ready to] mobilize whatever internal resources were require to finish the design and produce it in quantity for Christmas.

As first presented by Hideki Sato and his team of engineers, the original concept for Mars was little more than a Genesis with an extra 32-bit processor (a Hitachi SH-1, according to some reports) and an expanded color palette (128 out of 512 possible colors on screen). Joe Miller, who was in fact chief technical wizard at Sega of America, was appalled at the suggestion. "That is a horrible idea," he told them. "If all you're going to do is enhance the system, you should make an add-on. If it's a new system with legitimate software, great. But if the only thing it does is double the colors ...." There was some grumbling about this, but in the end Sega of Japan conceeded the point. They had several other hardware projects in the works, so this one was to be left up to the Americans. Mars was to be Sega of America's baby, although senior management staff from Sega of Japan would be present and oversee it through to production. By the time all was said and done that could be accomplished at that meeting, Nakayama was so excited at the prospect of Project Mars that he wanted its "core senior design team" to leave CES before it had even started and get started working on the new system right away. Miller, Sato, and the rest wound up attending the rest of the show, but went ahead and began the process during a series of late night meetings in Miller's hotel room over the next four days.

---

That is extremely vague, but if at all accurate would imply a very minimalistic upgrade of the Genesis making it not all THAT much better than the Sega CD already did. (if not weaker in some areas) That is if the "128 colors' simply implied doubling the subpalettes from 4 to 8 on the MD VDP and then adding the SH1 (or other coprocessor) on top of that.
Of course, that could also have implied dual VDPs, which would be a bit more significant, especially depending on other modifications. (ie what sort of overlay/pixel combining abilities are added and what the VRAM configuration is -ie if it double buffers and eliminates the DMA limitations, but given the 128/512 colors it seems to imply more of a super grafx like set-up if anything, or single VDP with expanded palette entries)

There's no mention of RAM at all, or sound upgrade for that matter, and the SH1 is onyl speculation, so there's really next to nothing to go on for Mars (or "Super Mega Drive" as it's sometimes referred to).

If Miller did accept an interview, even with no 32x discussion, perhaps he would still be willing to discuss some details of the original Mars Design Nakayma and Sato presented back in January of 1994.


If an interview with Sato was ever forthcoming it would definitely be interesting to know more details on the Saturn's development in particular (in addition to Mars), both in terms of the general design progression of the hardware that ended up as the final Saturn in 1994 (ie if the rumors of the Saturn getting a fairly hefty upgrade redesign starting in late 1993 is true or not) as well as any preceding designs that were abandoned outright. (and if there was ever a design with integral MD+CD compatibility and what that was like, if it existed) Was there ever a project internally referred to as "Giga Drive" or was that only something the media came up with? (or applied retroactively to real hardware, but not the original project designation)

Well, I think Sam Pettus articles can be used as a good reference. Maybe there are some mistakes on them but I think that their data make sense or unless are in the correct way.

I started my little research from magazine reports written between 1990 and 1994. As far as I have read and compared these reports with staff from SegaBase, forums, etc I think that the basic ideas are as follows:

1) Gigadrive was first mentioned in EGM 16 (November 1990) and the magazine suggests that it could be based on System 32 .
There was no more information about Gigadrive in the following time but we can get the idea that Sega was preparing a 2D cartridge-based 32 bit console for the future.

2) Between that time and 1994, there were lots of speculations about "planetary" systems, but all we have clear is that Jupiter was cancelled, Mars evolved to 32X and Saturn was finally launched in a different way that it was conceived because of the unexpected PSX 3D capabilities. ¿Which is the relationship between them?

2.1) Maybe Jupiter was not a true project itself and it was referred to researchs from both Sega America and Sega Japon about a 32 bit cartridge next gen console. I have found two reports about Jupiter:

2.1.1) EDGE Issues 5 and 6 (January 1994) support the idea that Jupiter is a low-cost cartidge-only Saturn (with less RAM) that will be released at the same time as the "complete" cd Saturn.
2.1.2) In EGM July 1994 . an interview with a Sega Product Manager (http://retrogaming.blog128.fc2.com/blog-entry-18.html). He refers to that Jupiter (=low cost Saturn) in this way:

EGM: Can Super 32X carts be used on the Sega Saturn?

OKAMURA: That's not possible. There's no compatibility between the two systems.

EGM: So that naturally means Mega Drive/Genesis software cannot be used with Sega Saturn?

OKAMURA: Yes.

EGM: Does that mean the Jupiter is dead?

OKAMURA: You mean a cartridge-only version of Saturn? No, there's no such thing.

EGM: Even in the future?

OKAMURA: No.

I think that this interview is quite interesting because if you read between lines you can deduce what Jupiter was, and is on agreement with Sam Pettus: Jupiter was the lost 32 bit console that was compatible with MegaDrive and also (with an adaptor) with the Saturn. In some way, it was the nexus between MD/MD32X and SS. So 32X-Mars designed by Sega America evolves from the idea of Jupiter. Saturn was a completely independent CD ROM project of Sega Japan, but the Japanese were still thinking about another console, the Jupiter.

At a certain point, Sega Japan decided to change the Sega Saturn hardware (in order to compete with PSX), abandoned Jupiter project and, as a result, 32X is the only legacy of the Jupiter concept. We will never exactly know when and why Sega Japan took this way. Maybe 32X was designed thinking that Jupiter console would also be released and that Jupiter and Saturn would have back-compatibility. In some way, I think that the lack of backwards compatibility of Saturn is one of the reasons for 32X failure.


But the whole story is even more difficult to understand because we must take on account that Tom Kalinske confirmed in an interview with Sega-16 that Sega America asked SGI for technologica support. At what time in the story? What was the prototype (Jupiter, Saturn, Mars, another one?) Sega America was thinking about?

As some of you have said, an important matter of discussion is the evolution of the design of Saturn. Did really Sega try to make an stepdown Model 1 (or 2) arcade board at any moment? Did Sega rejected the idea because of the high cost involved? All we know is that including Hitachi processors was an order imposed by Nakayama and that Saturn was modified out of time in order to compete with PSX.

Very good suggestion Da_Shocker. I'd love to see Christuserloeser put his money where his mouth is and make a better article.

Quote:

---

Originally Posted by Mafarka

EGM: Does that mean the Jupiter is dead?

OKAMURA: You mean a cartridge-only version of Saturn? No, there's no such thing.


I think that this interview is quite interesting because if you read between lines you can deduce what Jupiter was, and is on agreement with Sam Pettus: Jupiter was the lost 32 bit console that was compatible with MegaDrive and also (with an adaptor) with the Saturn.

---

That's not what he said at all! He said the Jupiter was a Saturn without the CD that used carts. He didn't imply in the least that it was a b/c console.

Quote:

---

Originally Posted by Chilly Willy

That's not what he said at all! He said the Jupiter was a Saturn without the CD that used carts. He didn't imply in the least that it was a b/c console.

---

Yes, you are right; maybe I have misunderstood the whole conversation.

Perhaps the idea of a backwards-compatible Jupiter is more in the words of the interviewer. As Okamura is asked for Saturn compatibility with 32X and Genesis and his answer is "not at all", the interviewer concludes that Jupiter project is no more. But, you are right, Okamura only talks about Jupiter as a cartidge-only Saturn.

This interview is the only staff I have found where someone from Sega at that time talked about Jupiter.

Quote:

---

Originally Posted by Mafarka

1) Gigadrive was first mentioned in EGM 16 (November 1990) and the magazine suggests that it could be based on System 32 .
There was no more information about Gigadrive in the following time but we can get the idea that Sega was preparing a 2D cartridge-based 32 bit console for the future.

---

The System 32 thing seems like total speculation and is generally unlikely: it might have been that SoJ had the concept based around a console capable of System-32 level 2D, but actually being directly related to the System 32 hardware is another thing entirely. (after all, none of Sega's previous home consoles had been direct arcade derivatives and for good reason: the home consoles were far more optimized for low-cost in areas where the arcade hardware was totally impractical for such and while there was often some similarity in terms of CPUs used and such, the actual hardware -especially the custom VDPs- had little to do directly with anything used in the arcade -in terms of general capabilities and use for porting yes, but in actual functionality and hardware design no)

Anyway, Pettus also mentions that the "Gigadrive" was also based around the context of competing with the 3DO, but that is also vague.

Also note that mentions of such a "Gigadrive" or "system 32 like" console in some magazines were often related to the Sega CD not yet announced in North America. (especially anything mentioned in 1990 or 1991) Any such comments in the context of 1992 onwards would be in a different context though.

As it is, there's no really good evidence of Sega's Pre Saturn/32x nextgen designs or how far along they got. That includes the apparent redesign of the Saturn in late 1993 into 1994 that Pettus mentions and has often been rumored. As the story goes: the original "Saturn" was to have been more or less the 1994 Saturn without the SH2s (SH1 acting as the main CPU), no VDP1 (or at least not the 3D functionality), no geometry DSP, and possibly less RAM. (possibly no 2 MB main block and just the 512 kB SDRAM for the SH1 plus the audio and video subsystems -which would mean 2.5 MB rather than 4.5 MB) There's no solid info on that though, so it's just speculation, especially in terms of what the 1993 Saturn actually was. (the evidence for the SH1 being the main CPU is the strongest as some early SoJ press releases -albeit vague- do mention the use of an SH1 as the main CPU, and the fact that the final Saturn used the SH1 for CD-ROM management seems to point heavily to that as well -the SH1 is WAY overkill for its place in the Saturn and it would make sense that it was left over from the previous design to save development time -ie building onto the 1993 hardware rather than building a new CD-ROM interface; the context would be that the SH1 would act as CPU and CD-ROM manager rather like the 68000 in the Sega CD)

Quote:

---

2) Between that time and 1994, there were lots of speculations about "planetary" systems, but all we have clear is that Jupiter was cancelled, Mars evolved to 32X and Saturn was finally launched in a different way that it was conceived because of the unexpected PSX 3D capabilities. ¿Which is the relationship between them?

---

It seems that Mars didn't so much evolve as got halted outright and restarted from scratch by SoA. (the details are vague, but that's definitely the case for the final 32x -it's possible that some add-ons directly based on SoJ's design were also considered -and indeed Pettus's article briefly mentions SH1 and 68000 based systems with different VDPs/coprocessors being considered -unfortunately Pettus is extremely unreliable when it comes to technical details as he often misinterprets when paraphrasing -direct quotes are most useful)

Quote:

---

2.1) Maybe Jupiter was not a true project itself and it was referred to researchs from both Sega America and Sega Japon about a 32 bit cartridge next gen console. I have found two reports about Jupiter:

2.1.1) EDGE Issues 5 and 6 (January 1994) support the idea that Jupiter is a low-cost cartidge-only Saturn (with less RAM) that will be released at the same time as the "complete" cd Saturn.
2.1.2) In EGM July 1994 . an interview with a Sega Product Manager (http://retrogaming.blog128.fc2.com/blog-entry-18.html). He refers to that Jupiter (=low cost Saturn) in this way:

---

It seems that Mars and "Jupiter" got mixed up a fair bit, but Jupiter (or the design apparently labeled as Jupiter) was to have been a direct Saturn derivative without the CD and with some other cuts, but not anything MD compatible.

Mars was to be a direct MD derivative with enhancements. (somewhat like the Neptune in that respect, but the info is extremely vague and I already speculated on possibilities based on Pettus's article above)

Quote:

---

EGM: Does that mean the Jupiter is dead?

OKAMURA: You mean a cartridge-only version of Saturn? No, there's no such thing.

---

Huh, I wonder if that means that there never was such a design: the "no such thing" comment sort of implies that, but it's not definitive as it's not claiming that there never was such.
I seem to recall that the "Jupiter" design was only briefly considered by SoJ on paper but never actually went beyond that.

Quote:

---

I think that this interview is quite interesting because if you read between lines you can deduce what Jupiter was, and is on agreement with Sam Pettus: Jupiter was the lost 32 bit console that was compatible with MegaDrive and also (with an adaptor) with the Saturn. In some way, it was the nexus between MD/MD32X and SS. So 32X-Mars designed by Sega America evolves from the idea of Jupiter. Saturn was a completely independent CD ROM project of Sega Japan, but the Japanese were still thinking about another console, the Jupiter.

---

No, I don't think so. There's a lot of speculation and a ton of mixed info on Mars and "Jupiter" and anything based on media speculation should be taken with more than a grain of salt.

I haven't even seen any direct information about the "Giga Drive" outside of vague statements in magazine articles which may have ended up relating to the Sega CD and nothing more.
For that matter, the only direct information on ANY MD/MCD compatible home console is in regards to project Mars and it seems in parallel with that (ie 1993) Sega had already been pushing the Saturn or at least the early derivative of that (with the added features supposedly added in a late 1993/1994 redesign as I said above) and that Saturn almost certainly was not MD compatible, or at very least the final Saturn shows no signs of ever facilitating compatibility. (even the 68000 used is incompatible with that of the MCD or MD)

Some have speculated that SoJ's Mars design was a direct derivative of some earlier planned MD compatible successor, and that may be possible. However, it seems quite possible that Mars was designed with the express purpose of being an interim console for 1994, especially if it's true that SoJ had originally planned to release the Saturn much sooner and Mars was a result of the delays corresponding to the redesign initiated in late 1993.
It's not even clear whether SoJ came to SoA with hardware already prototyped (or in development of such) or one that was only on paper. (possibly on paper with extensive documentation of the hardware/logic design, but not the actual TTL prototyping, let alone chip design)

Quote:

---

At a certain point, Sega Japan decided to change the Sega Saturn hardware (in order to compete with PSX), abandoned Jupiter project and, as a result, 32X is the only legacy of the Jupiter concept. We will never exactly know when and why Sega Japan took this way. Maybe 32X was designed thinking that Jupiter console would also be released and that Jupiter and Saturn would have back-compatibility. In some way, I think that the lack of backwards compatibility of Saturn is one of the reasons for 32X failure.

---

Again, I don't think that's the case, at least not from anything I've seen or read.
The Saturn as it was in 1993 was possibly as I described above (based around the SH1), and there may very well have been a significant redesign (added 3D capabilities and the SH2s, possibly more RAM -which would explain the bugs and poor documentation of some of the 3D hardware), but that doesn't mean the preceding design was going to be MD compatible. (there's no sign of such in the final Saturn at all, unlike something like the PC-FX where the video and sound hardware obviously imply that PCE/CD/SGX compatibility had been planned)

In terms of backwards compatibility: it's rarely if ever been a critical factor for success of a home console. (maybe for some handhelds it was critical, and definitely computers, but the cases of it being most critical on home consoles was never in conjunction with massive success -like the 7800's 2600 compatibility being a saving grace with its limited library, but it's mainly a bonus for the really prominent consoles -it didn't make or break the SMS, MD, or PS2 and lack of it didn't break the 5200, SNES, N64, PC-FX, PS3, or 360 -in the cases of problems with the latter group the biggest issues were not tied to compatibility directly but other bigger issues -the 5200 had much bigger problems in general including many non-hardware related ones, but the issue of compatibility was most significant due to the introduction of a 2600 adapter on the ColecoVision)
It's a nice feature to have, but rarely necessary in that context. We discussed this before:
http://www.sega-16.com/forum/showthread.php?t=10271
http://www.sega-16.com/forum/showthread.php?t=11345

If Jupiter would have exited at all, it seems to have been intended as a low-cost counterpart to the full Saturn and feature provisions for an upgrade to Saturn spec, but no MD/MCD compatibility. (it seems possible that the suggestion for such a cut-down console may have come after the late 1993/94 redesign efforts were underway: possibly a direct response to the added cost of such, but if that's true it still seems like the idea was dropped without much serious progress)
The Mars design may have been related to the delay of the Saturn as speculated, and indeed as a direct response to new competition from the Jaguar and 3DO (or at least perceived potential competition at the time) as well as an attempt to address the slump in revenue appearing in 1993 and deepening in 1994. (the same thing Nintendo addressed with some hefty software development projects initiated in 1993 and possibly increased emphasis on use of on-cart coprocessors)

In the context of short term impact, the 32x (or possible SoJ's Mars/Super MD) likely was more significant than Jupiter could have been if released in 1994 instead. (the price would have been significantly higher, albeit not compared to new Sega users -ie not much if any more of the cost of a Genesis+32x- but for staying competitive with the huge existing Genesis userbase and keeping revenue up in the short term)
In terms of the Saturn and general long-term outlook as well as early entrance to the 5th gen market in general: the Jupiter as such would almost certainly have made more sense, but it seems that the Jupiter was never even on the table in the context of what Mars was planned as. (it may not have even been presented to SoA before it had been canceled)

In the context of Mars (and 32x), that does indeed make intercompatibility more significant (due to it being a parallel new console rather than an old/aging system), but that wasn't the case. (or at least, I haven't seen any direct evidence of plans for the Saturn to include MD/MCD compatibility, let alone Mars/32x compatibility)

But there were bigger problems with the 32x/Saturn in general: probably most important being the screwed-up launch of the Saturn in May 1995. I'm not sure who's authority that was orchestrated under, but whatever the case it was a disorganized mess which frustrated and confused key 3rd party developers and major retailers who'd been left out of the loop of Sega's plans and also opened the door for Sony to steal Sega's thunder with the $300 announcement.
Had Sega kept to the September release schedule, not only would it have lessened the clash with the 32x (giving it almost a full year on the market before the Saturn was released), but would also have avoided the issues with distributors and developers (and avoided certain rushed games released prior to September), and it would also have meant that Sega would have been launching the Saturn full force at the same price point as the PSX with a bigger library of games than the PSX at about the same time as the PSX with full market hype behind it.
That was probably the most critical mistake of the Saturn.

Quote:

---

But the whole story is even more difficult to understand because we must take on account that Tom Kalinske confirmed in an interview with Sega-16 that Sega America asked SGI for technologica support. At what time in the story? What was the prototype (Jupiter, Saturn, Mars, another one?) Sega America was thinking about?

---

No, Sega didn't ask SGI for anything. SGI came to Sega with the proposal for a chipset (then in development) for use on a next-generation video game console.
That would have had to have been early 1993 and as the interview says SoA was very interested but SoJ engineers were skeptical. After modifications to the chip made to address SoJ engineers' criticism, they presented it again, but SoJ still found it unfavorable: supposedly due to cost effectiveness. (it may also have been due to concerns of how much more development time was needed among other things -and of course, that would have been in the context of what SoJ had planned in early 1993, not any changes made starting in late 1993)

And, of course, after Kalinske had to reject SGI's offer, SGI moved on to Nintendo who partnered with SGI in August 1993. (and made a public announcement regarding Project Reality that October -I believe Sony's PSX announcement was in November)

Quote:

---

As some of you have said, an important matter of discussion is the evolution of the design of Saturn. Did really Sega try to make an stepdown Model 1 (or 2) arcade board at any moment? Did Sega rejected the idea because of the high cost involved? All we know is that including Hitachi processors was an order imposed by Nakayama and that Saturn was modified out of time in order to compete with PSX.

---

No, Sega almost certainly did NOT plan to use any arcade hardware directly. I think the misconception comes from descriptions or design concept goals of being capable of System 32 level 2D or better (which the Saturn certainly did), and similar concept descriptions of model 1 level 3D (or at least capability of reasonable ports of model 1 games).
Generally that concept seems to have been applied in the same context as the "early Saturn" (ie before redesign in late 1993), and indeed that would fit rather well with the speculated design of the "early Saturn." (ie single SH1 handling software rendered 3D as well as 2D "sprites" and VDP2 handling 2D background rendering rather like the PCFX minus the PCE based graphics hardware -ie just the CPU rendering to background layers; it may also have been possible that it wasn't just VDP2, but a more limited precursor to VDP1 was there too -possibly more like the ASIC in the Sega CD- without the hardware quad warping, shading, etc -that would also explain why transparency doesn't work for warped quads but fine for flat quads -ie the warping feature was a late addition and didn't get the bugs fully ironed out)

Quote:

---

Originally Posted by Chilly Willy

That's not what he said at all! He said the Jupiter was a Saturn without the CD that used carts. He didn't imply in the least that it was a b/c console.

---

He didn't even seem to confirm the existence of the "Jupiter" (or cart based Saturn) at all. He didn't expressly deny it, but the "no such thing" comment seems to imply that it may not have ever existed. (but that's reading too much into it really, and "no such thing" could simply mean that it hadn't been produced, or possibly hadn't ever left the drawing board, but still may have been a real consideration at some point)

I do wonder if such a system was proposed (assuming it was proposed) after the Saturn's supposed redesign, or before. (it would make sense in the context of the post-redesign console in particular due to the added cost of the added hardware)

Some other stuff that came up on this before:
http://sega-16.com/forum/showthread....055#post277055

Quote:

---

Originally Posted by Chilly Willy

I don't think there's any "evidence" that SEGA beefed up the Saturn when they heard the specs on the PSX, but the story that circulated amongst people in the business was that the original Saturn had an SH1 processor as the main CPU, with 1MB of RAM, and the VDP2 (the 2D VDP) with 512KB of vram, and the 68000/soundchip. When they finally saw the PSX specs, they realized the Saturn would have its ass handed to it and quickly tossed a few more parts on it: the SH1 became relegated to just handling the CDROM, two SH2s with another 1MB of faster ram was added, and VDP1 (the "3D" sprite rendering VDP) was added along with more vram. That DOES explain why there are two 1MB blocks of unequal speed RAM in the Saturn, as well as why they are two VDPs hooked to three blocks of VRAM.

---

And later this popped up:
http://sega-16.com/forum/showthread....512#post279512

Quote:

---

Originally Posted by Team Andromeda

Yes it was around November time, covered in Edge issue 4 1994. [typo, should say 1993] Long after the 1st Saturn tech spec's were leaking out . In Edge Issue 5 the Saturn spec's were as follows (being brief)

Hitachi: SH7032 CPU RISC CHIP Running at 27 Mhz
Memory: 3Mb Ram
Sound: 32 channels, support for PCM and FM
Relase: November 94

---

And as it turns out:
http://sega-16.com/forum/showthread....534#post279534

Quote:

---

Originally Posted by Chilly Willy

the SH7032 is just an SH1 with no rom and 8KB of ram (among other things - the 7034 was an SH1 with rom). Those specs only boost my story - the Saturn was originally just an SH1 with less memory and only the 2D VDP. Note the date... Nov '94 just shortly before the PSX came out in Japan. That spooked SEGA something fierce, so they tossed on two SH2's with another 1MB of faster memory and the 3D VDP.

---

Note: again, the date on the former post was supposed to be November 1993, not '94, so the last comment about it being just before the PSX launched in Japan obviously doesn't make sense. (the Saturn was already released in Japan by then anyway ;))
However, November of 1993 is the time when the PSX was officially announced and tech specs were publicly released, so that would be a better argument.

And then you have a bit more discussion in general among those pages in that thread, but particularly these:
http://sega-16.com/forum/showthread....791#post279791
http://sega-16.com/forum/showthread....853#post279853
(take a look for your self: I'm not going to cut those up and I don't want to put the entire posts on here... but the point about the possible "plain sprite generator" or more primitive VDP1 is significant among some other things)

In fact, if the system had been configured with 1 MB main RAM for the SH1, 2x256 kB framebuffers 512 kB for "sprite" data (equivalent to the texture bank in the Saturn), 512 kB for VDP2 tile data, and 512 kB for Audio: you'd have 3 MB as in the spec in the Edge article.

I was pretty sure than in it's original spec, what became the Saturn had an NEC CPU, sadly until I have properly moved in and unpacked all my effects I can't search for a reference on that.

I'm also very sure that the 68000 was intended for sound CPU from day one, but as I say I'll have to confirm and provide references for these...

Quote:

---

Originally Posted by mick_aka

I was pretty sure than in it's original spec, what became the Saturn had an NEC CPU, sadly until I have properly moved in and unpacked all my effects I can't search for a reference on that.

I'm also very sure that the 68000 was intended for sound CPU from day one, but as I say I'll have to confirm and provide references for these...

---

I didn't argue anything about the 68EC000 not being present in the earlier Saturn, in fact I implied it was still there. (the sound system, VDP2, and SH1 were all there -supposedly- and possibly a more primitive precursor to VDP1, but not the 3D drawing modes, not the SH2s, and less RAM)

The mention of the NEC CPU, specifically a NEC V60 (or derivative) I believe is complete speculation simply due to the fact that Sega used it in the System 32 and model 1.
It's not a particularly efficient or attractive CPU for the time compared to the embedded RISC CPUs becoming available...
Magazines reported the V60, but I'm nretty sure that's all rumor mill stuff based on the arcade hardware.
Hell, some even refer to the V60 as a RISC CPU (even system-16 museum makes that mistake), it's a mid-80s CISC design.
Most definitely NOT the CPU to use if you're going to be doing a lot of intensive software rendering. (even the SH1 was limited for that) And, of course, Sega's model 2 had moved on to the powerful Intel 960. (that wasn't particularly cheap compared to some of the low-cost emphasized RISC CPUs on the market either -including ARM and MIPS, and of course Hitachi's offerings)

IF Sega was to go with NEC, NEC did have some embedded RISC designs too, particularly the V810 (Virtual Boy and PC-FX), albeit that was more or less in the same performance range as the SH1.

In particular there's this:

Quote:

---

Originally Posted by Team Andromeda

Basic CPU spec's , some of it wrong and some it correct .

Here's EDGE in August 1993

http://farm5.static.flickr.com/4102/...9dd73fc8_b.jpg

And Mean Machines in July 1993

http://farm5.static.flickr.com/4096/...5a9f61e7_b.jpg


There were other mag's too . And sorry in the summer of 1993 none of the press (or most outside SONY) had any idea of the PSX/PS console (not to be confused with Snes CD Rom . This was a time when SONY was seen as SEGA partners, and before the 3DO even shipped

---

Which seems to get some additional information vaguely correct as it mentions Hitachi in context of the CPU, but then says that it's a customized Hitachi derivative of the V60 that runs at a higher clock speed which is WAY off base. (and I don't think Hitachi ever even had anything to do with the V60, not even as a 2nd source manufacturer)

Edge is more vague, but also more accurate due to that as it simply claims a 32-bit RISC CPU, nothing about the V60.
Both list 16k polygons per second, which seems to heavily imply the sort of performance the SH1 would have been doing in software. (ie significantly less than the 32x and 3DO or Jaguar for that matter, but better than SVP)

Of course a later edition of Edge that November lists this:

Quote:

---

Originally Posted by Team Andromeda

Hitachi: SH7032 CPU RISC CHIP Running at 27 Mhz
Memory: 3Mb Ram
Sound: 32 channels, support for PCM and FM
Relase: November 94

---

And that's a fair bit more detailed, though still very vague. The CPU is quite specific though, and as Chilly Willy mentioned, it's an SH-1.

Quote:

---

Originally Posted by Chilly Willy

the SH7032 is just an SH1 with no rom and 8KB of ram (among other things - the 7034 was an SH1 with rom). Those specs only boost my story - the Saturn was originally just an SH1 with less memory and only the 2D VDP.

---

One odd thing though is the 27 MHz figure (or even 25 MHz) as the SH1 was only offered up to 20 MHz in 1993/1994 I believe, or at least this gave me that impression: http://www.hotchips.org/archives/hc6...S4/HC6.4.2.pdf

And here:
http://www.rockbox.org/wiki/pub/Main...cpu_sh1_pb.pdf
http://www.renesas.com/products/mpum...h7032_root.jsp
http://www.renesas.com/products/mpum...30_landing.jsp
http://am.renesas.com/products/mpumc...h7020_root.jsp
http://pdfdata.datasheetsite.com/web/380863/SH7021.pdf
etc

Suprised nobody has mentioned the pre-release prototype that had a 32X shaped cart slot at the back and "High Performance CD-ROM/Cartridge Entertainment System" on the top.

To me that implies game crtridges, but maybe I'm reading too much into it.

Quote:

---

Originally Posted by Da_Shocker

Well if you have something better I would love to check it out.

---

Quote:

---

Originally Posted by j_factor

It was written in like 2000. It has inaccuracies, but I think it was the best he could do at the time.

---

Quote:

---

Originally Posted by Guntz

Very good suggestion Da_Shocker. I'd love to see Christuserloeser put his money where his mouth is and make a better article.

---

Oh, come on. It's not too hard to write something like he did:
"Hayao Nakayama awoke in his room in La Concha, Las Vegas. It was a cold winter morning and he hadn't slept too well. A Norwegian Blue was parroting in the tree right before his window. Hayao sat up and looked around. It was a somewhat empty room with next to no furniture.
He got up and took his stuff from the one chair that stood at the other side of the room, next to the bathroom door.
It would be today when he would meet with Sega of America's Joe Miller and Tom Kalinske. Tom, who had single handedly fought the dragon and saved the company. Hayao was consumed by rage just when he thought about him. It was Tom's success that had been the reason why Hayao could not find rest. Hayao had long been working on an elaborate plan on how he could undermine Tom's success. And it would be today that he would finally carry out that plan. Hayao smiled as he walked towards the bathroom. It would be today that he would sabotage Sega of America by shoving a 32X up Tom's gastrointestinal tract."

Quote:

---

Originally Posted by mick_aka

Suprised nobody has mentioned the pre-release prototype that had a 32X shaped cart slot at the back and "High Performance CD-ROM/Cartridge Entertainment System" on the top.

To me that implies game crtridges, but maybe I'm reading too much into it.

---

Umm, the Saturn was always to have a cart slot... at least given a lot of the descriptions and interviews, and of course the ST-V was cart based and the final Saturn had a cart slot, so I don't see your point.

The only thing with that prototype/mock-up/pre-released unit is a slightly different cart slot shape which implies next to nothing beyond cats being initially planned to be shaped differently. (now if there was a close-up pic of the connector in the slot and it appeared to use a MD/32x like 64-pin connector, that WOULD be significant, but otherwise it's not really much of anything)

It's quite clear from the hardware design (and even reasonably indicative from the 1993 "specs" published) that MD/MCD compatibility wasn't planned. 32x, of course, came after that anyway, though Mars may already have been in development. (perhaps only on paper, but there's no evidence one way or another)

Sega never announced publicly that the Saturn was to be backwards compatible, and even the 1993 magazine publications support that it was a total departure from previous hardware. (see the quote from Team Andromeda above with attached scans -particularly mean machines mentioning it being radically different and totally incompatible with all previous Sega consoles)

And even if there was any plans to make it compatible, none of that remained in the final version released in 1994. There's pretty much zero evidence of direct compatibility being planned judging by the Hardware alone. (again, by start contrast of the PC-FX which includes all the custom audio and video hardware from the PCE and the Super Grafx, but just not the CPU, BIOS, or card slot)

Quote:

---

Originally Posted by Christuserloeser

Oh, come on. It's not too hard to write something like he did:
"Hayao Nakayama awoke in his room in La Concha, Las Vegas. It was a cold winter morning and he hadn't slept too well. A Norwegian Blue was parroting in the tree right before his window. Hayao sat up and looked around. It was a somewhat empty room with next to no furniture.
He got up and took his stuff from the one chair that stood at the other side of the room, next to the bathroom door.
It would be today when he would meet with Sega of America's Joe Miller and Tom Kalinske. Tom, who had single handedly fought the dragon and saved the company. Hayao was consumed by rage just when he thought about him. It was Tom's success that had been the reason why Hayao could not find rest. Hayao had long been working on an elaborate plan on how he could undermine Tom's success. And it would be today that he would finally carry out that plan. Hayao smiled as he walked towards the bathroom. It would be today that he would sabotage Sega of America by shoving a 32X up Tom's gastrointestinal tract."

---

:lawl:

If nothing else you can be pretty sure that the Genesis and Sega CD wouldn't have been nearly as popular in the US if it hadn't been for Katz and Kalinske's management... well assuming you didn't find some other extremely capable, consumer entertainment savvy, experienced businessman to take their place.
Sega of America was pretty weak before Katz came along and they lost the Master System because of that.

Granted the MD likely wouldn't have been too much different in terms of Japanese and perhaps European success, but Sega's North American market was transformed by the actions of Katz and Kalinske.

Management of marketing and advertising is what made the Genesis what it was (just as with the VCS and NES before it) and the key reason the SMS and TG-16 were weak in North America. (albeit not in Japan: the SMS and MD did rather well all things considered -ie the competition- though the screw-up with the weak SG-1000 was significant and the main reason Nintendo got the critical edge early on -granted Sega was a bit of a mess at the time with Rosen with Nakayma and Okawa and CSK just about to buy it back)

Quote:

---

Originally Posted by mick_aka

Suprised nobody has mentioned the pre-release prototype that had a 32X shaped cart slot at the back and "High Performance CD-ROM/Cartridge Entertainment System" on the top.

To me that implies game crtridges, but maybe I'm reading too much into it.

---

Where can I find this prototype? Or could you post an image of it?
I´v never seen a Saturn prototype than a small picture with nothing like "High Permormance..."

Quote:

---

Originally Posted by Mafarka

Where can I find this prototype? Or could you post an image of it?
I´v never seen a Saturn prototype than a small picture with nothing like "High Permormance..."

---

http://i253.photobucket.com/albums/h...aturnproto.jpg

Quote:

---

Originally Posted by mick_aka

http://i253.photobucket.com/albums/h...aturnproto.jpg

---

That slot does NOT look like a 32X slot... it's almost 100% identical to the current Saturn slot... say 99.99% identical. It's MAYBE 70% like the 32X, if that much.

"32X shaped cart slot" was actually a direct quote from an article accompanying that image, I wasn't suggesting that it was one or looked like one, and if you look back in this thread you'll find that I stated that I was quite sure no cart backwards compatibility was ever planned.

The more rounded edges of the slot on the system above, may however account for why many of the very early 3rd party card addons seem to match that shape more than the later more squared edged cart slot.

Although I'm sure that the 3rd party carts that used 32X and MegaDrive casings (early ST-Key for example) were more to do with those casing being very cheap to come by.


Incidentally a contact in Japan I buy a lot of games through reckons that some of those units went into stores in demo kiosks, but I've found no further evidence of that so if anyone has more info I'd appreciate it...

I have some observations, but I'm not going through and quoting everybody. The Edge and Mean Machines Articles are interesting but not definitive in any way. It is interesting that Mean Machines claims the Saturn would have a "V60 based" chip that Sega and Hitachi worked on. I also think it is interesting, but not definitive, that early SNES CD-ROM specs look just like the SFX-2 with more RAM. We all know that the CPU was by far not the only component in any system design.

The V60 was used in both the System 32 and Model 1 to totally different effects. Comparing the 16Mhz V60 to the 20MHz SH1 in the Saturn is tenuous at best. The 25Mhz Intel i960-KB in the Model 2 board sounds much more like the target these articles speculated about.

The dual SH-2's in the Saturn and 32X are the only known Hitachi solution to either the Model 1 or Model 2 hardware, much less the V60 based System 32.

Repeating speculation does one thing, and one thing only, it further muddies the fact pool.

Quote:

---

Originally Posted by sheath

I have some observations, but I'm not going through and quoting everybody. The Edge and Mean Machines Articles are interesting but not definitive in any way. It is interesting that Mean Machines claims the Saturn would have a "V60 based" chip that Sega and Hitachi worked on. I also think it is interesting, but not definitive, that early SNES CD-ROM specs look just like the SFX-2 with more RAM. We all know that the CPU was by far not the only component in any system design.

---

They're vague enough that you can interpret them in a number of ways, but for the SNES CD, I'd bet it was more likely that the NEC 810 CPU was to be used and not the Super FX DSP. (I do wonder if the Super FX DSP is the same or directly related to the RISC DSP in the Flare 1/Slipstream/Multisystem chipset -they were both designed by the same engineer -Ben Cheese- and the original Super FX was at a similar clock rate to the previous 12 MHz DSP too, and it has 512 bytes of on-chip RAM just like the Slipstream DSP, and both are described similarly, but I don't know enough about the hardware and haven't heard from anyone who's looked at the details for both)

Anyway, I think Edge is a lot more reasonable than Mean Machines, granted they kept it simpler with the initial specs, but as such they avoided more inaccuracies: on top of the odd V60 reference, Mean Machines also lists 16,000 polygons on-screen, which is a pointless metric and would never be seriously used as almost any compute/console cable of rendering any polygons could render 16k, it's just a matter of how long it would take. :p (Edge lists it as 16,000 polygons/second, which makes a good bit of sense)

Quote:

---

The V60 was used in both the System 32 and Model 1 to totally different effects. Comparing the 16Mhz V60 to the 20MHz SH1 in the Saturn is tenuous at best. The 25Mhz Intel i960-KB in the Model 2 board sounds much more like the target these articles speculated about.

---

No, they were used to similar effects, but had very different supporting hardware. In those systems, the CPUs weren't doing any rendering, just the game logic and such. (the only difference is that 3D game logic tends to be a bit more intensive in some cases, though it really depends on the case as well as the amount of hardware support)
I mean, you could put a 16 MHz 68000 in the PSX and still have a very powerful system. (the problem there is the lack of hardware Z-buffering, so software Z-sorting would be rather significant, and some more complex games used more resource for game logic too)

It's a totally different case for Software rendering though, and a fairly powerful RISC CPU like the SH1 vs an older, slow CISC processor on a 16-bit bus is a no brainer.
If they couldn't use the SH1, then they'd have gone with a MIPS or ARM design most likely, maybe one of NEC's RISC CPUs depending on pricing.

The i960 is a pretty powerful CPU, but it wasn't used for software rendering either in the model 2, but for the same purpose as the V60 as a general purpose CPU to handle game logic. Hell, Namco's System 21 only used a pair of 68000s for the CPUs and the powerful System 22 used a 68EC020 (less powerful than the 3DO's CPU), but it's the hardware that matters in those cases.

In the case of such an early Saturn, as with the 32x and DOS PCs, CPU resource would be THE rendering grunt, so totally different.
That said, a 25 MHz i960 would have been much more powerful than a SH1 and possibly more powerful than 2 SH2s. (if nothing else it has a FPU which would be very significant for 3D -something that nothing had until the N64 -albeit I'm not sure the N64 had a proper FPU)

Quote:

---

The dual SH-2's in the Saturn and 32X are the only known Hitachi solution to either the Model 1 or Model 2 hardware, much less the V60 based System 32.

---

I'm not sure what you mean there... they're just the CPUs, not the real heart of the arcade/console hardware, except for the 32x as CPU rendering is all it really does. (as with DOS PCs)
So I don't think "hitachi solution" would really apply at all, and prior to the SH2s, the SH1 could easily have been the option. In fact, if Sega was ever planning on releasing a console by late 1993, that would have been the main option among those used. (and really, the SH1 makes very little sense to use purely as a CD-ROM controller, which is probably the strongest evidence that the SH1 was originally meant to be more than that -same as with the Sega CD, the 68k would be way overkill if all it was used for is managing the CD-ROM operations, but in that case it also acts as the general CPU for the MCD and as a coprocessor for the MD)

Quote:

---

Repeating speculation does one thing, and one thing only, it further muddies the fact pool.

---

Yes, but speculation, especially educated speculation is exactly all we have to go on. Hence, I was presenting evidence that made much more sense in the context of said speculation than without, but yes, we have near nothing to go in in reality.

Ie: we have pretty much no idea what the Mars design was really like.
-We don't know if the Saturn was really heavily modified starting in late 1993
-We don't know the development stages of the 32x (though we do know it was developed from scratch in only 5-6 months)
-We don't know who's authority it was that the Saturn be launched in May of 1995
-we don't know absolutely that the Saturn (or precursor to that) was ever intended to be compatible with the MD or MCD. (but the final hardware and interviews seem to point to compatibility not being planned)
-we don't know the exact context for initiating Mars in the first place, or the original intentions for supporting in in parallel with the Saturn. (speculation with some supporting evidence ranges from the market slump to fear of the Jaguar and 3DO to fear of the PSX, to filling a "gap" in the market, to "not being able to get the Saturn out soon enough" or a combination of those)
-We don't know the full reason for Sega's shift in budget in 1996 and related focus on Saturn.
-We don't know why the MD was configured in the manner it was with the backwards compatible components used rather oddly. (though that's likely up to time constraints on the design and perhaps shifting priorities)
-We don't know if Sega had ever seriously considered adding video overlay to the MCD.
-We don't know the context or real design of the project referred to as Jupiter, if it even existed.
We don't know for sure that Sega never considered a direct derivative of their arcade boards, but from all actual hardware, it's clear than that was never done for any console actually released. (or known prototypes)
-we don't know the full context of Kalinske and Rosen leaving in 1996
-we don't know the full details of any internal management conflicts or indeed frustration on the part of SoJ staff.
etc, etc.

I wanted to post this in a new thread (but can't until 50 posts, despite having been here 2½ years :().

Was mooching around on Google Image Search earlier and found this image, seemingly originating from a Spanish forum:

Attachment 6109

Almost certainly a fake (the guy who posted it jokingly claimed to have found it in his basement), but it's a bit of a beauty nonetheless isn't it?

I just wondered if anyone had seen this before?


EDIT: Hmm, just found a slightly clearer version on a French forum. Maybe it was published in a magazine (looks like a scan)?

Attachment 6110

Quote:

---

Originally Posted by Chilly Willy

That slot does NOT look like a 32X slot... it's almost 100% identical to the current Saturn slot... say 99.99% identical. It's MAYBE 70% like the 32X, if that much.

---

If I'm not mistaken, SEGA completed and finalized Saturn's design in March 1994. The first prototype was sent out about a month later. EGM actually leaked the first public info regarding the project codenamed "Aurora" in its December 1993 Issue. I recall the public getting first details about "Mars" in January of 1994. During 1993, the press was reporting on both Nintendo's Reality and SEGA's "upcomming 3DO contender" called "Jupiter".

Jupiter itself is just System 32 with Model 1 components rather, one VDP with similar silicon to Model 1's basic 3D rendering. Jupiter's memory and RAM set up from what I've read is identical to that of System 32 hardware. Because it was planned to be Cartridge only, Jupiter would have also had a set up similar to SNES carts. Added converters and 3D backup chips.

I'm pretty sure Jupiter was NOT related to Saturn. Saturn's earliest design in late 1993 specifically was Cartridge based with the possibility of a CD tray kinda of combination of Genesis/Mega Drive Model #2.

http://trademarks.justia.com/744/69/...-74469937.html


http://trademarks.justia.com/744/95/...-74495742.html

Around here, when someone says "Jupiter" they mean a Saturn without the CD (and associated ram and SH1), and possibly without the low work ram. That may not be what had been called "Jupiter" at various times, but it's what WE mean NOW. It's easier to simply call it the Jupiter as that precedes Saturn. ;) :D

Lest it get lost on the previous page:

Quote:

---

Originally Posted by MY99

I wanted to post this in a new thread (but can't until 50 posts, despite having been here 2½ years :().

Was mooching around on Google Image Search earlier and found this image, seemingly originating from a Spanish forum:

Attachment 6109

Almost certainly a fake (the guy who posted it jokingly claimed to have found it in his basement), but it's a bit of a beauty nonetheless isn't it?

I just wondered if anyone had seen this before?


EDIT: Hmm, just found a slightly clearer version on a French forum. Maybe it was published in a magazine (looks like a scan)?

Attachment 6110

---

Quote:

---

Originally Posted by MY99

I wanted to post this in a new thread (but can't until 50 posts, despite having been here 2½ years :().

Was mooching around on Google Image Search earlier and found this image, seemingly originating from a Spanish forum:

Attachment 6109

Almost certainly a fake (the guy who posted it jokingly claimed to have found it in his basement), but it's a bit of a beauty nonetheless isn't it?

I just wondered if anyone had seen this before?


EDIT: Hmm, just found a slightly clearer version on a French forum. Maybe it was published in a magazine (looks like a scan)?

Attachment 6110

---

That's intriguing, whatever it is, but I'm not even sure what it might be. Is it supposed to be an actual system or just a mock-up case? (if the latter, it could easily be for the Saturn or maybe even a 32xCD system . . . or maybe just a CD-X like system that's labeled as 32-bit for some reason -arguably true in some contexts, but no more so than the MD itself or any 68000 based machine, really)

Quote:

---

Originally Posted by MrSega

If I'm not mistaken, SEGA completed and finalized Saturn's design in March 1994. The first prototype was sent out about a month later. EGM actually leaked the first public info regarding the project codenamed "Aurora" in its December 1993 Issue. I recall the public getting first details about "Mars" in January of 1994. During 1993, the press was reporting on both Nintendo's Reality and SEGA's "upcomming 3DO contender" called "Jupiter".

---

TeamAndromeda (among others) pulled up some old magazine articles addressing "leacked" information on Sega's in progress 1993 designs, and most of those either used the actual name Saturn, or used a more generalized/vague name. (there may have still been some references to Gigadrive even)

In any case, what eventually became the Saturn was almost certainly being worked on in earnest during 1993, and probably 1992 as well. (that would also mesh with SoJ declining to partner with SGI in 1992, prior to going to Nintendo -which became Project Reality, and -ultimately- the N64)

Now, whether or not the platform was definitely CD-ROM based from day 1 is kind of beside the point. The core hardware of the Saturn could work with or without the CD-ROM subsystem, so it's very well possible that it started life as a cart based system and/or was being developed with a corresponding CD-ROM drive/interface but keeping the option open as to whether it would launch purely as a CD-ROM system. (ie could have had a cart system with separate CD add-on or parallel launch of cart, cd-addon, and standalone cd equipped console)
TBH, since we'd only be talking about prototype hardware at this point (1993), it wouldn't really matter yet anyway. You wouldn't actually be building the final console (outside of mock up cases) until the hardware had reached mass production requirements, and alpha dev kits could (and often do) include all sorts of crap that the final release versions lack.

Quote:

---

Jupiter itself is just System 32 with Model 1 components rather, one VDP with similar silicon to Model 1's basic 3D rendering. Jupiter's memory and RAM set up from what I've read is identical to that of System 32 hardware. Because it was planned to be Cartridge only, Jupiter would have also had a set up similar to SNES carts. Added converters and 3D backup chips.

---

That never existed . . . or at least, there's no sensible reason for it to have. This is rumor mill crap, and usually applied to the name "Gigadrive" (recycling the code name surrounding Sega CD rumors).
There's no feasibility in a design based on the Model 1 and System 32 . . . there are some technical similarities to the System 32's sprite hardware and Saturn VDPs, but they're not direct derivatives either. (evidence for related engineering though)
More realistically, Sega may have used the model 1 and System 32 as conceptual starting points (aiming at hardware capable of handling similar game styles) . . .

There's no way the model 1 hardware could have much of anything to do (beyond game design concept) with home console hardware. It was a design based around a bank of fast, expensive FPU coprocessors. (as is the Model 2 -albeit with texture rendering capabilities and a better CPU)

This was typically the case for Sega: design console hardware from the ground up, new, unique, but conceptually similar (in some respects) to existing arcade hardware as well as previous home console hardware (in some cases). Consoles need a much greater emphasis on low-cost, embedded hardware that is efficient to mass produce and scales well as volumes increase and as time progresses. (new revisions and/or dropping component costs)

Quote:

---

I'm pretty sure Jupiter was NOT related to Saturn. Saturn's earliest design in late 1993 specifically was Cartridge based with the possibility of a CD tray kinda of combination of Genesis/Mega Drive Model #2.

---

This depends on context: there's the actual project(s) codenames Jupiter internally at Sega, and projects (or rumored projects) labeled as "Jupiter" in the media or word of mouth discussions. The latter certainly includes a "cart only Saturn" and was so significantly noted as such, that Sega staff recognized "Jupiter" as meaning cart-based Saturn derivative. (whether that was actually a thing internally as well, or just a response to rumor mill labels is the question)
"Jupiter" also is sometimes used to refer to the initial proposition that resulted in the 32x in early 1994. (another totally different design from the Saturn)

Also, I'm pretty sure the Saturn was being designed MUCH earlier than 1993. A design that complex needs well over a year to reach mass production . . . several years really, depending exactly how resources were distributed. (it would make sense that they started with several different projects on paper before refining them into the definitive one -at least as far as the chipset design features- and then move forward with the actual chip design/prototyping/simulation followed by first silicon and trouble shooting -and additional fixes/revisions if necessary)

At a guess, the Saturn design was probably established around 1992 (maybe even earlier), with the basic design of the chipset/memory mapping (at least for the core custom hardware) laid out on paper. At that point, certain key things (especially design of complex custom chips) should have been locked down on paper, and actual design of the silicon, TTL wirewrap, and programming of a simulator would get under way. Actual prototyping work should have been well under way during 1993, and I'd have expected the first revision custom chips to appear by some time that year at least (followed by testing and bug fixing).
In such a design, provisions for certain areas of the hardware to be left open and semi-modular for alternate off the shelf (or semi-custom) parts to be used (like the CPU and sound subsystem, perhaps the CD-ROM subsystem), so changes to those areas could be made more easily and be left fluid until late in the design cycle before being frozen for mass production.

This last part is why claims of Sega changing the CPU arrangement "at the last minute" (namely the winter 1993/1994) could very well be true, so could things like changes to the amounts of RAM present and such. (the slow lo-RAM added to main, perhaps) It's probably limited to only using a limited selection of very similar CPUs, so the inferences some have made that the system was using an SH1 at one point is quite possible, as would be a single SH2 (could have both been tried). In fact, it's likely that they very least used the SH1 for earlier prototypes since the SH2 didn't even reach first silicon until October 1993, so usable engineering samples wouldn't have been available until some point thereafter. (the SH1 reached first silicon in September 1992, so would have been available much sooner)

There's even magazine articles specifically mentioning the SH1, specifically Edge issue 4, Novermber 1993 listing this:

Quote:

---

Hitachi: SH7032 CPU RISC CHIP Running at 27 Mhz
Memory: 3MB Ram
Sound: 32 channels, support for PCM and FM
Relase: November 94

---

The SH7032 is an SH1 with no on-chip ROM and 8 kB scratchpad RAM. (since no SH1 offered an actual cache, and the ROM was used for microcontroller applications, this probably would have been Sega's best SuperH option for a main CPU prior to the SH2)
This came up here: http://www.sega-16.com/forum/showthr...2X-hype/page12

Whether or not Sega was serious about using the SH1 at that point, or whether that was just for prototyping is unclear. But if the claims of the "last minute" changes are true, an upgrade to dual SH2s from the single SH1 would make sense in that context.

Quote:

---

Originally Posted by kool kitty89

That's intriguing, whatever it is, but I'm not even sure what it might be. Is it supposed to be an actual system or just a mock-up case? (if the latter, it could easily be for the Saturn or maybe even a 32xCD system . . . or maybe just a CD-X like system that's labeled as 32-bit for some reason -arguably true in some contexts, but no more so than the MD itself or any 68000 based machine, really)

---

I think the picture is either gonna be a old concept pic that was published in a magazine (and not a widely-circulated one), or it's some japester's attempt to create something that looks like that.

What's great about it is, design-wise it looks like a proper Sega console. If Sega released 'a potential Gigadrive' around 92/93, I could easily imagine it looking like that.

In a fantasy land, perhaps you could've paid something close to Neo-Geo prices for one, in order to settle down and play arcade-perfect Galaxy Force II :)

Quote:

---

In any case, what eventually became the Saturn was almost certainly being worked on in earnest during 1993, and probably 1992 as well

---

Yep confirmed by SEGA then boss Tom and also a small interview with the Thunderhawk Team at Core - how confirmed that had seen early desgins spec's of the Saturn in early 1993

Quote:

---

This last part is why claims of Sega changing the CPU arrangement "at the last minute"

---

It wasn't last min the CPU the SH-2 was chosen very early in and SEGA and Hitachi went public about it, the big last min change was having a dual SH-2 set up - which weren't in early Tech leaks

Quote:

---

Originally Posted by Team Andromeda

It wasn't last min the CPU the SH-2 was chosen very early in and SEGA and Hitachi went public about it, the big last min change was having a dual SH-2 set up - which weren't in early Tech leaks

---

What early tech leaks? The earliest actual documents I have date to early 1994 (march or so).

Quote:

---

Originally Posted by zyrobs

What early tech leaks? The earliest actual documents I have date to early 1994 (march or so).

---

Mean Machines and Edge went public with early system leaks of Hitachi CPU running at 27mhz and those spec's on in the launch issue of EDGE way back in August of 1993 and in December of 1993 EDGE again did a update with more spec's including the soundboard

Quote:

---

Originally Posted by kool kitty89

Unfortunately it seems unlikely that either of those things would be possible: Melf has already been trying to do so for quite some time with total lack of interest from Japanese Sega engineers/staff and Joe Miller as well.
In the latter case, Joe Miller agreed to an interview that excluded the 32x as he's so tired of going over that and rehashing details he's already been asked for many times before. (in that case, such information from miller should be out there, but perhaps all such interviews were too dumbed down to be useful -ie no actual nitty gritty tech details on the Mars/32x design -Apparently Sam Pettus interviewed Miller on the subject, but his quotes and paraphrasing leave much to be desired in terms of detail)

---

Actually, I spoke to Miller and we discussed the 32X in detail. He clears up the story about the initial meeting and dispels myths about who was responsible for what.

Personally, I'm highly skeptical that Pettus interviewed Miller at all. Judging by the shortness of his quote, I think he grabbed it from somewhere instead. Miller was very forthcoming with me about a host of topics, including the 32X, and while Pettus' quote is accurate, it's less than half the story.

Quote:

---

In particular I really wonder if the Mars design was as limited as Joe Miller's quote makes it out to be as in Pettus's Project Mars article:

If Miller did accept an interview, even with no 32x discussion, perhaps he would still be willing to discuss some details of the original Mars Design Nakayma and Sato presented back in January of 1994.

---

According to Miller, there weren't even code names at the time they were initially designing the 32X, and that famous design of a Genesis 2 with simply a larger color palette is just one of a bunch of designs that SOA and SOJ people worked on together.

Cool Melf, I'm definitely checking out that interview when its up. :D

Quote:

---

Originally Posted by Team Andromeda

It wasn't last min the CPU the SH-2 was chosen very early in and SEGA and Hitachi went public about it, the big last min change was having a dual SH-2 set up - which weren't in early Tech leaks

---

Hitachi yes, but SH2 no, at least not from any of the info I've seen. In fact (as Chilly Willy noted at one point), the leaked published specs you posted from EDGE are strong evidence that Sega was planning to use an SH1 as the main CPU. (the SH7032 is an SH1 core with 8 kB screatchpad RAM)

Actually I'll just grab this from my above post:

Quote:

---

There's even magazine articles specifically mentioning the SH1, specifically Edge issue 4, Novermber 1993 listing this:

Quote:

---

Hitachi: SH7032 CPU RISC CHIP Running at 27 Mhz
Memory: 3MB Ram
Sound: 32 channels, support for PCM and FM
Relase: November 94

---

The SH7032 is an SH1 with no on-chip ROM and 8 kB scratchpad RAM. (since no SH1 offered an actual cache, and the ROM was used for microcontroller applications, this probably would have been Sega's best SuperH option for a main CPU prior to the SH2)
This came up here: http://www.sega-16.com/forum/showthr...2X-hype/page12

Whether or not Sega was serious about using the SH1 at that point, or whether that was just for prototyping is unclear. But if the claims of the "last minute" changes are true, an upgrade to dual SH2s from the single SH1 would make sense in that context.

---

Quote:

---

Originally Posted by Chilly Willy

Err - the SH7032 is just an SH1 with no rom and 8KB of ram (among other things - the 7034 was an SH1 with rom).

---

Designing a system that could allow any similar CPU type to be used would be perfectly reasonable to do, so any Super H CPU in that speed range should have worked. (so changing from SH1 to SH2 might have been trivial, engineering wise -albeit not trivial in terms of the performance boost)

The SH2 didn't even exist (physically) until late 1993 (first silicon in October, engineering samples some time after that), so the SH1 (first silicon September 1992) would have been the only option prior to that: Sega may have been banking on using the SH2 from the start, but it would make perfect sense to use the SH1 in the interim. (and perhaps as a backup if the SH2 saw excessive delays)


On an interesting note: The SH1 and SH2 actually use similar size dies too and the same .8 micron process. The only added manufacturing cost of the SH2 would be the higher pin count of the package (depending on yields), though the actual price for Sega (or anyone for that matter) should have been inflated by the SH2's general capabilities. (and Hitachi's R&D investment)