From PC to Next-Gen Consoles: Largest Performance Gap...

[Crazyace]

One point - were Sega making a profit selling DC's at $200?, various reports stated that there was quite a loss involved, so the various comparisions between DC and PS2 might be better off involving that.
If DC had an Elan chip it would have improved some things, but how many DC games are T&L bound, and how many have their looks due to other limitations..

 

[DeanoC]

Strictly speaking the Dreamcast CPU, and the PlayStation 2 CPU aren’t fully 128-bit. They are actually classed as 32-bit processors. If they were fully 128-bit, they could even out-perform the XBOX, which uses a 32-bit 733 MHz Intel Pentium 3.

Arghh. What is it with counting bits in processor thats so hard Its not even like there counting the bits on difficult processors.

For the record:
              Integer      Float       Integer SIMD       Float SIMD      Address bits
XBOX     :     32            80              64                     128             32
XBOXVS :      0              0                0                      128             8
PS2EE    :     64            32             128                      0 [1]          32
PS2VU0  :     16             0                 0                     128             16
PS2VU1  :     16             32[2]           0                     128             16
GC[3]     :     32             32              128                   64              32
Apple G5 :     64            64[4]           128                   128            42
Pentium4 :     32            80               128                   128             32[5]
AMD64    :      64            80               128                  128             40
GBARM   :      32             0                  0                      0               32
SH4DC   :      32             32[6]            64[7]                0               32 
Notes:
[1] PS2EE can use VU0 for float SIMD
[2] VU1 doesn't have scalar float caps but the EFU provides some scalar float functionality.
[3] Not certain about Gekko, having never programmed one.
[4] G5 have a fused multiple add with 100+ bits effective precision
[5] Xeon have extension to 36 bits
[6] DC has extended float scalar ops (approx sin/cos and dot product)
[7] Can't remember how big the DC's integer SIMD was (64 feels right)

 

[PC-Engine]

Apparently devs who know about it differ from you, who doesn't.

Developers who know about what? The definition of a word that requires no programming experience? You're running out of straws...

To illustrate that once again, you don't know what you're speaking about. Anyhow back on topic, NAOMI2 != DC. Furthermore, as I asked before, was NAOMI limited to the arcades? If you can't answer the question, then no biggie; there's no shame in it.

Umm...when did anyone say N2 equal DC? You're lost dude. N1 = DC but of course in your delusional world it doesn't because it's a totally different architecture that costs a lot more and has a different name and JAMMA connectors and an arcade cabinet and arcade joystick and buttons. Looks like you're the one who doesn't know what you're talking about. Maybe you should ask SEGA how they shrunk NAOMI 1's circuitry into a DC and sold it for $200. I think they even shrunk the arcade cabinet and monitor into the DC too.

One point - were Sega making a profit selling DC's at $200?, various reports stated that there was quite a loss involved, so the various comparisions between DC and PS2 might be better off involving that.
If DC had an Elan chip it would have improved some things, but how many DC games are T&L bound, and how many have their looks due to other limitations..

When DC was released in Japan, SEGA was obviously losing money on each unit, how much nobody knows.

When it was released in the US 1 year later SEGA was probably still losing money. Whether or not it was close to breaking even nobody knows. The orginal DCs were manufactured in Japan. Later they were moved to China. The original DCs also had liquid filled heat pipes.

DC games don't look as good as say a N2 game most probably because of lighting and geometry.

 

[Deepak]

What is XBOXVS??

 

[PC-Engine]

What is XBOXVS??

Vertex Shader maybe?

 

[PC-Engine]

you guys are classic...

If you then factor in the ps2 cost sony 450$ (At least) to make. The dreamcast in 1998 cost 200$ to make. Sega could have done a double dreamcast (Basicly a naomi 2 ) for 400$. Which would have been cheaper than an actual ps2.

You are trying to convince us, Sega "could have" pulled of a double dreamcast , being more expensive than the Dreamcast, and NOT suffer the consequences? They couldn't even keep themselves alive with the cheaper Dreamcast and now you want any credibility on them being able to pull off a more expensive NAOMI2 Dreamcast a year later? (Not to mention that such an attempt by Sega wouldn't have been unnoticed by Sony and they in turn could have made a "double PS2"... which again changes the picture all together. )

Sega wasn't in any other position. GET OVER IT! Anything else is really grasping at straws.

Now can't we have one thread not derrail to the usual topics and inturn discuss the real topic of this thread?

Your post doesn't make any sense. DC was released before PS2 and also cheaper so it was less powerful. If it was released the same time as PS2 and had a price equal to PS2 it would benefit from the ELAN chip and 2nd GPU and more memory. Don't know why you think SONY was the only one who had access to powerful technology back then. Maybe you should get over it? I think his point was that SEGA had access to competitive technology. He wasn't talking about SEGA having the financials to actually manufacture, market, and take losses on the machine. OTOH if MS used the N2 technology back then instead of entering the market over a year late...

 

[Phil]

Your post doesn't make any sense. DC was released before PS2 and also cheaper so it was less powerful. If it was released the same time as PS2 and had a price equal to PS2 it would benefit from the ELAN chip and 2nd GPU and more memory. Don't know why you think SONY was the only one who had access to powerful technology back then.

Exactly my point: Sega wasn't able to keep the Dreamcast afloat which had a lower budget (your claims) than PS2, 1 year prior. What makes you think Sega would have had the same budget as PS2 a year later and keep their project afloat? What proof do you have? It's pretty clear to me, by the events and Sega pulling out of the console market, that they had a very limited budget available. On the other hand, the PS2 cost as much as it did at launch to make because Sony had pretty concrete plans on how they would drop the price and what sales they were expecting. They fabbed their own chips, meaning their strategy allowed them to drop manufacturing costs strategically over the console's life. Would Sega have been in a similar position had they allowed for a similar high budget? I doubt it. They wouldn't have lasted even a year.

Maybe you should get over it? I think his point was that SEGA had access to competitive technology. He wasn't talking about SEGA having the financials to actually manufacture, market, and take losses on the machine. OTOH if MS used the N2 technology back then instead of entering the market over a year late...

Get over what? That you are making up situations which are not feasible in real life?

Sega launched with Dreamcast when they did because they felt they needed to have the head-start, most likely because they knew that they had a very limited budget. Simply having better hardware wouldn't have given them the mindshare they would have needed. It also wouldn't have guaranteed support - support the PS2 was getting along with all the well known franchises and software in development which were loved on the original PlayStation.

And what's the point in making up scenarios in which the company in question has more resources than it could have? It's pretty much a fact that Sega didn't have the budget Sony had and that is closely tied in with the outcome of the prior generation along with other factors such as mindshare, own fabbing of chips, cost cutting, break-even point, internal/external resources, R&D, market acceptance (how well will my product sell?) and others. Simply changing one or two factors will get flawed results (as you are doing). Fact is: Sega didn't. Only they know of the true reasons and simply suggesting/speculating different is at best a laughable attempt in ignoring the real outcome, in other words, reality. If it's anyone grasping at straws, it's definately the one not accepting reality. I really think it's time to move on and get over it.

 

[Devourer]

WHich would easily be on par with a ps2.

Where did PS2's VU - the most distinguishing point in its architecture - go?

what are you talking about ?


a naomi 2 would have been capable of 10 million polygons with 4 lights (was it more ?) sustained. it would have also had a 200mega pixel fillrate .all the way up to 2000mega pixel fillrate.

It would have also been able to do fsaa , bump maping , volumetric effects and a host of other things.

Would, would, would...

In reality, all that was in the arcades and no one can tell what kind of machine they could have actually produced for the home market, with DVD and all.

Also, I agree with Phil post, which is just way more competent than saying "If Sega launched DC in 2000 they would have put 128 MB RAM". There are way more factors than available technology when marketing a new console.

 

[pahcman]

Wow talk bout missing all points phil, ty, devourer and gang...

I wonder you guys read and understand the thread... NO ONE is saying Sega, in their state then, can financially support a DC-N2 in 2000, BUT the technological/cost feasibility is there. If someone bigger/richer like MS was present then, what to say Xbox couldnt run on N2 that meet and probably surpass the graphics of PS2. What makes you think Sony is that far in technology then, now and future?

As for ram and costs, jvd said it best man. Go back and read what he said. Keep note arcade boards are more expensive then consoles. No royalties from softwares for start. Look at 246/Chihoro/Triforce and come back to us. Thanx.

A N2 console can just run at half ram and still look as good, proven by Naomi > DC ports. Those look virtually unchanged from arcades. Texture compression and the less need of caching everything helps man.

Of cos, before you blame the Sega guys, just note, this "IF would IF" preposition will never come up IF ONLY "mr one" hadnt brought up IF Sony had... back at page 3 or 4...

But this is not the PC market where x86 still dominates, this is the embedded console market where you can adopt radical design such as PS2 without compatibility to old technology. Today PS2 is seen as less powerful machine than newer Xbox, but if PS2 had had more memory, and if Toshiba had not misestimated future GPU trend in designing GS...?

Read and understand before posting. It helps.

 

[Devourer]

Wow talk bout missing all points phil, ty, devourer and gang...

I wonder you guys read and understand the thread... NO ONE is saying Sega, in their state then, can financially support a DC-N2 in 2000, BUT the technological/cost feasibility is there.

I got the point perfectly and I disagree. Feasibility WASN'T there. Imho of course. Read the posts before accusing others of missing the point

 

[Devourer]

Of cos, before you blame the Sega guys, just note, this "IF would IF" preposition will never come up IF ONLY "mr one" hadnt brought up IF Sony had... back at page 3 or 4...

Mr One didn't make utterly bold statements with this ifs and woulds. PC-Engine e JVD are makind definite statements whith their ifs and woulds. I hope one can still have the right to disagree to definite statements made out of ifs and woulds.

 

[PC-Engine]

Your post doesn't make any sense. DC was released before PS2 and also cheaper so it was less powerful. If it was released the same time as PS2 and had a price equal to PS2 it would benefit from the ELAN chip and 2nd GPU and more memory. Don't know why you think SONY was the only one who had access to powerful technology back then.

Exactly my point: Sega wasn't able to keep the Dreamcast afloat which had a lower budget (your claims) than PS2, 1 year prior. What makes you think Sega would have had the same budget as PS2 a year later and keep their project afloat? What proof do you have? It's pretty clear to me, by the events and Sega pulling out of the console market, that they had a very limited budget available. On the other hand, the PS2 cost as much as it did at launch to make because Sony had pretty concrete plans on how they would drop the price and what sales they were expecting. They fabbed their own chips, meaning their strategy allowed them to drop manufacturing costs strategically over the console's life. Would Sega have been in a similar position had they allowed for a similar high budget? I doubt it. They wouldn't have lasted even a year.

Maybe you should get over it? I think his point was that SEGA had access to competitive technology. He wasn't talking about SEGA having the financials to actually manufacture, market, and take losses on the machine. OTOH if MS used the N2 technology back then instead of entering the market over a year late...

Get over what? That you are making up situations which are not feasible in real life?

Sega launched with Dreamcast when they did because they felt they needed to have the head-start, most likely because they knew that they had a very limited budget. Simply having better hardware wouldn't have given them the mindshare they would have needed. It also wouldn't have guaranteed support - support the PS2 was getting along with all the well known franchises and software in development which were loved on the original PlayStation.

And what's the point in making up scenarios in which the company in question has more resources than it could have? It's pretty much a fact that Sega didn't have the budget Sony had and that is closely tied in with the outcome of the prior generation along with other factors such as mindshare, own fabbing of chips, cost cutting, break-even point, internal/external resources, R&D, market acceptance (how well will my product sell?) and others. Simply changing one or two factors will get flawed results (as you are doing). Fact is: Sega didn't. Only they know of the true reasons and simply suggesting/speculating different is at best a laughable attempt in ignoring the real outcome, in other words, reality. If it's anyone grasping at straws, it's definately the one not accepting reality. I really think it's time to move on and get over it.

You're talking about SEGA's financial ability. I'm talking about SEGA's technological capability. We're not talking about the same thing. Nobody denied the fact SEGA ran out of money and pulled the plug. You're arguing with yourself

 

[rabidrabbit]

Guys, why don't you start a new thread called "Dreamcast: The Little Console That Would" instead of keeping this thread OFF TOPIC.

Sure, DC could've been as powerful or even more powerful on some areas than PS2 had it been designed/released later.
Sure, xbox could've been as powerful or even more powerful on some areas than PS2 had it been designed/relased earlier.
Sure, you could've manufactured more powerful and advanced tech than PS2 had you the money available.

These are just games consoles, they are not the absolute cutting-edge of the microprosessor technology, as much as the marketing people would like us to believe so.

 

[Phil]

PC-Engine:

...and my point is that you can't argue on their technological capability without factoring in their financial ability. It is meaningless and has no bearing on reality nor on the subject of this thread, which, can I remind was on next generation consoles and the performance gap to PC that can be expected.

Oh and the above is also directed at pahcman and gang that wish to continue this pointless OFF-TOPIC argument.

 

[PC-Engine]

PC-Engine:

...and my point is that you can't argue on their technological capability without factoring in their financial ability. It is meaningless and has no bearing on reality nor on the subject of this thread, which, can I remind was on next generation consoles and the performance gap to PC that can be expected.

Oh and the above is also directed at pahcman and gang that wish to continue this pointless OFF-TOPIC argument.

If you don't want to talk about it then don't fuel it.

NAOMI 2 was proof of SEGA's technical capability. Not being able to make a console based on that technology is financial not technical assuming that was the reason why they didn't.

 

[Panajev2001a]

Deano, I thought that the SH-4 FPU had two modes: scalar FPU and Vector Engine (so it would have SIMD FLOPS) which we can see in normal FMAC instructions and specialized dot products and the like.

This is based on the SH-4 core.

FPU

• * On-chip floating-point coprocessor
  * Supports single-precision (32 bits) and double-precision (64 bits)
  * Supports IEEE754-compliant data types and exceptions
  * Two rounding modes: Round to Nearest and Round to Zero
  * Handling of denormalized numbers: Truncation to zero or interrupt generation for compliance with IEEE754
  * Floating-point registers: 32 bits x 16 words x 2 banks
  (single-precision x 16 words or double-precision x8 words) x 2 banks
  * 32-bit CPU-FPU floating-point communication register (FPUL)
  * Supports FMAC (multiply-and-accumulate) instruction
  * Supports FDIV (divide) and FSQRT (square root) instructions
  * Supports FLDI0/FLDI1 (load constant 0/1) instructions
  * Instruction execution times
  o Latency (FMAC/FADD/FSUB/FMUL): 3 cycles (single-precision), 8 cycles (double-precision)
  o Pitch (FMAC/FADD/FSUB/FMUL): 1 cycle (single-precision), 6 cycles (double-precision)
  Note: FMAC is supported for single-precision only.
  * 3-D graphics instructions (single-precision only):
  o 4-dimensional vector conversion and matrix operations (FTRV): 4 cycles (pitch), 7 cycles (latency)
  o 4-dimensional vector inner product (FIPR): 1 cycle (pitch), 4 cycles (latency)
  * Five-stage pipeline

http://www.renesas.com/http://makeashorterlink.com/?B2C142219

 

[Fafalada]

[3] Not certain about Gekko, having never programmed one.

Allow me

                Integer        Float       Integer SIMD      Float SIMD     Address bits
GC        :     32             64          0(1)              64             32/52(2)

1) Gekko has integer SIMD 'like' instructions for 'string' operations (store/load string), but x86 had that since 8bit days and noone classified it as SIMD so.
2) 32 Physical/52 Virtual

And because I like to nitpick , I will mention that EE has (very)limited 128bit number integer arithmetic too. Question, shouldn't bitness chart include memory bus width too though?

 

[Ty]

NAOMI 2 was proof of SEGA's technical capability. Not being able to make a console based on that technology is financial not technical assuming that was the reason why they didn't.

If that was your argument then you're more off base than I guessed. It was proof of ImgTech's technological capability, not Sega's! And your earlier argument that the technology could exist is patently obvious SINCE it already existed (NAOMI). Breaking it down for you further;

1> NAOMI is proof of ImgTech's technological capability since they designed the heart of it. NAOMI is proof of Sega's ability to license technology, not create it.
2> NAOMI exists so of course NO ONE is talking about whether or not the technology exists because it does already - duh.

Therefore the argument is around whether or not Sega could have used NAOMI2 (you mentioned NAOMI2 first I believe) type of technology for a console moves into the financial aspects as well, which Sega certainly couldn't have done since they couldn't even float DC1 at the time.

If other people such as pachman want to pretend that someone other than Sega had a license to the technology then sure, I can easily imagine situation of that company shrinking NAOMI2 into a console. When? At what price? I don't know but I know they wouldn't be off-the-shelf-parts.

 

[PC-Engine]

NAOMI 2 was proof of SEGA's technical capability. Not being able to make a console based on that technology is financial not technical assuming that was the reason why they didn't.

If that was your argument then you're more off base than I guessed. It was proof of ImgTech's technological capability, not Sega's! And your earlier argument that the technology could exist is patently obvious SINCE it already existed (NAOMI). Breaking it down for you further;

1> NAOMI is proof of ImgTech's technological capability since they designed the heart of it. NAOMI is proof of Sega's ability to license technology, not create it.
2> NAOMI exists so of course NO ONE is talking about whether or not the technology exists because it does already - duh.

Therefore the argument is around whether or not Sega could have used NAOMI2 (you mentioned NAOMI2 first I believe) type of technology for a console moves into the financial aspects as well, which Sega certainly couldn't have done since they couldn't even float DC1 at the time.

If other people such as pachman want to pretend that someone other than Sega had a license to the technology then sure, I can easily imagine situation of that company shrinking NAOMI2 into a console. When? At what price? I don't know but I know they wouldn't be off-the-shelf-parts.

SEGA designed the NAOMI 1/2 systems not IMGTEC. N1/2 wouldn't exist if SEGA didn't design it. Parts don't magicaly build nicely performing systems by themselves

I guess Nintendo has no technical capability either since they designed the GCN buy working with technology partners. Parts just mysteriously start working when you throw them together with no design and engineering thought. Hey maybe we should all go buy a cpu a gpu and design our own consoles?

Keep grasping...

 

[Ty]

SEGA designed the NAOMI 1/2 systems not IMGTEC

Keep grasping...

Lol. Now we're going to have an argument about what "design" means. Here's a clue for you since you seem to be lacking in this department.

CPU : Hitachi SH-4 64-bit RISC CPU (200 MHz 360 MIPS / 1.4 GFLOPS)
Graphic Engine : PowerVR 2 (PVR2DC)
Sound Engine : ARM7 Yamaha AICA 45 MHZ (with internal 32-bit RISC CPU, 64 channel ADPCM)
Main Ram : 32 megs <-- Whoa, wonder if Sega "designed" this?
Main Memory : 32 MByte <-- Or this?
Graphic Memory : 16 MByte <--This? NOPE!
Sound Memory : 8 MByte <--We getting any closer? NOPE!
Media : ROM Board (maximum size of 172MBytes) / GD-Rom <-- Ok, maybe this.

Looks like they were all off-the-shelf parts.