Adding a multi-vector processor to X2 a good idea?

People walk into stores and say "whats the most powerful console out and how much is it?

The guy or gal behind the counter says "well PS3 can push 4 billion polygons... while xenon can only do one billion." (of course he/she got this iinfo from gamepro) Based on having awed the consumer with his/her knowledge he finishes the interaction with "Based on power, I'd go with the Ps3."

There is the essence of theoretical numbers, consumer appeal, magazine hype and console makers specs. It all comes down to what spec the sales person can latch onto to make the sale...

 

working at a gamestop through out highschool and still holding a job there for the discount i can honestly say in 8 years i have never heard anyone ask that .

The thing we get asked is which is the most popular and which has the most games for .... whatever age group they ask .

After that we get asked which is the cheapest .

 

hmph... i have stood in either old electronic boutiques and babbages and heard conversations like that (playstation 1 era)

or the consumer would walk in and say "I heard the playstation was the best... is that true?"

maryland must be weird...

 

when they say i heard so and so was the best why do you think they are talking about which is more powerfull ?

I have heard fan boys discussing specs but none of them know what they are talking about and they are normaly 14 years old and already have a console and are not looking to buy one

 

usually the ages are rather higher (college to late twenties).

as far as best versus power... parents and women say "best" while shopping for their kids... older folk not shopping for their kids usually say x console is more powerful than y console so why would i buy y console...?

 

If thats the way you want to describe it i've never heard it asked . No one cares about the power of the consoles unless they are fan boys . But hey believe what you want

Oh and I thought babbages was gone ? I know the only one over here closed down .

 

JVD's comments give a strong hint as to why MS wants to get out first with more games at a cheaper price.

 

Because of this, there are two defined scales. There is horsepower, a 'theoretical peak maximum' if you will, and there's brake-horsepower, or 'real world performance'. Generally cars are listed at bhp ratings, and presence of the bhp qualify confirms that scale is being used.. Similarly we have different measures for temperature. 100 degrees doesn't mean anything until qualified as either Celsius or Fahrenheit. There's nothing wrong giving real information as long as it's qualified. Seems to me those who mess up much of these specs are magazines. We all know how way-off-the-mark internet articles can be when talking console "facts".

As it is, without a universal standard for console performance measurements how do you communicate to your developers what sort of performance they should aim for on your hardware? You could give average figures like Ninty, but that assumes average use. The PS's 66 Mps peak may not apply in most situations, but what if one dev comes up with a novel use for that? By giving peak specs you give the educated developers a chance to see what aspects of performance are possible, and they can work out what level of performance they'd get when choosing what effects to use. Of course the press will get wind of such figures and banter them around anally, which fanboys will regurgitated incessantly in the multitudes of forums spreading the FUD, and I wouldn't be surprised if Sony, MS et al deliberately considered this in their marketting strategies. But 'til now and Allard's claims of 1 teraflop performance for Xenon, obviously a massaged figure to counter KK's Teraflop comments for Cell, I don't know that performance figures have ever been lies but statistics for use by those in the know. I wasn't paying attention to the release of this generation.

I'll add that Sony have a notorious reputation for hyping PS2. I missed the original claims and have searched the web ceaselessly (well, a couple of times!) for what was said by KK. Much of his broken promises were actually subjective and, taken in context of the leap in graphical performance for the time, not too out-there. There were a couple of seemingly bald-faced lies, and I have NO idea what George Lucas was thinking when he said PS2 was rendering Episode 1 in realtime, but the claims against Sony seem exaggerated to me. Seems to me more often than not, the press is to blaim, feeding unvalidated, unqualified data to the public who haven't the interest (or intelligence?) in researching the reality.

 

Of course this is not what happened at all. The first anyone in the media heard figures like 66 mpps was in a press release by Sony. In a press release. Not in dev-kit documentation, and the press release wasn't created to promote the platform to developers. It was created to educate consumers about an upcoming Sony product. This is the same as what Microsoft did. And if Sony can give insane peak figures, why not Microsoft?

Allards "1 teraflop of targeted computing performance" is also a subjective comment that is correct if "taken in context". He did qualify the statement (I'm not sure what "targeted computing performance" means, but if I were a lawyer I'd say that is vague enough to make any statement "correct" i.e. "10 hamburgers of targeted computing performance" could be correct if "taken in context"). If the press is to blame for abusing Sonys hype, then they are also to blame for abusing Microsofts hype, and no blame can be leveled against either company. There is no distinction here. And obviously apologism for one kind of hype alone and condemnation for the opposing hype, is the very definition of bias and fanboyism.

The same argument applies. I'ts a theoretical number that has to be viewed through critical eyes. If someone is misled by these claims, how can Microsoft be to blame? It's the fault of the gaming press and the individual for being so stupid.

Or, maybe it's a ridiculous misleading number that amounts to nothing more than hype. This is the argument I am making, you disagree?

If this is your feeling, then why participate in the conversation?

 

Heh, you can assume 'red' = 'blue' because both are 'color'?

 

In principal I agree with the argument, but I don't think it applies in this context. The claims of 66 mpps for PS2 were 'real' figures, based on conventional, established ways of measuring performance. The 256 gflops/Cell is again based on conventional methods of measuring performance. The '1 Teraflop' of Xenon is not (assuming they haven't got some astounding hardware we don't know of). By all accounts people are looking at a 90 GFlop CPU and maybe 200 GFlop GPU using existing measures. To come with 1 teraflop it looks like MS have dug around and found more figures they could add to their overall rating. That's why they've invented a new term 'targetted performance' to 'qualify' their statement, without explaining what the 'targetted' actually means. If you're going to invent a new measure/scale, you need to explain it. Sony didn't even a new measure for PS2 when they announced it specs and you could measure the hardware's performance to those scales. You couldn't take a Xenon and measure it with existing understanding of Flop performance and gt anything like 1 teraflop, unless as I said MS realy have got some amazing and unknown powerhouse in there! This claim is already showing signs of FUDding gamers. On the CnVG forum, a CnVG news article mentions the '1 teraflop' statement, and in the subsequent discussion someone says 'by comparison your PC maybe runs at 20 GFlops'. Two totally different ways of measuring but as the scale (flops) is the same, people assume the figures are comparable. The 66 mpps is[/is] comparable to a 30 mpps peak of a GForce2 (or whatever figure/card)

As I said before, you need to convey a sense of what your new hardware creation is about. Sony could hardly just say "PS2 will be better" without in somewhere describing how. That'd be silly. They want to get people interested before it comes out and saving their money to buy it. This needs a performance/quality measurement. In the light of a lack of an agreed hardware standard Sony went with peak figures which is as good as any method. I have no qualms with MS claiming performance figures or stats as long as we know what they are. But statistics are dark art, rarely ever truthful. Just as governement's re-work how they record statistics to show, for example, a reduction in unemployment, console companies are looking increasingly to be going to rework their measurements to inflate their figures. The first obvious use of such statistical reworking for me is that XB2 1 teraflop figure, though I am no expert on previous hardware hype and can only comment on this generation from what history I've read. PS2's hardware stats on the other hand were using an existing measure - Sony didn't invent a new way to record Flops so to inflate their values TTBOMK

 

Microsoft is presumably measuring the work resources for the total system of the next Xbox which would rightfully include fixed function floating-point computing power. Programmable floating-point speed is a different specification altogether.

Neither measure fully describes a system's floating-point performance, which would have to take into account how efficiently those calculations could be applied to work loads of variable size and predictability.

 

bbot : Those slides seem to contradict themselves. The slide claims 1 teraflop, but subsequent slides claim 40 (or whatever it was) GFlops for the same processor, while the graphs at the end show something far short of 1000 GFlops in the timeframe of NV40.

This kinda shows the problem faced. What measure do companies use?

 

If you added up all the "fixed functionality" FLOPs in xenon's GPU, I'd expect you'd get a large part beyond just 1 TFLOP.

Like DeanoC said here, people counting just programmable FLOPs have been doing it wrong all along.

The only reason programmable CPU FLOPs has been so concentrated on in the spec wars, is that it is a easy to derive but meaningless number that makes the PS2 (and by extension, PS3) look good.

 

40 gflops seems to be programmable shader flops only on NV40. Which is many times less than what you'll have on next generation parts.

Note the ratio of "programmable FLOPS" (40) to "fixed function FLOPS" (1000) claimed by NVidia. That's a factor of 25 (!!).

 

...and hence is born yet another "crucial" spec to suggest "performance"- the ratio of fixed FLOPs to programmable FLOPs... :l

 

Hmm, how many NURBs were taken?

Maybe as a complete system, but I don't think NV2A could do that alone.(wasn't it downgraded to 100mpps anyhow?)

You work at a gamestop for the games discount?
I'm guessing your in college and this isn't your life career then. I don't think I would have ever pictured you in retail.

Didn't Gamestop eat Babbages and Funcoland? BTW, I always wondered if Babbage's was named after Charles Babbage.

Is it possibly that the entire Xenon system could even have 1 teraflop of performance, or is it just something that was said because it doesn't have to mean anything right now since specs aren't final? Or maybe it's of the "some other hardware would require 1 teraflop of performance to achieve the same results as a dedicated processor we're using" variety. Like maybe an 68000 would have to be capable of 1 teraflop of performance to equal xenon.

 

Well i'm still in college and currently on disability but aside from that i work in a pc store (local) making pcs and figuring out what i want to do with my life . I work at the stop cause i get 20% off and since i know everyone i end up working like 4 hours a week and i normaly spend that money on games plus my sisters , my father , my uncles and my cousins all use my discount .

Yea tahts why i thought they all closed down . I know that babbages were either shut down or renamed gamestop , same goes with funcoland .

I'm from jersey and in one of our malls willow brook , we have a gamestop , a mini gamestop which was a babages and a funcoland now renamed gamestop..... so its kinda nuts

 

Some of the funcolands lasted a bit longer, but babbages seemed to disappear overnight. I always assumed babbages just changed its name to be something a bit hipper and less confusing, but gamestop's policies seem to be some sort of wierd amalgam between babbage's and funcoland's.(or can I still trade in my old nes stuff at gamestop?)