StarDestroyer.Net BBS

mplsjocc wrote: They had the "Super Mario Club"

Sorry, that's not the same sort of thing as a modern online community, the requirements for building such are far greater then Nintendo's online experience.

Okay, so they've had a few jaunts into online activity, but it doesn't compare to what went into creating Xbox Live, nor can any of that experience or technology carry over into helping with the Revolution.

You apparently didn't read that link...

The Kernel wrote: Stop talking about concepts you don't even have a vague conception of. You are wrong about the PowerPC not having OOO Execution and you are also wrong about assuming 50% performance across the board (EPIC uses in order execution and is the most powerful instruction set implementation in the word at many things).

Don't talk smack about shit you know next to nothing about. The PPC in the X360 is NOT a PPC970, nor a POWERx derivative (which have OOOE). In fact, as speculated = here it is the same CPU core as in the PS3, but without the vector units and there are three of them. You may argue against "speculation" but then your argueing that there is another chip exactly identical to the CPU core in PS3 that has never been seen or heard before. I find that exceedingly unlikely, especially considering that only this is the only CPU at IBM has the ability to reach 3Ghz+. In short, it is almost certainly the same, which makes a IOE processor.

EPIC doesn't count as while it is IOE it's also a 6-issue core (compared to 2 for the PPC and 3 for Pentiums) with a huge amount of cache (1.5-9MB L3, 256KB L2), and uses the IA-64 instruction set which was designed to lessen it's IOness, and it still loses to Pentiums and Athlons in integer ops. Integer performance is also the same weakness the much slimmer PPC has to face too. For more realistic CPUs, the 50% claim stands. If you dispute the 50% claim, then you dispute the guy who wrote that, not me. Floating point ops will be much better, which is the PPC's strength, but that's useless in emulation so backwards compatibility is still simply impossible.

HyperionX wrote: Don't talk smack about shit you know next to nothing about. The PPC in the X360 is NOT a PPC970, nor a POWERx derivative (which have OOOE). In fact, as speculated = here it is the same CPU core as in the PS3, but without the vector units and there are three of them. You may argue against "speculation" but then your argueing that there is another chip exactly identical to the CPU core in PS3 that has never been seen or heard before. I find that exceedingly unlikely, especially considering that only this is the only CPU at IBM has the ability to reach 3Ghz+. In short, it is almost certainly the same, which makes a IOE processor.

That makes little sense considering that the alpha dev kits use PPC 970's.

Even if true though, you still are wrong about a 50% across the board performance hit as well using reported IPC as a performance metric for determining emulation performance.

Performance of this chip will be highly dependent on application. In order chips are easier to keep fed since you don't have to worry about dependencies or execution units being busy. The only real problem with an in order chip is cache misses, which can lead to stalls while loads are fetched. This isn't nearly as big of a problem when you are dealing with an emulator as the primary execution code could be small enough to fit in a Level 1 data cache.

EPIC doesn't count as while it is IOE it's also a 6-issue core (compared to 2 for the PPC and 3 for Pentiums) with a huge amount of cache (1.5-9MB L3, 256KB L2), and uses the IA-64 instruction set which was designed to lessen it's IOness, and it still loses to Pentiums and Athlons in integer ops.

Itaniums problem with integer performance center around the design philosophy, not it's in order nature. Integer code contains a number of conditional branches which makes it harder for an IA-64 compiler to create larger instruction bundles, which means several of the Itanium execution units might be dormant. This is why compiler performance is so important to the Itanium.

Integer performance is also the same weakness the much slimmer PPC has to face too. For more realistic CPUs, the 50% claim stands. If you dispute the 50% claim, then you dispute the guy who wrote that, not me. Floating point ops will be much better, which is the PPC's strength, but that's useless in emulation so backwards compatibility is still simply impossible.

The person who wrote that admits himself that the performance differences are highly dependent on the situation. Obviously when you are dealing with software that has a great deal of cache access, you are going to have performance loss using an in order design, but this does not include every integer heavy application, especially one custom written for the hardware at a low level the way an emulator is.

mplsjocc wrote:You apparently didn't read that link...

I read it. What specifically would you say is analagous to setting up Xbox Live?

Praxis wrote: That last sentence is totally ridiculous.

Not at all, it's the best business decision they could ever make. Microsoft is feverent to break into the Japanese market, Nintendo could sell themselves for four times their actually value and Microsoft would be glad to pay. Or did you forget that the purpose of business is to provide shareholder value?

Even if that wasn't the route they took, Nintendo would be better off playing to their strengths. They should get out of the console (not the portable) hardware market and focus on making games, a la Sega.

Nintendo stated that the day they leave the console market is the day they stop making games.

The Kernel wrote:Not at all, it's the best business decision they could ever make. Microsoft is feverent to break into the Japanese market, Nintendo could sell themselves for four times their actually value and Microsoft would be glad to pay. Or did you forget that the purpose of business is to provide shareholder value?

Even if that wasn't the route they took, Nintendo would be better off playing to their strengths. They should get out of the console (not the portable) hardware market and focus on making games, a la Sega.

Ummm... you realize how much of an oversimplification all of this is? Are you saying that Nintendo should sell out to Microsoft at any price? It's likely that all big companies talk about mergers every now and then, but it's generally almost impossible to get them all to agree on terms.

mplsjocc wrote:Nintendo stated that the day they leave the console market is the day they stop making games.

Confirmed...

Iwata put an end to rumors that Nintendo might go the way of Sega (as in making games only), saying, "When we withdraw from the home game console, that's when we withdraw from the video game business." We'll see, we'll see.

From Here

mplsjocc wrote:Nintendo stated that the day they leave the console market is the day they stop making games.

Which is exactly the sort of stubborn, prideful attitude that is going to drive them into irrelevence. A company needs to play to their strengths and Nintendo has shown no ability to move beyond the corporate mindset of the 1980's where they ruled the home gaming market.

Their pride lost them the 4th generation console wars (the N64) and saw them bleed massive amounts of business to Sony. Again their pride caused them to make bad decision with regards to third parties which caused them to desert in droves which helped kill the Gamecube's success. Unless Nintendo learns how to adept with the times, they are going to get slaughtered in the coming years.

Master of Ossus wrote: Ummm... you realize how much of an oversimplification all of this is? Are you saying that Nintendo should sell out to Microsoft at any price? It's likely that all big companies talk about mergers every now and then, but it's generally almost impossible to get them all to agree on terms.

Microsoft would pay an enormous sum for Nintendo, this has been made clear by repeated attempts to buy them out. Nintendo is worth a lot more to Microsoft then their percieved value as they would be the key to breaking into Japan so that Microsoft could sell all sorts of things to the Japanese consumers.

The Kernel wrote:

mplsjocc wrote:Nintendo stated that the day they leave the console market is the day they stop making games.

Which is exactly the sort of stubborn, prideful attitude that is going to drive them into irrelevence. A company needs to play to their strengths and Nintendo has shown no ability to move beyond the corporate mindset of the 1980's where they ruled the home gaming market.

Their pride lost them the 4th generation console wars (the N64) and saw them bleed massive amounts of business to Sony. Again their pride caused them to make bad decision with regards to third parties which caused them to desert in droves which helped kill the Gamecube's success. Unless Nintendo learns how to adept with the times, they are going to get slaughtered in the coming years.

Their pride in what they do has kept them in business for close to 100 years. I like Nintendo, and I'm glad that they don't puss out and conform to every quick fad that the casual gamer drives the market towards. With a new president in the mix, the company is showing signs of changing and being a bit more aggressive. I, for one, look forward to E3, and bet that Nintendo steals the show..again.

What's funny is that once MS wins (which they will, for reasons mentioned already) the console game scene will die even faster than it is already.

mplsjocc wrote: Their pride in what they do has kept them in business for close to 100 years. I like Nintendo, and I'm glad that they don't puss out and conform to every quick fad that the casual gamer drives the market towards.

Then you are praising the corporate culture that during the 80's was the worst monopoly this side of Microsoft. They fixed prices, throttled supply in order to keep demand up, and bullied retailers with threats of pulling their products and destroying their business.

With a new president in the mix, the company is showing signs of changing and being a bit more aggressive. I, for one, look forward to E3, and bet that Nintendo steals the show..again.

The Kernel wrote:

HyperionX wrote: Don't talk smack about shit you know next to nothing about. The PPC in the X360 is NOT a PPC970, nor a POWERx derivative (which have OOOE). In fact, as speculated = here it is the same CPU core as in the PS3, but without the vector units and there are three of them. You may argue against "speculation" but then your argueing that there is another chip exactly identical to the CPU core in PS3 that has never been seen or heard before. I find that exceedingly unlikely, especially considering that only this is the only CPU at IBM has the ability to reach 3Ghz+. In short, it is almost certainly the same, which makes a IOE processor.

That makes little sense considering that the alpha dev kits use PPC 970's.

That's because the alpha dev kits are nothing more than dual core Macs. You won't see those in beta kits.

Even if true though, you still are wrong about a 50% across the board performance hit as well using reported IPC as a performance metric for determining emulation performance.

That's true, sort of. It won't be 50% on every application, but nearly every application will suffer a hit and those that get >50% hits that balances the average are especially problematic because in an emulator they need to run at full speed or emulation will stall.

Performance of this chip will be highly dependent on application. In order chips are easier to keep fed since you don't have to worry about dependencies or execution units being busy.

In other news, 2 + 2 = 5, up is down, black is white. This is a total load of 100% wrong BS. This is specifically the weakness of IOE: dependencies. An IO process must wait for dependencies to resolve before continued computation and the pipeline stalls, one of it's main critical weaknesses against OOO processors. In general while there are variances there are virtually no apps in which a IO processor outperforms an OOOE processor because of limitations like these, except for multithreaded apps and one having a surplus of IO cores. Executions units are rarely busy because the pipeline is stalled so often, which is a BAD thing because you want them to be busy as possible.

The only real problem with an in order chip is cache misses, which can lead to stalls while loads are fetched. This isn't nearly as big of a problem when you are dealing with an emulator as the primary execution code could be small enough to fit in a Level 1 data cache.

This is most definitely not the only "real" problem, but another one. Actually cache hits are worse for IO than for OOO too because L1 cache has about 2-4 cycle latency and L2 has 10-20ish cycles; Latencies that OOOE processors can mask but an IOE processor can not. And I call your BS that any significant portion of code in an emulator could fucking fit in 32KB of L1 cache. Even zsnes (~500KB) doesn't fit in something that small, nevertheless a fully fledged x86 emulator + Windows kernel and the graphics API. Not even L2 cache could conceivably do it. It's going into main memory here folks.

EPIC doesn't count as while it is IOE it's also a 6-issue core (compared to 2 for the PPC and 3 for Pentiums) with a huge amount of cache (1.5-9MB L3, 256KB L2), and uses the IA-64 instruction set which was designed to lessen it's IOness, and it still loses to Pentiums and Athlons in integer ops.

Itaniums problem with integer performance center around the design philosophy, not it's in order nature. Integer code contains a number of conditional branches which makes it harder for an IA-64 compiler to create larger instruction bundles, which means several of the Itanium execution units might be dormant. This is why compiler performance is so important to the Itanium.

Which by the way is a total red herring. x86 has the same branch conditionals, so does PPC, and so does every other ISA ever made. What you're talking is about the fundemental limits of parallelism and how the IA-64 ISA deals with them. The fact that Itanium simply does not perform that well on integer (given it's resources) is a hit against IO processors at integer in general.

Integer performance is also the same weakness the much slimmer PPC has to face too. For more realistic CPUs, the 50% claim stands. If you dispute the 50% claim, then you dispute the guy who wrote that, not me. Floating point ops will be much better, which is the PPC's strength, but that's useless in emulation so backwards compatibility is still simply impossible.

The person who wrote that admits himself that the performance differences are highly dependent on the situation. Obviously when you are dealing with software that has a great deal of cache access, you are going to have performance loss using an in order design, but this does not include every integer heavy application, especially one custom written for the hardware at a low level the way an emulator is.

The first half is true, the second half came from someplace not credible. Get a link first if your gonna make wild claims about how emulators don't fall under IOE slowdowns.

Face it, backwards compatibility is, if not impossible highly unrealistic barring adding a real x86 chip (plus a proprietary media/IO chip in the Xbox1, but that's another story).

The Kernel wrote:

mplsjocc wrote: Their pride in what they do has kept them in business for close to 100 years. I like Nintendo, and I'm glad that they don't puss out and conform to every quick fad that the casual gamer drives the market towards.

Then you are praising the corporate culture that during the 80's was the worst monopoly this side of Microsoft. They fixed prices, throttled supply in order to keep demand up, and bullied retailers with threats of pulling their products and destroying their business.

You could say similar things about IBM and Apple in the 80's, and now IBM is a strong proponent of OSS and even Apple is running an OS based on NeXTStep and BSD, rather than their old proprietary stuff.

Things change, you know.

Nintendo is one of the leading innovators in the gaming industry today. And certainly not a monopoly. Getting schooled taught them lessons.

Praxis wrote:Anyone think there is a chance of installing Mac OS X on this system with some kind of hack? Perhaps through Mac-on-Linux?

Good luck getting the drivers for the custom hardware

If the XBox 360 has the capability to run Linux very well, and the Revolution does not have that possibility, I may buy both and just hack the XBox for Linux.

Microsoft has Virtual PC for Mac, the biggest problem with porting it to the G5 line is the lack of an instruction to cheaply swap the bit order.

Given Microsoft will have custom-implementation of practically all the core components, this would be trivial to fix. All the graphics calls can be redirected to native code, as well as the sound and any known library calls. And they might use a JIT compiler to convert x86 bytcode on the fly to something else.

This can vastly help in providing backwards compadibility.

ggs wrote:

Praxis wrote:Anyone think there is a chance of installing Mac OS X on this system with some kind of hack? Perhaps through Mac-on-Linux?

Good luck getting the drivers for the custom hardware

If the XBox 360 has the capability to run Linux very well, and the Revolution does not have that possibility, I may buy both and just hack the XBox for Linux.

Microsoft has Virtual PC for Mac, the biggest problem with porting it to the G5 line is the lack of an instruction to cheaply swap the bit order.

Given Microsoft will have custom-implementation of practically all the core components, this would be trivial to fix. All the graphics calls can be redirected to native code, as well as the sound and any known library calls. And they might use a JIT compiler to convert x86 bytcode on the fly to something else.

This can vastly help in providing backwards compadibility.

You'll note the entire arguement above, and the fact that the XBox 360 is NOT using G5s.

And BTW, Virtual PC 7 is already G5 compatible.

The whole Kernel vs HyperionX and me debate has been about whether emulation is possible or not.

Praxis wrote:The whole Kernel vs HyperionX and me debate has been about whether emulation is possible or not.

Ultimately it depends on how Microsoft has implemented the CPU instruction set. Which isnt something we could see without pulling apart an actual final hardware version of the CPU. An emulator isnt good enough.

The problem with Microsoft is that they bought too many parts from people. If they wanted to be backward compatible, they would likely need a Nvidia card, or emulate it. Either way, Nvidia would get royalties, which MIcrosoft doesn't want.

mplsjocc wrote:The problem with Microsoft is that they bought too many parts from people. If they wanted to be backward compatible, they would likely need a Nvidia card, or emulate it. Either way, Nvidia would get royalties, which MIcrosoft doesn't want.

It doesnt matter for backwards compaidibility that everything is perfectly identical.

It is perfectly doable to intercept how the Xbox games manipulated the video card and transform them into Xbox2 calls. There are a few OpenGL -> DirectGraphics and a DirectGraphics -> OpenGL layers around too.

Also WINE takes a similar stratagy

Okay, from what i understand about this backwards compatablitiy issue is that emmulating graphics that were designed for the old Nvidia gpu on the new ATI gpu is nearly impossible. The ammount of processing power required to actually make the emulation even remotely possible just isn't feasible right now. I think that Microsoft is looking at cost effectiveness rather then a larger library of games to choose from. In any instance, if microsoft were to release a backwards compatable Xbox2 wouldn't the cost and overall clock specs would match those of most extremely highend gaming pc. Prices would soar and sales would take a severe hit. The Xbox2 would be the worst flop since the DreamCast. Keep in mind, before I'm flamed, that I know VERY little about this stuff, and this is just a rough idea that someone gave me.

mplsjocc wrote:Okay, from what i understand about this backwards compatablitiy issue is that emmulating graphics that were designed for the old Nvidia gpu on the new ATI gpu is nearly impossible. The ammount of processing power required to actually make the emulation even remotely possible just isn't feasible right now.

Just for record: it still takes power to emulate Glide (3dfx), so I would *really* like to see Micro$oft trying to emulate new nVidia or ATI GPUs.

lol micro$$$$oft the $ means they like to make money lol

Link

Time quotes Bill Gates saying Sony's next-gen console will "walk into" the Xbox 360 follow-up to the best-selling sci-fi shooter.

For months, the Web was rife with rumors that Microsoft would ship the sequel to Halo 2 for the Xbox 360 simultaneously with the launch of the PlayStation 3. Now, that has been confirmed--by none other than Bill Gates himself.

In next week's issue of Time Magazine--which features Gates holding the Xbox 360 on the cover--Gates mentioned plans for the release of the much-anticipated follow-up to the best-selling console shooter. In a parenthetical sidebar in the article, Time says the following:

"Note to the hard-core faithful: the next version of Halo will not, repeat not, be ready in time for the launch of Xbox 360. It will be part of the all-important second wave next spring. 'It's perfect,' Gates says, radiant with bloodlust. 'The day Sony launches [the new PlayStation], and they walk right into Halo 3.'"

While Microsoft reps were not available for comment as of press time, Gates' comments appear to confirm the theory that Microsoft will release Halo 3 either simultaneously or near the release of Sony's next-gen console. It also marks further official confirmation that a third Halo is indeed in the works, and will be released for the Xbox 360.

Although such statements might be interpreted as grandstanding, such a maneuver definitely would have an impact. Halo 2 sold over 6.5 million copies to date, and is credited with doubling the subscriber base of Xbox Live from around 750,000 to over 1.5 million. It is also credited by industry-watchers as one of the main reasons for boosting sales of the original Xbox last holiday season.

Wow, so this really is true. Even the gamespot editors think this will have an impact considering Halo 2 has sold over 6.5 million copies.