TEXAN*s SEGA enthusiast thread - past and future hardware choices

[amk]

amk, Soul Calibur doesn't use bump mapping. The only places I've seen it are in the background for Sega Rally 2's title screen, and on some coins in Shenmue II.

I was under the impression that it hadn't been used at all.

 

[Fafalada]

No, actually.

That was PA reporting context, not the "geometry submissions to GPU" which is the usual 'standard' people always quote on this.
Comparing the numbers of two contexts is just adding to the nonsense, not clearing anything up.

Not to mention workloads per pixel shot-up dramatically on PS2 after a year or so, which made a lot more difference then few polygons here and there anyway.

 

[Sonic]

Here's a cool link I just found on google when looking up dreamcast and bump mapping.

http://yam.20to4.net/dreamcast/index_old.html

 

[elsence]

I remember that I read before a couple of months at trusted reviews, a rumor about PSP2 GPU. (POWERVR based)

http://www.trustedreviews.com/video...umour--PSP2-to-Offer-Xbox-Quality-Graphics/p1

http://www.imgtec.com/News/Release/index.asp?NewsID=449

Although of cource this was not possible from a technological standpoint in 1999, i just want to see if it was possible on 25nm manufacturing process...

I mean, does anyone knows:

1.how many million transistors
2.What die size (X mm2) and on what process (X nm)

is each core of the of the POWERVR SGX543 GP-GPU?

I suppose the lowest SKU is SGX543MP2 (200MHz, 2 core ver?), since this (SGX543 tech) is for the higher end mobile market i suppose it is targeting 640X480 screens? (i mean with this 2 core ver)

I just thought if it was possible to have a design such this with 25nm at 100MHz.

If each core is around or below 15 million transistors and they alter the design for speed and not power consumption they possibly could achieve something like 100Mhz (the lowest speeds for 25nm designs was 144MHz Voodoo3, 125MHz G400/TNT2 / and finally 115MHz Kyro 1 with 12 million transistors if i remember correctly)

I think that OpenGL™ ES 2.0 type of visuals even in 320X240 (1 core version 100MHz, 33 million polygons per second, fill rates in excess of 0,5Gpixel/sec with 2,5X scene depth complexity ) would look better than PS2 type resolution visuals...

Like i said above, this was not possible from a technological standpoint in 1999 (heck, this was not possible from a technological standpoint even in 2002) but i just had the curiosity...

 

[function]

That was PA reporting context, not the "geometry submissions to GPU" which is the usual 'standard' people always quote on this.
Comparing the numbers of two contexts is just adding to the nonsense, not clearing anything up.

So when it talks about number of polygons per frame under the "Rendering" heading, does that actually mean the number of polygons actually drawn to the framebuffer?

This seems to suggest so:

http://www.progaminsyndicate.com/tu...PI,SDK & Consoles/PS2/PerformanceAnalyser.pdf

Would this include or exclude polygons which are rejected based on the Z buffer?

 

[TEXAN*]

Is this the correct thread to be posting this grandmaster/shifty?

Blaze has two more SEGA licensed consoles in the works -

http://www.incgamers.com/News/18526/new-consoles-licensed-by-sega-emerge

New Consoles Licensed By Sega Emerge
11 Sep 2009 at 13:21:26 by James Chalmers

Two new consoles licensed by Sega will be hitting retailers before Christmas.

Blaze, which first came to prominence when it released the Xploder cheat cards, will be the company responsible for the new consoles. In recent years the company has had success with a handheld Sega console along with a remake of the Mega Drive and it is now planning two more consoles for the Christmas period, according to CVG.

It's quite fitting the news comes out during the same week that the Sega Dreamcast celebrates it's 10th anniversary. We have already seen the announcement earlier this week of a brand new Dreamcast game being developed. Or is this all just a coincidence?

Which consoles do you think Blaze will be releasing next? Could it be the Sega Genesis or in fact a portable Dreamcast? Let us know your thoughts in the comments section below.

Speculation is running rampant that it may be a Dreamcast re-release, they've already done Nomad, and MegaDrive/Genesis, I doubt it would be GameGear & Master System as that would be a step down from it's previous two offerings.

So what other options are there?

32X, SEGA/MEGACD, Saturn & Dreamcast.

 

[Sonic]

It could be portable DC. SEGA did manage to make the thing a SoC at some point.

 

[TEXAN*]

Renesas SH3707 (Aurora)

 

[grandmaster]

2. Listening to their fanbase about the design of the perheprials. Let me tell you something about the general public...they're morons. The company that designed one of the best and worst controllers of all time (speaking to the genesis 6 and 3 button controllers respectively) - managed to forget everythign good they'd learned and produce the Dreamcast controller...which was just god awfully....Do you remember the D-pad? Do you remember the pain? I do

I'm fairly sure the controller was "designed" via focus group, which said that the mainstream audience can't cope with too many buttons.

 

[Shifty Geezer]

Is this the correct thread to be posting this grandmaster/shifty?

Absolutely.

 

[Sonic]

Time to stir up the rumor mill. SEGA might be working on new hardware that should cost between $900 - 1200. This isn't for the home.

 

[TEXAN*]

Sonic, can you elaborate?

New hardware? Not for the home?

Even their high end RINGEDGE arcade board costs sub 300USD.

 

[TEXAN*]

PLAYSEGA has been updated, 40 of the best MegaDrive/Genesis titles are now available. They're even throwing in a Saturn pad.

http://www.playsega.com/index.php

There's also a 5 day free trial.

 

[function]

I'm fairly sure the controller was "designed" via focus group, which said that the mainstream audience can't cope with too many buttons.

I remember Sega saying something to that effect in the run up to the DC's release - they specifically wanted a controller that would allow them to reach out beyond the Dual Shock using crowd. Expand the market and all that.

This was probably one of the reasons that they wanted to increase controller feedback, thinking about it. A personalised pad with games able to access a given controller's VMU screen and speaker should add to user friendliness as well as overall utility.

Time to stir up the rumor mill. SEGA might be working on new hardware that should cost between $900 - 1200. This isn't for the home.

I hope it's something 3D (and that the ghost of Megadrive VR lives on)!

Arcade machines are one of the few devices that could make comfortable use of goggle-less 3D displays.

 

[TEXAN*]

Goggle free 3D display technology.

That would give arcades such an edge that the home won't catch up for atleast a decade.

Imagine a HotD5 where the zombies are literally coming out of the screen.

Frightening stuff.

 

[Lazy8s]

Interesting revelations about older technology like Dreamcast still pop up through the years as homebrew programmers, modders, and hackers get access to dev tools and explore the system.

Modding done with Shenmue has managed to switch the Passport model of Ryo, once believed to be too detailed for anything but a "head demo", into the actual game. The Dreamcast handles it rather admirably.

http://www.youtube.com/watch?v=c2agmLEuK3g

Since Shenmue's opening, prologue cut-scene already features a Passport-like model of Shenhua in a full game environment, and the Passport disc features video of an unused cut-scene featuring the full, Passport model of Ryo down at the harbor, this shouldn't come as much of a surprise.

 

[Mobius1aic]

.............and the Dreamcast lacked the ability to synthesize emotion.

OMG truedat yo.

Well like many of said, the PS2 was a looming monster, more powerful, DVD playback, many already set franchises.

 

[Simon F]

Can anyone comment on why bump mapping was never used, and what kind of difference it could have made if it was? Simon perhaps?

I just think it was a case of there not being any N-th generation games to make use of it. The bump mapping could also allow effects other than just lighting but I guess no one had the chance to explore it.

Or was there not enough fill rate or bandwidth available?

I don't think that's a likely. At 640 x 480 @ 50/60Hz (note not 25/30), you have enough theoretical texture fill for than 5 visible layers. Putting an extra layer for the bump map on all the walls, floors and characters should not be an issue unless they are burning a lot of texture fill for transparent planes.

Namco claimed they had fully exploited the DC with Soul Calibur when they announced Soul Calibur 2 for PS2, Xbox and GC but not DC. Perhaps if they had lower poly but bump mapped characters in a DC version... ? (Presumably they could have done higher poly and bumped characters for xbox.)

Doesn't that claim pop up every now and again for every console only to be superseded by a better game?


[update] I should add that after reading this thread yesterday, I dusted off my DC and got "a class D license" several times in a row [/update]

 

[fearsomepirate]

The fact that they've been so successful in the hands of richer, better run companies

There's an important matter of timing. Blowing money on Seganet (it cost them far more to run than it brought in) was idiotic in the late 90s, when not all that many people had Internet access, and most people accessed it through a modem. By contrast, Sony is the first company to make money on an Internet-enabled machine. But notice they didn't actually start building Ethernet into the system until 2004, and they didn't spend nearly that much on services. MS spent a ton more on Live, but the early part of this decade was the wrong time to do it. There's simply no way that was a good idea in 1998.

And no, VGA support didn't bankrupt Sega.

It's money they spent on the hardware that they didn't need to spend. That's the overall theme of Dreamcast: wasting money on features not enough people really cared about to make it worth it.

 

[function]

I just think it was a case of there not being any N-th generation games to make use of it. The bump mapping could also allow effects other than just lighting but I guess no one had the chance to explore it.

Could you elaborate on the other effects you had in mind? You get used to hearing that there's little flexibility in DC like hardware.

There's an important matter of timing. Blowing money on Seganet (it cost them far more to run than it brought in) was idiotic in the late 90s, when not all that many people had Internet access, and most people accessed it through a modem. By contrast, Sony is the first company to make money on an Internet-enabled machine. But notice they didn't actually start building Ethernet into the system until 2004, and they didn't spend nearly that much on services. MS spent a ton more on Live, but the early part of this decade was the wrong time to do it. There's simply no way that was a good idea in 1998.

I think the potential for making money was there before the PS2 Lite and Xbox Live (which made money even while the Xbox made a loss), but certainly narrowband was working against Sega. The European approach - charge people to use the system online by making them use your preferred ISP - was perhaps a better way of making it pay for itself but was always going to artificially limit its popularity.

Getting online up and running and established first could have had some long term value (as Xbox Live has demonstrated) and set Sega up well for the broadband era, but I guess that's academic at this point.

It's money they spent on the hardware that they didn't need to spend. That's the overall theme of Dreamcast: wasting money on features not enough people really cared about to make it worth it.

Assuming the $30 per modem approximation was accurate, by Jan 2001 you would have been looking at well over $200 million worth of modems sitting in people's DCs around the world. That would have been a lot of money to feed game development and advertising in the early stages of the DC's life.

A single hit MMORPG for the platform could have covered the modem costs long term, but again that's academic at this point. Selling modems and BB adapters as low margin peripherals might have been a wiser path to take. At the right time they could have used network adapter / keyboard / game bundles as high-value offerings to jump start online.