Japan execs were upset that Kalinske was allowed to resign w/o taking blame for 32X

[gamevet]

Some may have been, others not so much. Again if the price drops were all reactionary we'd see them all happening after Sony made the drop first. But we don't see that. We see some happening in reaction to Sony, but others happen before Sony drops their price. You can especially see this in Japan. In Japan we see it launch at 44,800 Yen, and PS1 at 39,800 Yen. Saturn drops to 34,800 Yen with Virtua Fighter Remix and 29,800 Yen with just the console around June of 1995. We see Sony then drop to 29,800 in July to match. The next big drop we see in Japan is Sega drops the Saturn to 20,000 Yen in March of 1996 with the release of the Model 2. Sony at this time drops to around 24,800 Yen. Later in July of 1996 Sony drops to 19,800 Yen. For the rest of it's life in Japan Saturn remains at 20,000 Yen, though they still continue to reduce costs on the hardware as we can see from other revisions.

So this isn't entirely all Sega reacting to Sony, you see both reacting to each other. If Sega truly couldn't afford to sell the Saturn at these prices I don't think we'd see them making these price cuts before Sony made them. I'm not saying they'd have the same profit margins as Sony, but that they clearly felt they were able to make those price cuts safely.

It absolutely was. The reason they had that 3 free games deal, is because they had to sell the Saturn for $50 more than the PlayStation. It's been talked about in discussions about Saturn vs Playstation. Sony also marked the PlayStation down to $129 in 1998. At that point, they might as well have brought a tanker full of gasoline to Sega of America's HQ and torched the place, because that's how much damage it would have done to Sega to reach that price.

Sure, having less chips would lower it more, but the changes from VA0 to VA1 are still pretty dramatic internally. I wouldn't be surprised if it was enough to make that drop to $300 safe to do. And then we see from there Sega continues to make changes to get the cost down to allow them to continue to drop the price. In Japan we see once it hits the 20,000 Yen mark that it just stays there for the rest of it's life. This is probably when the hardware stopped being sold at a loss due to the price not dropping but cost reductions still happening.

Yeah if they didn't have to deal with the exchange rate, or import taxes. It's also possible that Sega of America was still holding onto old hardware that wasn't selling.

And no, the changes from VA0 to VA1 weren't really much to talk about. VA15 was where the meaty changes happened, including changing over to the cheap Sanyo CD player. I've heard that those things are garbage.

Yeah I've seen that, and while it does have some good info, there are also some flaws in it. For example for software sales he's using Famitsu charts which don't paint the entire picture, they only show what made the weekly charts. They can show you trends, but they don't say everything. For example he has the attach rate for Saturn at 5:1 using that data, yet we know from Sega's FY98 report that it was higher than that. Sega's own report states they had sold 8.8 Million Saturns and 80 million games. That would give it an attach rate of about 9:1. Considering how bad it performed outside of Japan, we can assume Japan had a high attach rate to bring those numbers up.

He does point out that that 8:1 attach rate might very well include the 3 pack-in games sold with the console here. I know one thing for sure. Best Buy was clearing out unsold software in the spring of 1997. It had quite a lot of PlayStation games too, but the Saturn stuff was being sold for $8 each and I picked up about a dozen games that many would consider the better games for the Saturn. The PlayStation stuff was like a Military Strategy game (I bought the thing, but the name slips my mind, it sucked!) some other goofy crap games that really weren't that great and maybe some sport titles.

That all said, the main point I'm trying to make here is that I think the fear of a $400-$500 Saturn was a bit unwarranted from Sega of America. At least, not enough to warrant them pushing for the 32X idea. If they had simply prepped for a good launch and not done the 32X, they probably could have launched Saturn at around $299-$349 in the fall of 1995 thanks to the cost reductions that had been made up to that point. A good launch at that price would have led to more console sales, and more software sales. Having not done the 32X would have lead to them having a stronger launch line up outside of Japan as well, which would again lead to better first impressions and better sales. Having that momentum and better success outside of Japan would have most likely made the hardware cost issues not as severe as they were.

Yeah, $350 would have been reasonable, but with Sony having their console at $300 and a solid lineup of sports games, they were still better prepped than Sega for that Holiday season. I ended up getting a PlayStation in January of 1996, because I wanted to play Gameday and War Hawk.

[zyrobs]

It would still lower your manufacturing costs. You only have to make one board instead of the 3+ that were in the VA0 Saturn. You can also see other chips were starting to be consolidated down. This is clear in the CD-ROM block and you can see it with RAM chips being consolidated down. Would it save exactly $100? Who knows. But it clearly made it so they could make that kind of a price drop safely. So I'd say if they were launching in the fall with this revision, they probably would have been able to price match the PS1 at launch. That combined with better software would have made a massive difference.

The VA1 streamlined production, the bill of materials was most likely the same as the VA0. The Access LED and Reset daughter boards for the VA0 were part of the main PCB during manufacturing, they were then literally cut off the corners, so they cost no extra. The SH1 sub-board was integrated into the main PCB on the VA1 but those sub-boards were probably something they got stuck with when the whole CD-ROM was still considered an add-on, so they had no chance but to use those on the VA0.
However the VA1 had its own daughter board for the controller (and its own EMI shield too).

RAM chips were never consolidated at any point. The two main RAM chips were replaced with a single SGRAM chip (with twice the pins), but this was most likely not done to make things cheaper because they used both designs side by side. The reason the Saturn has 16 board types is because 7 of those boards had a separate SGRAM variant (of which 2 were never even released).

It took until the Model 2 and then the VA6 where they started changing the design to be cheaper to make, but that was late 95/early 96.

[zyrobs]

And no, the changes from VA0 to VA1 weren't really much to talk about. VA15 was where the meaty changes happened, including changing over to the cheap Sanyo CD player. I've heard that those things are garbage.

Sanyo drives were used from VA6 and onwards, just pretty rarely. VA0 also had a second-sourced drive by Hitachi, it's so rare that I've seen it described as a prototype once.

The first cost saving measure was the Model 2 case (made the case much simpler, removed the access led and the memory reset button), then the VA6 (combined the CD block to one chip and used a different PLL), VA10 (combined the sound to one chip), and VA15 (combined two SH2s to one giant ASIC). They were also VERY aggressive at using up inventory, at one point they had 4 different revisions in production with different chipsets (single-chip or two-chip CD block, SGRAM or SDRAM memory).

[TrekkiesUnite118]

It absolutely was. The reason they had that 3 free games deal, is because they had to sell the Saturn for $50 more than the PlayStation. It's been talked about in discussions about Sony vs Playstation. Sony also marked the PlayStation down to $129 in 1998. At that point, they might as well have brought a tanker full of gasoline to Sega of America's HQ and torched the place, because that's how much damage it would have done to Sega to reach that price.

Talking about Japan here. In Japan we see Sega beat Sony to certain price points consistently, long before they happen in the US. In Japan Saturn hits the equivalent of ~$325 in June of 1995 with a Virtua Fighter Remix bundle. This was a price cut of about $90 from it's original launch price in Japan. PS1 at this point was still at it's launch price of ~$375. In July PS1 goes to ~$275 in Japan, Saturn was already around that price with the console by itself without Virtua Fighter Remix.

So that price cut there doesn't seem to be in direct response to Sony, but instead seems to be in response to cost cutting and the VA1 revision coming along. Some of these bundles appear to be VA0s, and others are VA1s. We see this pattern pretty consistently in Japan with Sega getting the Saturn to a certain price point, then Sony lowering to match a few months later. Even in the US it's not that reactionary past the first one. Sega hit's $199 around end of April, start of May in 1996 and stays there until June of 1997 when they go down to $149. Sony hits $199 in May of 1996 and stays there until about March of 1997 when it goes down to $149.

Also the 3 free games deal being because they had to sell Saturn for $50 more than the PS1 doesn't really make sense considering both systems were $199 before that promotion started.

Yeah and they didn't have to deal with the exchange rate, or import taxes. It's also possible that Sega of America was still holding onto old hardware that wasn't selling.

And no, the changes from VA0 to VA1 weren't really much to talk about. VA15 was where the meaty changes happened, including changing over to the cheap Sanyo CD player. I've heard that those things are garbage.

Sorry but I think going from a full size motherboard and a daughter board for the CD-ROM drive to a 2/3 size board that has everything on it save for the controller ports is a pretty dramatic revision. Sega seemed to think so as well since they changed the model number from HST-3200 to HST-3210 with this revision and rated the power consumption lower. I'd imagine producing less boards and smaller ones would reduce costs a bit as well as reduce the amount of assembly time to put them together. So I could see that adding up to reduce costs.

VA2-VA3 aren't as dramatic of changes. The main one that happens is they change to a smaller Power Supply that mounts to the bottom. I'd imagine that would help reduce assembly time as it's easier to assemble. It also reduces the complexity of the top case. The power supply itself is also smaller and less complex so I'd imagine that saved some costs too. Probably not huge savings, but still savings.

VA4 and VA5 is when it switches to the Model 2 case and we see the memory reset button, access LED, etc. dropped. I'd imagine the less fancy Model 2 case was where more of the savings came from here as it's a less complicated case in general to produce. At this point the Saturn hits the 20,000 Yen mark in Japan in March of 1996 and seems stays there for the rest of it's life in Japan. In the US it hits the $199 mark with these revisions. It doesn't drop in price again until June of 1997 when it goes down to $149 to match the PS1 that went down to that price a few months earlier.

VA6 is where the next major change happens and that's when we see the CD-ROM block consolidated down to 1 Chip. VA10 consolidates the SCSP, M68K and it's RAM down to 1 Chip, it's also where we start to see cheaper CD-ROM drives showing up. It should also be noted that these revisions all came out in 1996. So with the price holding at ~$200 from March of 1996 to June of 1997 I'd imagine the cost cutting from these revisions would have helped reduce losses.

VA15 is the last revision we see where the SH-2s are combined into 1 chip. This revision seems to hit around the later half of 1997. By this point the US price would be $149 for both the PS1 and Saturn but in Japan they'd still be at 20,000 Yen for Saturn and 19,800 Yen for PS1. Sony would drop to 18,000 at the end of 1997. From this point into early 1998 is about when Saturn production stops. Sony wouldn't drop their price again until the end of 1998 to $130 in the US, and about 15,000 Yen in Japan.

So while the price cuts in the first year and a half are aggressive from both Sony and Sega, you see it start to slowdown after the start of 1996. Yet behind the scenes they were still reducing the cost of the system. Around this time is when the Saturn starts to become profitable as they are able to reduce costs while maintaining the same price. The major issue here though is that Saturn was already failing outside of Japan and the Japanese market wasn't enough to keep things a float. This is why I think blaming it all on the cost isn't really the true issue. The real issue is the completely botched western launch that put them in a state they could never recover from.

He does point out that that 8:1 attach rate might very well include the 3 pack-in games sold with the console here. I know one thing for sure. Best Buy was clearing out unsold software in the spring of 1997. It had quite a lot of PlayStation games too, but the Saturn stuff was being sold for $8 each and I picked up about a dozen games that many would consider the better games for the Saturn. The PlayStation stuff was like a Military Strategy game (I bought the thing, but the name slips my mind, it sucked!) some other goofy crap games that really weren't that great and maybe some sport titles.

Do you honestly believe that amount of software sold came from US clearance bins and the 3 free games promotion? The 3 Free games promotion pushed maybe 1 million systems, so that would only be 3 million games in the 80 million game bucket. I honestly doubt even that much software was printed for the Saturn outside of Japan. If you add up all of the Famitsu data on the site he uses, you get 35 million games sold in Japan. But that data doesn't even account for half of the Saturn's Japanese library, and the games it does list that we have final numbers for seem to be off by 15-25%. We know about 9 million Saturns were sold worldwide, and of those about 6 million were sold in Japan. So if that's anything to go off of that would mean the Japanese market accounts for about 2/3s of the Saturn's total sales. If we apply that to the 80 million games sold number we'd get about 53 Million games sold in Japan, and 27 Million sold in the rest of the world. With about 2 million Saturns being sold in the US, that would imply about 17 million games sold in the US (14 if we don't include the 3 free games deal), with the remaining 1 million Saturns and 10 million games being the rest of the world and Europe.

So those promotions and clearance bins may have gotten an artificially high attach rate in the West, the real bulk of those 80 million games sold seems to come from Japan. Sure PS1 was dominating in Japan, but the Japanese Saturn numbers do beat the Japanese N64 numbers. Which is why I think the bigger issue isn't the cost of the system, but instead just how badly it was handled outside of Japan.

Yeah, $350 would have been reasonable, but with Sony having their console at $300 and a solid lineup of sports games, they were still better prepped than Sega for that Holiday season. I ended up getting a PlayStation in January of 1996, because I wanted to play Gameday and War Hawk.

Which is why I'm saying the bigger issue wasn't the price, but Sega of America dicking around with the 32X and not getting ready for a solid western Saturn launch with quality titles.

The VA1 streamlined production, the bill of materials was most likely the same as the VA0. The Access LED and Reset daughter boards for the VA0 were part of the main PCB during manufacturing, they were then literally cut off the corners, so they cost no extra. The SH1 sub-board was integrated into the main PCB on the VA1 but those sub-boards were probably something they got stuck with when the whole CD-ROM was still considered an add-on, so they had no chance but to use those on the VA0.

I'd still hazzard a guess that the streamlining production would have reduced the overall time and labor costs for assembling the systems, which would have helped get the price down. Also wouldn't the overall smaller board be cheaper to produce?

RAM chips were never consolidated at any point. The two main RAM chips were replaced with a single SGRAM chip (with twice the pins), but this was most likely not done to make things cheaper because they used both designs side by side. The reason the Saturn has 16 board types is because 7 of those boards had a separate SGRAM variant (of which 2 were never even released).

My bad I forgot that on some revisions those chips are on the other side of the board.

It took until the Model 2 and then the VA6 where they started changing the design to be cheaper to make, but that was late 95/early 96.

I'm not sure it really took until the Model 2 for things to start getting cheaper. Most of the internal changes were already done in the last Model 1 systems save for the Access LED and Memory reset button removal. In fact aren't VA4 and VA5 the same as VA2 and VA3 just in a Model 2 case?

[Gryson]

Some new 32X info that hasn't been posted yet in this thread. I was looking more into Al Nilsen's interviews and he goes into a bit more detail about comments he made that I previously posted.

Keep in mind that Nilsen left Sega in 1993, so he was not there in Jan 1994 when the 32X was conceived. That's important to remember because he refers to the add-on he saw as "the 32X", but it obviously was something different (unless we're going to doubt the entirety of the team who actually made the 32X).

We first learned about the 32X when Tom, Shinobu, and I were at Sega Japan and they made a presentation to us when we were walking through one of the R&D groups, showing us. You know, here's what the capabilities are, how much it was going to cost, it would cost $150. So you're adding $150 onto a $100 game console. Any new system goes and takes time to learn how to do it... Now you've got another $150 component and you're confusing the marketplace. You're cutting down on development resources so you're not going to go and get the great product that you want. And it was something we just didn't want. While I was at Sega I killed the product four times. It was like, "No. No. No. We can't do this." And it just kept being pushed and pushed. And finally, when I left, the agreement I get, Sega Japan pushed harder and it went into the marketplace.

Here's the interview. Note that I cut a few comments out in the middle if you're interested in hearing the whole thing.

I suspect Nilsen is referring to the "upgraded Genesis" concept here. Could the upgraded Genesis have been an add-on? That would make sense. Sega was playing around with the SVP at the end of 1993. One of the problems with the SVP was the ridiculously high cost. They may have considered the add-on concept as a way to avoid having to sell an SVP with every game that used it.

So, Japan was pushing an add-on to boost sales of the Genesis from 1993. SOA didn't want it and turned it down numerous times. Eventually, SOA struck on the idea of making the add-on more powerful via the SH-2s, which they figured would be a better seller than the less powerful upgraded Genesis.

It's interesting to imagine a SVP-like add-on released in 1993. Not that it would have been a good idea or anything, but a fun thought.

[Team Andromeda]

And I'm saying it may not be the same "spiced up MD". It could just be the Sega CD they're talking about. Rumors in Magazines should be taken with massive grains of salt.



Sure, but again we don't know if they're the same ones. The name "Gigadrive" was thrown around to a lot of different projects. What's relevant here is that in regards to Saturn and 32X is that Sega of America wanted something cheaper based on the Genesis, Japan offered them a Genesis with more colors, Sega of America countered with 32X. It was something that came about more by Sega of America begging Japan to let them do it, not due to some crazy knee-jerk reaction to the Jaguar as people try to claim. The interview people point to to back that claim up is simply one out of context piece of the larger picture.

Fair point and I agree with you over rumours; but when multiple outlets start reporting much the same rumour, there's usually an element of truth in it. Like I found out when sussed out of my brain with regards to the talk of the Saturn add-on expansion graphical chip.

Leaving that all aside. We had Joe Miller, Scot Bayless & Marty Franz all in different interviews all talk about and confirm the call from Japan to make what became the 32X and not in Tom's SEGA Japan bashing style. So I feel it's a little bit disingenuous for Kawagoe-san to say it was all SEGA America's idea. I've no time for SEGA America in the 32-bit era, but SEGA Japan looked into a spec up MD and to me made that call 1st. Sure we never really know about how much of the GigaDrive really was true, but I would imagine SEGA drew up plans for a counter to the Super Grafx, long before the Jag/3DO.

Also, I quite agree with you over price, that to me was never the real issue with the Saturn, but even SOJ was a little worried at the high price with having both the Saturn and Jupiter projects and all that was down to the high price of CD - Rom and memory needed.

I got my Saturn in May of 1995 and had a Die Hard Gameclub store near where I'd lived. I can say with most certainty, that the small amount of import software available for the Japanese Saturn,

That was true for many consoles. The early days of the Mega Drive software in Japan weren't great. The same was true for the PS2, Dreamcast, they also had poor launch software too. So to single out the Saturn isn't really fair

Yeah, and people still believed that Sega could continue on with the Saturn.

SEGA West was going to lose money if it did nothing or did something, if it did nothing sure it would save money, but then you'll have no money coming in with no software. I know you'll say it was different in America and maybe it was, but with the 32X and Mega CD. SEGA was getting a bad rep in the press in Pal land for bringing out new machines only to then move on to different things or poorly support the system with software.

To me, it would have been better if Sega West looked to spend a few million on bringing over Grandia (yes I know a huge translation job) and then titles like Gun Griffon 2 DOA Ect: which wouldn't have been the biggest translation jobs.

It was published in EDGE. Sega continued to use Hitachi for their processors even though Hitachi had little prior experience with microprocessors, suggesting there was a relationship beyond just "manufacturer-buyer."

It was. Also, its never given much credit, but when SEGA and Hitachi made their close partnership in September 1993. SEGA was able to break NCL distribution channel and use the likes of Hitachi Mediaforce 5000 plus outlets, to get Saturn into loads of retail, which helped with sales and made the Saturn a huge seller in Japan at the start.

The problem here is that, for many years, the internet has been filled with outrageous

This is 1989 and 1990, not many knew what the internet was back then, much less had access to it.

The Jaguar reference here is revealing, because Sega already had a product to compete against the Jaguar: the Saturn.

The Jaguar was meant to come in at a price of below £199 and you then also had SEGA looking to add parts to the Saturn to counter the PSX spec, which some felt would mean the Saturn could be delayed.

[stu]

Some new 32X info that hasn't been posted yet in this thread. I was looking more into Al Nilsen's interviews and he goes into a bit more detail about comments he made that I previously posted.

Keep in mind that Nilsen left Sega in 1993, so he was not there in Jan 1994 when the 32X was conceived. That's important to remember because he refers to the add-on he saw as "the 32X", but it obviously was something different (unless we're going to doubt the entirety of the team who actually made the 32X).


Quote "We first learned about the 32X when Tom, Shinobu, and I were at Sega Japan and they made a presentation to us when we were walking through one of the R&D groups, showing us. You know, here's what the capabilities are, how much it was going to cost, it would cost $150. So you're adding $150 onto a $100 game console. Any new system goes and takes time to learn how to do it... Now you've got another $150 component and you're confusing the marketplace. You're cutting down on development resources so you're not going to go and get the great product that you want. And it was something we just didn't want. While I was at Sega I killed the product four times. It was like, "No. No. No. We can't do this." And it just kept being pushed and pushed. And finally, when I left, the agreement I get, Sega Japan pushed harder and it went into the marketplace."

Here's the interview. Note that I cut a few comments out in the middle if you're interested in hearing the whole thing.

I suspect Nilsen is referring to the "upgraded Genesis" concept here. Could the upgraded Genesis have been an add-on? That would make sense. Sega was playing around with the SVP at the end of 1993. One of the problems with the SVP was the ridiculously high cost. They may have considered the add-on concept as a way to avoid having to sell an SVP with every game that used it.

So, Japan was pushing an add-on to boost sales of the Genesis from 1993. SOA didn't want it and turned it down numerous times. Eventually, SOA struck on the idea of making the add-on more powerful via the SH-2s, which they figured would be a better seller than the less powerful upgraded Genesis.

It's interesting to imagine a SVP-like add-on released in 1993. Not that it would have been a good idea or anything, but a fun thought.

Sounds to me like what was maybe needed was for Al Nilsen to have stuck around at SOA for another couple of years. Maybe then we would not have had the clusterfuck "launch of 32x then rushed Saturn launch" debacle ..lol

Seriously though, Al Nilsen makes good points about addons, they are normally a terrible idea: they confuse the marketplace, divert precious development and marketing resources etc.
In the previous interview that you linked, he mentions that with the Sega CD there were reasons for going ahead with the product eg it allows developers to learn how to create console gaming experiences on a new storage format, it also will appeal to an older more sophisticated audience.

Finally, it seems to me that what is happening is that both sides are trying to spin the narrative that "the other side is at fault" for the launch of the 32x. We started off with the American side blaming the Japanese team for it, then we have the Japanese team saying that the addon was requested by the American team, then finally it comes out that actually the Japanese team had tried pushing a 32x like addon on to the US market on a number of occasions and that they had been rebuffed, however the main opposition to the product leaves the company and then the product is upgraded again and is finally launched ( or pushed out the door) on to the US market. I guess it depends on who to believe.

[axel]

Some new 32X info that hasn't been posted yet in this thread. I was looking more into Al Nilsen's interviews and he goes into a bit more detail about comments he made that I previously posted.

Keep in mind that Nilsen left Sega in 1993, so he was not there in Jan 1994 when the 32X was conceived. That's important to remember because he refers to the add-on he saw as "the 32X", but it obviously was something different (unless we're going to doubt the entirety of the team who actually made the 32X).



Here's the interview. Note that I cut a few comments out in the middle if you're interested in hearing the whole thing.

I suspect Nilsen is referring to the "upgraded Genesis" concept here. Could the upgraded Genesis have been an add-on? That would make sense. Sega was playing around with the SVP at the end of 1993. One of the problems with the SVP was the ridiculously high cost. They may have considered the add-on concept as a way to avoid having to sell an SVP with every game that used it.

So, Japan was pushing an add-on to boost sales of the Genesis from 1993. SOA didn't want it and turned it down numerous times. Eventually, SOA struck on the idea of making the add-on more powerful via the SH-2s, which they figured would be a better seller than the less powerful upgraded Genesis.

It's interesting to imagine a SVP-like add-on released in 1993. Not that it would have been a good idea or anything, but a fun thought.

But at least in 1993 it wouldn't be competing alongside the Saturn. In the U.S. it would have a two year head start, which is an optimistic lifespan already. The fact that Nilsen had to kill it four times is kind of hilarious. How did nobody else see why this was a bad idea?

Have to wonder what a $150 upgrade would buy in 1993. The SH-2 would not have been ready, you have either the SH-1 that had just come out, or maybe a faster 68k. Probably the video mixing cable and framebuffer would have to be there too since adding more colors was one of the early goals.

[Team Andromeda]

But at least in 1993 it wouldn't be competing alongside the Saturn. In the U.S. it would have a two year head start, which is an optimistic lifespan already. The fact that Nilsen had to kill it four times is kind of hilarious. How did nobody else see why this was a bad idea?

Have to wonder what a $150 upgrade would buy in 1993. The SH-2 would not have been ready, you have either the SH-1 that had just come out, or maybe a faster 68k. Probably the video mixing cable and framebuffer would have to be there too since adding more colors was one of the early goals.

Very true, this wasn't the 32X at all.

[zyrobs]

Sorry but I think going from a full size motherboard and a daughter board for the CD-ROM drive to a 2/3 size board that has everything on it save for the controller ports is a pretty dramatic revision. Sega seemed to think so as well since they changed the model number from HST-3200 to HST-3210 with this revision and rated the power consumption lower. I'd imagine producing less boards and smaller ones would reduce costs a bit as well as reduce the amount of assembly time to put them together. So I could see that adding up to reduce costs.

A dramatic revision is not necessarily dramatically cheaper. They probably saved some money since they cut the LEDS from four to two, used less cable ties, but not dramatically cheaper. These are few dollar savings at this scale.

The PCB cost is a different thing, it's not as simple as "smaller board = teh cheap".
There are many more factors like how many traces there are, how many copper vias have to be put on there, how many layers the boards uses, and so on. The VA0 put most chips on the top, VA1 put a lot of the RAM to the bottom, that makes things more complex (the chips have to be glued down prior to wave soldering). There was the ribbon cable and two connectors for the controller sub board. The MPEG card slot used a THT connector too. That's 100+ more holes to drill on the board.
Someone with experience with PCB production would need to analyse all of this, I don't think neither you nor I are qualified for estimating this. Also remember that they went back to using very large boards with VA6 onwards, so I don't think PCB size alone was that big of a factor.

I think the main goal wasn't cost reduction but streamlining production. It could possibly make a difference in whether other assembly plants are suitable for producing it or not, though I wouldn't know much about that... VA0 was all made in Japan and only in 2 plants, except for the overseas models made in a third plant in Taiwan. From VA1 and onwards they start producing it at like eight different countries.

VA2-VA3 aren't as dramatic of changes. The main one that happens is they change to a smaller Power Supply that mounts to the bottom. I'd imagine that would help reduce assembly time as it's easier to assemble. It also reduces the complexity of the top case. The power supply itself is also smaller and less complex so I'd imagine that saved some costs too. Probably not huge savings, but still savings.

VA4 and VA5 is when it switches to the Model 2 case and we see the memory reset button, access LED, etc. dropped. I'd imagine the less fancy Model 2 case was where more of the savings came from here as it's a less complicated case in general to produce. At this point the Saturn hits the 20,000 Yen mark in Japan in March of 1996 and seems stays there for the rest of it's life in Japan. In the US it hits the $199 mark with these revisions. It doesn't drop in price again until June of 1997 when it goes down to $149 to match the PS1 that went down to that price a few months earlier.

The power supply was already on the bottom by VA1. The basic assembly method the consoles follow remains the same from VA1 Saturn all the way to the Dreamcasts. VA0 Saturn was closer to the Megadrive and Mega CD 2, down to using that godawful bad THT solder that is such a pain in the arse to remove.

VA2/3 changed the CD drive to 21pin, which is potentially significant for several reasons: the 20-pin drives were horribly built with parts on both sides, a lot of extra components (one of the MCUs even used external RAM!), and complex calibration process. 21pin drives reduced that complexity a lot.
The VA2/3 boards also used individual resistors and THT electrolytic caps, VA1 used SMD caps and resistor arrays. This change remained till the end of production.

VA2/3 was used in both model 1 and 2. They put whatever motherboard was available in both cases.
VA4/5 is the same as VA2/3 except for one chip, the DCC. I have no idea why that change was made, because the boards are identical otherwise. Maybe the old chip type was going out of production. That would also explain why they used anything from VA2-5 in Model 2s.

The model 2 definitely made things much cheaper, it removed the transparent try part of the case. That's a major thing, you now only need one mold and one type of plastic. This is not counting that they reduced the LEDs to one, the plastic light peg to one, and removed another pushbutton.

VA6 is where the next major change happens and that's when we see the CD-ROM block consolidated down to 1 Chip. VA10 consolidates the SCSP, M68K and it's RAM down to 1 Chip, it's also where we start to see cheaper CD-ROM drives showing up. It should also be noted that these revisions all came out in 1996. So with the price holding at ~$200 from March of 1996 to June of 1997 I'd imagine the cost cutting from these revisions would have helped reduce losses.

The VA6 only combined the SH1 and it's CD block ASIC into one chip, and the VA10 only combined the 68k and SCSP (and came with the same compatibility problem that the Genesis 3 had; apparently the shrunk down Yamaha made 68k was similar in both). The RAM remained there all the same. You can't exactly put a DRAM chip inside an ASIC that simply.
The CD drive from VA6 onwards was almost the same as the VA2-5 one, with the only difference being the MCU changed to a 32-pin one, which was definitely cheaper but who knows by how much. Also at this point piracy may have been a factor; the VA6 makes a subtle change in the pinout that confused a lot of modchips.

VA15 is the last revision we see where the SH-2s are combined into 1 chip. This revision seems to hit around the later half of 1997. By this point the US price would be $149 for both the PS1 and Saturn but in Japan they'd still be at 20,000 Yen for Saturn and 19,800 Yen for PS1. Sony would drop to 18,000 at the end of 1997. From this point into early 1998 is about when Saturn production stops. Sony wouldn't drop their price again until the end of 1998 to $130 in the US, and about 15,000 Yen in Japan.

So while the price cuts in the first year and a half are aggressive from both Sony and Sega, you see it start to slowdown after the start of 1996. Yet behind the scenes they were still reducing the cost of the system. Around this time is when the Saturn starts to become profitable as they are able to reduce costs while maintaining the same price. The major issue here though is that Saturn was already failing outside of Japan and the Japanese market wasn't enough to keep things a float. This is why I think blaming it all on the cost isn't really the true issue. The real issue is the completely botched western launch that put them in a state they could never recover from.

The Playstation was reduced to use half the internal components while the Saturn only really reduced something like 10%.
There's also that Hideki Sato interview where he mentions they simply cut production numbers to lower their losses - they made less consoles. By 1998 they almost completely halted production with only at most 300k units produced (and the ones I opened up had 1997 timestamps on a lot of chips too!).

We know about 9 million Saturns were sold worldwide

I need to go through the serials again but I very much doubt they even made that many consoles. 7-8 million is closer to the mark.

[TrekkiesUnite118]

The Playstation was reduced to use half the internal components while the Saturn only really reduced something like 10%.

Sure, but I'd imagine if the system was doing better outside of Japan we'd see it be in production longer with more reductions done. For the major components (CPUs, VDPs, Sound CHips, etc) it looks like they reduced things by about a 3rd by the time we get to the VA15 in 1995. I'd imagine the next thing to tackle would have been combining VDP1 and VDP2 if possible, or consolidating some of the smaller chips together like the SMPC and what not. In 1999 is when we se more dramatic changes happen to the PS1 like dropping the parallel port and what not. So I'd imagine if Saturn got a similar reduction we'd see things like the MPEG port get dropped as well as maybe the serial port since those weren't used that much beyond relatively niche accessories.

I think the bigger issue really was more how bad things were handled outside of Japan. If the system was launched better and therefore doing better the cost issue probably wouldn't have been as big of a problem.

There's also that Hideki Sato interview where he mentions they simply cut production numbers to lower their losses - they made less consoles. By 1998 they almost completely halted production with only at most 300k units produced (and the ones I opened up had 1997 timestamps on a lot of chips too!).

The one issue with this though is what period is Sato talking about here? The entire life of the console or just that period right after launch from 1994 to mid 1995? As the period from Launch to Mid 1995 seems to align with stories where people say there was demand for Saturn but it was harder to find in stores than the PS1. Looking at serial numbers it seems fewer Saturns were made in 1994 and production seems to ramp up more in 1995 and peaking in 1996.

I need to go through the serials again but I very much doubt they even made that many consoles. 7-8 million is closer to the mark.

I'm just going off of what Sega reported. Their numbers in their FY98 report stated 8.8 Million were sold at that point. It seems maybe another 500kish were sold after that point to get it a bit over the 9 Million mark from what most places report.

[Gryson]

Finally, it seems to me that what is happening is that both sides are trying to spin the narrative that "the other side is at fault" for the launch of the 32x.

I don't really agree with this take at all. I think there's too much bias here to see some "USA vs Japan" fight that never existed at the time.

Just about all of the people involved with the 32X have been very honest and upfront about how it was conceived. These statements from Nilsen and Miller just reinforce what Kalinske has said about Nakayama giving them free reign. SOA had the ability to decide what products it marketed and could reject anything they didn't want.

What we see with this add-on is Japan trying to support its subsidiary by using its R&D power to come up with a product concept in an attempt to boost a rapidly declining market. SOA didn't want it, Japan revised it, SOA proposed their own add-on, Japan consented. That's called cooperation. And I definitely think the parent company shares any responsibility with the daughter company in such a situation.

I have seen no real blame being applied in any of the quotes I've posted here (well, aside from that initial news article, but that wasn't from anyone involved), or anyone saying they were forced to do something.

But at least in 1993 it wouldn't be competing alongside the Saturn. In the U.S. it would have a two year head start, which is an optimistic lifespan already. The fact that Nilsen had to kill it four times is kind of hilarious. How did nobody else see why this was a bad idea?

Have to wonder what a $150 upgrade would buy in 1993. The SH-2 would not have been ready, you have either the SH-1 that had just come out, or maybe a faster 68k. Probably the video mixing cable and framebuffer would have to be there too since adding more colors was one of the early goals.

I think it's likely it was the SVP or something similar given the timeline. The SVP was announced in mid-1993, and Virtua Racing was ultimately released in mid-1994 in the US. Given the $100 price tag on Virtua Racing, it would make sense that Sega played around with the idea of including it in a separate attachment. That doesn't necessarily agree with what Joe Miller said (Miller implied the "upgraded Genesis" was something more than the SVP), but it could have undergone changes since Nilsen saw it.

We also have rumors of an SVP attachment:



From GamePro, April 1994.

Sega was trying to find a way to respond to Star Fox. As you say, a product like that released in 1993 wouldn't be competing against the Saturn at all.

[gamevet]

Talking about Japan here. In Japan we see Sega beat Sony to certain price points consistently, long before they happen in the US. In Japan Saturn hits the equivalent of ~$325 in June of 1995 with a Virtua Fighter Remix bundle. This was a price cut of about $90 from it's original launch price in Japan. PS1 at this point was still at it's launch price of ~$375. In July PS1 goes to ~$275 in Japan, Saturn was already around that price with the console by itself without Virtua Fighter Remix.

It was said that it cost Sega $380 to build a Saturn. Again, removing a small little board, a resistor or 2 and a bridge isn't going to magically cut production costs down to $325. Sega was most certainly taking a loss with that price cut. Don't look at price cuts being there, just because build costs go down. Labor in a place like Malaysia wasn't killing the profits. They pay people pennies on the dollar for labor. Reduction of labor over a missing board wouldn't amount to much, considering that the parts are still being put onto one board. It's the silicon that is expensive, with the Saturn having over 25 ICs on its board, with the cheapest probably being around $5 and the most expensive ones being in excess of $40.

So that price cut there doesn't seem to be in direct response to Sony, but instead seems to be in response to cost cutting and the VA1 revision coming along. Some of these bundles appear to be VA0s, and others are VA1s. We see this pattern pretty consistently in Japan with Sega getting the Saturn to a certain price point, then Sony lowering to match a few months later. Even in the US it's not that reactionary past the first one. Sega hit's $199 around end of April, start of May in 1996 and stays there until June of 1997 when they go down to $149. Sony hits $199 in May of 1996 and stays there until about March of 1997 when it goes down to $149.

Also the 3 free games deal being because they had to sell Saturn for $50 more than the PS1 doesn't really make sense considering both systems were $199 before that promotion started.

NO! NO! NO! NO! NO!

There aren't enough changes between VA0 and VA6 that would amount to that kind of savings in 1995. Have you actually ever bought resisters, diodes and capacitors? They're fucking cheap!

https://www.amazon.com/EDGELEC-LED-E...dDbGljaz10cnVl

They extended it, because Sony had lowered the price. It was easier for them to lose money on software, than hardware.

Sorry but I think going from a full size motherboard and a daughter board for the CD-ROM drive to a 2/3 size board that has everything on it save for the controller ports is a pretty dramatic revision. Sega seemed to think so as well since they changed the model number from HST-3200 to HST-3210 with this revision and rated the power consumption lower. I'd imagine producing less boards and smaller ones would reduce costs a bit as well as reduce the amount of assembly time to put them together. So I could see that adding up to reduce costs.

This is like a $5 savings. Labor is dirt cheap in that part of the world. Why do you think it is that the US no longer manufactures their own products, outside of automobiles?

VA2-VA3 aren't as dramatic of changes. The main one that happens is they change to a smaller Power Supply that mounts to the bottom. I'd imagine that would help reduce assembly time as it's easier to assemble. It also reduces the complexity of the top case. The power supply itself is also smaller and less complex so I'd imagine that saved some costs too. Probably not huge savings, but still savings.

VA4 and VA5 is when it switches to the Model 2 case and we see the memory reset button, access LED, etc. dropped. I'd imagine the less fancy Model 2 case was where more of the savings came from here as it's a less complicated case in general to produce. At this point the Saturn hits the 20,000 Yen mark in Japan in March of 1996 and seems stays there for the rest of it's life in Japan. In the US it hits the $199 mark with these revisions. It doesn't drop in price again until June of 1997 when it goes down to $149 to match the PS1 that went down to that price a few months earlier.

VA6 is where the next major change happens and that's when we see the CD-ROM block consolidated down to 1 Chip. VA10 consolidates the SCSP, M68K and it's RAM down to 1 Chip, it's also where we start to see cheaper CD-ROM drives showing up. It should also be noted that these revisions all came out in 1996. So with the price holding at ~$200 from March of 1996 to June of 1997 I'd imagine the cost cutting from these revisions would have helped reduce losses.

VA15 is the last revision we see where the SH-2s are combined into 1 chip. This revision seems to hit around the later half of 1997. By this point the US price would be $149 for both the PS1 and Saturn but in Japan they'd still be at 20,000 Yen for Saturn and 19,800 Yen for PS1. Sony would drop to 18,000 at the end of 1997. From this point into early 1998 is about when Saturn production stops. Sony wouldn't drop their price again until the end of 1998 to $130 in the US, and about 15,000 Yen in Japan.

So while the price cuts in the first year and a half are aggressive from both Sony and Sega, you see it start to slowdown after the start of 1996. Yet behind the scenes they were still reducing the cost of the system. Around this time is when the Saturn starts to become profitable as they are able to reduce costs while maintaining the same price. The major issue here though is that Saturn was already failing outside of Japan and the Japanese market wasn't enough to keep things a float. This is why I think blaming it all on the cost isn't really the true issue. The real issue is the completely botched western launch that put them in a state they could never recover from.

VA15 didn't appear in the Sega Saturn until 1998. It was the real cost cutter of the bunch. The rest of the revisions are save $5 here and $3 there. They aren't as substantial as you believe them to be. Actually, just the retooling and changes would have been a slight setback to production, which might have negated any cost savings they'd introduced.

https://segaretro.org/Sega_Saturn/Hardware_revisions

"This Is Cool" Skeleton Saturns: exact same as the normal Japanese units, except for the shell and the boxing, and they came with transparent controllers as well. They had VA13 and VA15 motherboards inside. VA13 seems to have been in the HST-0020 Special Campaign Original machines, while the VA15 ones were in HST-0021 models. These were among the last Saturns ever produced. The HST-0020s were made in late 1997, and the VA15 ones in 1998. They were all made by Seiyo Denshi. Estimated production based on the serials seen: ~30,000 to ~40,000. There is a rumor that made its way to Wikipedia stating that these units were made for the USA market, but this is false, no USA Skeleton machine exists. The only American skeleton unit was the Brazilian Tectoy model, which was most likely a rebranded Derby Stallion Skeleton unit.

Do you honestly believe that amount of software sold came from US clearance bins and the 3 free games promotion? The 3 Free games promotion pushed maybe 1 million systems, so that would only be 3 million games in the 80 million game bucket. I honestly doubt even that much software was printed for the Saturn outside of Japan. If you add up all of the Famitsu data on the site he uses, you get 35 million games sold in Japan. But that data doesn't even account for half of the Saturn's Japanese library, and the games it does list that we have final numbers for seem to be off by 15-25%. We know about 9 million Saturns were sold worldwide, and of those about 6 million were sold in Japan. So if that's anything to go off of that would mean the Japanese market accounts for about 2/3s of the Saturn's total sales. If we apply that to the 80 million games sold number we'd get about 53 Million games sold in Japan, and 27 Million sold in the rest of the world. With about 2 million Saturns being sold in the US, that would imply about 17 million games sold in the US (14 if we don't include the 3 free games deal), with the remaining 1 million Saturns and 10 million games being the rest of the world and Europe.

So those promotions and clearance bins may have gotten an artificially high attach rate in the West, the real bulk of those 80 million games sold seems to come from Japan. Sure PS1 was dominating in Japan, but the Japanese Saturn numbers do beat the Japanese N64 numbers. Which is why I think the bigger issue isn't the cost of the system, but instead just how badly it was handled outside of Japan.

No, I'm saying that legit good games were hitting the clearance bins at Best Buy, because they weren't selling. I was picking up games like Marvel Super Heroes, Street Fighter Collection, AMOK, Scorcher, Bomberman, Tempest 2000, Megaman X4, Decathlete and Soviet Strike. Those are the ones I remember off of the top of my head. I do also remember passing up on Legend of Oasis for a stupid Madden title, that really sucked!

I now remember one of the PlayStation games I'd bought, it was Panzer General. It was horrible!

Which is why I'm saying the bigger issue wasn't the price, but Sega of America dicking around with the 32X and not getting ready for a solid western Saturn launch with quality titles.

I think it goes beyond that though. EA didn't show up with a Madden game that year and when Sega finally did bring out their own NFL game, it sucked!

I'd still hazzard a guess that the streamlining production would have reduced the overall time and labor costs for assembling the systems, which would have helped get the price down. Also wouldn't the overall smaller board be cheaper to produce?

Boards are cheap. You're talking about $10 for the larger boards and $3-$5 for the smaller ones. Labor is also cheap, when it is all outsourced to the cheapest labor forces they can find.

I'm not sure it really took until the Model 2 for things to start getting cheaper. Most of the internal changes were already done in the last Model 1 systems save for the Access LED and Memory reset button removal. In fact aren't VA4 and VA5 the same as VA2 and VA3 just in a Model 2 case?

I had to buy all of those kind of components when I was attending electronics courses. My costs were about 10 cents for a resistor, $1 for an LED (Light emitting diode) and a switch would be about $3.50. I had to buy a breadboard kit that included buffer chips, capacitors, a test board, an open circuit board, IC sockets, small ICs (And, OR and whatever other gateways they were). I believe the kit was about $50 and I'm pretty sure that my book store had about a 30% markup. I still have my old kit. All of those items would have been a lot cheaper in bulk for a company like Sega.

[stu]

I don't really agree with this take at all. I think there's too much bias here to see some "USA vs Japan" fight that never existed at the time.

Just about all of the people involved with the 32X have been very honest and upfront about how it was conceived. These statements from Nilsen and Miller just reinforce what Kalinske has said about Nakayama giving them free reign. SOA had the ability to decide what products it marketed and could reject anything they didn't want.

What we see with this add-on is Japan trying to support its subsidiary by using its R&D power to come up with a product concept in an attempt to boost a rapidly declining market. SOA didn't want it, Japan revised it, SOA proposed their own add-on, Japan consented. That's called cooperation. And I definitely think the parent company shares any responsibility with the daughter company in such a situation.

I have seen no real blame being applied in any of the quotes I've posted here (well, aside from that initial news article, but that wasn't from anyone involved), or anyone saying they were forced to do something.

Well it certainly seems that the author of that original article that you first quoted had a axe to grind against Kalinske and Nakayama, the part you highlighted demonstrates that quite clearly. It would also seem from what you also said that Irimajiri had no love for either of them too. I would argue that the collaboration of Kalinske and Nakayama was far more successful than anything Irimajiri would achieve at the company .

I would agree with your views though that the 32X was a collective decision (obviously not the right one - with the benefit of hindsight) The 32x was designed in the US with SOJs blessing and assistance, its clear that SOJ had offered upgrade modules to the Genesis in the past, which were turned down. However it does seem clear that SOA started to feel pressure in the marketplace and obviously didn't feel comfortable with the pricing of the Saturn and therefore chose the route they chose.

It sounds like this reporter felt Kalinske should have been fired and disgraced or something and that Nakayama should of been punished the same way for allowing it to happen. As you stated the both the parent company and subsidiary worked cooperatively on the project and were trying to help each other, the fact that the plan didn't work out obviously regrettable but as they say .. shit happens. As usual people are silent when things are going good but are quick to criticize when things go down the shitter. And I generally aim that comment at the press not at anyone in SOJ.

[Gryson]

Well it certainly seems that the author of that original article that you first quoted had a axe to grind against Kalinske and Nakayama

There were certainly factions within Sega (as is true in any company), and I think Nakayama was always trying to resist interference from chairman Isao Okawa. Apparently, early on Okawa tried to take a more active role in Sega but Nakayama rejected it. We all know how that story turned out in the end. And Nakayama has spoken several times about pressure from other directors at Sega against entering the home console industry. Kalinske claimed Okawa was in that group.

The article in the OP is discussing general dissatisfaction within the executive ranks with how the situation was handled at SOA (as well as other examples). My suspicion is that the comments about Kalinske come from Irimajiri, but of course they are kept anonymous. Irimajiri has reason to put down Kalinske, too, since the Saturn technically failed under Irimajiri's watch at SOA (although I think it was clearly doomed before he took over). He would benefit from shifting blame.

Also, to speculate a bit more, I think there's a good chance the 1996 management change at SOA was forced. Kalinske, Nakayama, and Rosen all left their executive roles at SOA at the same time. Kalinske swears he wasn't forced out, but there's a general rule about a CEO who leaves when the company is in the red. I highly suspect pressure was applied on Nakayama via the board in Japan, and as a result all power at SOA was given to Irimajiri. This is the impression I get from reading about it in the Japanese newspapers, but of course there's nothing said explicitly. Maybe 'forced' is too strong a word, though. Agreed upon? They wanted to give Irimajiri complete power so the old guard had to go.

Anyway, I don't really see any Japan vs US kind of fighting at this point. It's more about who held power within the company.