Homebrew CTRX - PS1 Emulator

[Foxi4]

Next he'll be telling us that the 3DS is faster then the Wii U because it uses PPC.

Xenoblade Yawnicles is a downport, not an upport for a reason - the hardware is weaker, spyro just refuses to acknowledge it.

Now, in the interest of this thread I suggest moving any discussion about supposed inferiorities or superiorities of the platforms to its own thread and leave this one to discussion pertaining the amazing PS1 emulator for the 3DS. I myself didn't expect one to pop up so quickly, which is very encouraging.

 

[MyJoyConRunsHot]

Spyro is once again spouting nonsense with no sources to back up his claims, I've explained this to him in different threads before to no effect. The Wii is at least twice as beefy as the 3DS, the N3DS cuts it closer, but has a weaker, mobile GPU.
Before someone accuses me of using an obsolete benchmark, at least it's a benchmark, not empty words - it's a sample of one versus zero samples. A gap like this (155% difference between the 3DS and the Wii) cannot be made up with fancy tricks - stop living the dream, the 3DS is weaker than the Wii, and so is the N3DS which is practically double the processing power of the original 3DS (so about 1319.36 DMIPS, less than the Wii), at least when working in the default 268MHz mode (it can be clocked up to 804MHz on paper, however there are some limitations in reaching that clockspeed).

EDIT: Corrected my own derp, the N3DS is weaker than the Wii, I accidentally compared it to GC's score. Apologies, it's a bit late.

You're once again spreading ignorance over the boards, making baseless claims without presenting evidence? Not to mention you ignore the links and quotes I and other users(biggest contributer Granville, props to him) provided because it came at odds with your ignorance. Plus, up till now, you haven't explained in technical detail as I and others have to the contrary, how the Wii's PPC and hollywood is stronger than that of the 3DS(architecture, pipelines, cache, polycount, shader effects, etc). In this post alone, all you've done is repeat yourself 3 times unnecessarily and then proceed to use DMIPS like it's your holy bible despite me debunking it several threads ago with statements from tech specialists and users alike, one of which outright stated that it's not representative of real world performance. "3DS ez waeker dan Wii, look at teh dimps" is pretty much the extent of your post.

Now I'm not here to engage in a 60 page long beating of a dead horse so I'll start with the low down, lay down some sources, and quote some assessments on the topic by experienced computer hardware . Then you guys can return to your usual gospel whilst the OP proceeds to turn your statements into a textbook case of gbatemp myopia with this project:

Not only is the ARM11 MPcore newer and uses more modern technology for better efficiency, but it also has: more transistors, more/faster cache, faster bus( wider bit bus), OoO completion, smaller nm, Thumb2, multi-core, 4-core VPF, etc. Many, many other superiorities over the ancient PPC in the GCN/Wii. The clockspeed might be a bottleneck but, for the N3DS, thats where the additional cores comes in. And all this is not even mentioning that developers stated the N3DS received a single-threading speed boost over the old 3DS making the speed it processes linear code even faster than it already was.

As for the GPU, that has already been confirmed by developers to be, for the screen size, superior to the Wii's, sporting more modern shaders, lighting and according to the revised 2008 Pica200 spec chart issued by DMP themselves, seems to indicate 40 million triangles at 100Mhz. RE:R demonstrates the graphical capabilities while Ironfall, while sporting similar graphics, also displays up to 40 enemies on screen at a time while maintaining framerate(stated by the devs), smoke/particle/water effects, and explosions all within huge maps. And this is OLD 3DS software. Certainly blows the GCN out of the water.

References:

Spoiler

# of transistors
3DS
http://bit.ly/1MMbX6V
Wii
http://www.anandtech.com/show/858/2
(gamecube, but applies to Wii since both use the same CPU)
Bits
http://pastebin.com/Ybsn0fyA
Pipeline
3DS
http://sandsoftwaresound.net/raspberry-pi/arm11-microarchitecture/
Wii/GC
http://bit.ly/1gT8wgv

You can argue with me, but you can't argue with the more experienced.
Info on poorly made 3DS ports:

Spoiler

Nintendo procrastinated for years before upgrading to actual 21st century hardware architecture. The Gamecube and its PowerPC CPU was still rooted firmly in the 90's era of hardware, and the Wii wasn't a substantial step beyond that either (largely using moderately overclocked components with only very subtle improvements). Same goes with the GBA and DS, which both used much older ARM chips from the 90's. It wasn't until 2011 with the release of the 3DS did they create a piece of tech with 21st century hardware and design, a true multicore CPU (the extra ARM7 CPU in the DS is said to have been largely used as a sound processor in DS games) along with a GPU that could support modern shader effects. It was a new and jarring experience for them i'm certain to step into a field that other console makers and developers had long since become accustomed to (such as Capcom).

If you try to port PowerPC code to an ARM chip, you're going to run into trouble if you take too many shortcuts and don't optimize properly. Even if the ARM chip is STAGGERINGLY more powerful than the original PowerPC CPU. Vita even has even had serious issues with ports of now ancient PS2 games in spite of having hardware that zooms around anything from that generation. And as i've said what seems like hundreds of times before, the GBA was a constant dumping ground for crap SNES/Genesis ports in spite of its own staggering advantage over those consoles. The approach and language for mobile devices is completely different. But it's easier to make the excuse that the hardware is at fault than admit you and your team are either unable or unwilling to put the time and effort in to reprogram it properly. Lest you get called out for poor porting like Konami with Snake Eater 3D (especially when Kojima had been hyping up the port and claiming it was built from the ground up for 3DS).

Capcom and VDDev have both approached 3DS development the proper way by custom building an engine around the 3DS' architecture and have provided a taste for what the system can do (well beyond Wii). Even Capcom's early and lazier efforts took better advantage of the hardware than most other games on the system (they improved their engines by leaps and bounds for Revelations and MH4). New 3DS is at an even better advantage due to the even greater memory, vram and two additional CPU cores. A poor Wii port from Monster Games however isn't going to give anyone a decent idea of what those advantages are.

The straight number comparisons don't show the entire picture...what we are missing is a look at X3DS's code. You see, the extra performance you gain with a multicore processor depends on the software algorithms used, and how they're implemented. Xenoblade Wii, which was running on that system's single, more powerful core, would have been written largely as a string of linear code. Multicore processors are limited by the fraction of the current task that can be run parallel on different cores. It is theoretically possible for the developer to translate the code in such a way that 90%+ of the code could be run parallel. This would truly take advantage of the multicore processor. If, however, the developer translates the code in a way that only 50% can be run parallel, then half the time, you're only using the power of one core anyway. There's no way for us to see the code Monster Games has written (or copied), so we can't see how they've split it up between cores. Remember though, it's possible to program the old Xenoblade in a way that makes use of the extra power the 3DS has, but it would need to be built more or less from the ground up (not,easy...possible). I agree with grans, it's not a hardware limitation, it's on the devs.

 

[Deleted User]

A once-over of 3dbrew as well as other reports says the clockspeed wasn't increased, only the number of cores and singlethreading.

This was actually found to not be true. The N3DS allows CIAs to specify whether to use 268mhz or 804mhz.

 

[BlueOnClues]

A new level of 3DS homebrew is beginning. Man, keep up the work! I thought it would've been impossible for a PS1 emulator, but you seemed to do it.

Here's a dancing Crash for good luck!​

looks like crash is single ready to mingle ( he's banging )

 

[teampleb]

This was actually found to not be true. The N3DS allows CIAs to specify whether to use 268mhz or 804mhz.

I don't even feel like arguing with him. He's getting anal about this. He was wrong about the 3DS's GPU being faster too. The 3DS outputs 15.3 mil polygons a second. The Wii is triple that in most circumstances.

 

[doctorgoat]

http://all-things-andy-gavin.com/video-games/making-crash/

crash bandicoot is unbelievably complex, invoking him is dangerous

 

[Foxi4]

You can argue with me, but you can't argue with the more experienced.

Here's where you're wrong - nobody can argue with you because you're consistently not listening to anything anyone is saying unless it fits your narrative.

The CPU is more modern, it is efficient and it does have OoO execution as well as other bells and whistles, but it's running at a low clockspeed because the 3DS is a portable and battery life is more important than raw performance in this context. The fact that the CPU is multicore puts additional obstacles in your way since parallelism is always wasteful - it's not an advantage, it's a necessary evil as we're nearing the end of what can be feasibly done with silicon.

The N3DS could only "eclipse" the Wii when running full-blast in "804MHz mode", or at least at a higher clock than the default 268MHz, that's a fact, I don't really care if you feel like disputing it or not because nothing is at stake.

What you're showing as evidence are low-level factoids and maximum values that have little to do with real-life performance. The console simply does not utilize the CPU to the fullest because it's designed to be portable, that's why the cores are running at clocks lower than standard - it's done on purpose and by design.

Now, if you would kindly stop derailing the thread, it'd be really nice. If you want to discuss this further, take it to a dedicated thread.

 

[granville]

Pica200 draws a tad over 15 million polygons at 200mhz, the 3DS variant is 268mhz to match the CPU. That's a moderate boost to the polygon count. 3DS is also very low resolution. Even the second screen (itself low res and usually only displaying 2D graphics) and 3D effect together isn't adding up to many pixels. We don't know what resolution the Pica200 was originally benched at, but it was almost certainly much higher than the 3DS supports. The Pica200 Samurai tech demo ran at a substantially higher resolution than the 3DS can dish out-


Polygon counts benched at high resolutions aren't usually accurate whatsoever of what a GPU is capable of at lower resolutions. Substantially lowering the resolution drastically improves overall performance elsewhere, including the amount of polygons and other effects rendered at once.

It's usually a bad idea to pass judgement on hardware based on ports. It's not always the case, but the vast majority of handheld ports are downgraded even when there's a substantial advantage in power. GBA received a ton of SNES and Genesis ports, Vita got a number of PS2 ports. Pretty much across the board however, both handhelds consistently received downgraded versions of games from those consoles. I struggle to think of a case where a handheld port released without inferior audio, visuals and/or performance. GBA was quite a boost over both SNES and Genesis, and Vita's advantage over PS2 shouldn't need explaining. Intact handheld ports, let alone upgraded ones, are insanely rare. With Xenoblade, performance actually turned out ok, with very few isolated instances of removed scenery. But it's a mystery why the textures were heavily downgraded with New 3DS' huge ram and vram advantage over Wii (3DS' ram isn't even the slow kind often used in mobile devices, it's FCRAM)...

The 3DS has done some impressive things (given the hardware) in good hands, enough to at least rival Wii. VDDev's Ironfall being a particularly impressive technical achievement. Though even Capcom took a surprising role early on in pushing the hardware. Many if not most other developers have struggled to get even low end Dreamcast quality visuals running on the system sadly.

 

[teampleb]

3D effect together isn't adding up to many pixels.

Except for the fact it has to render the scene twice in 3D mode.

 

[granville]

Except for the fact it has to render the scene twice in 3D mode.

In my experience with gaming in 3D on PC, it has an impact sure, but it's not so cut and dry as "you have to render everything twice". It doesn't tend to cut performance in half. I've found it much more strenuous to render a game at much higher resolutions and with other visual effects applied.

And there are ways to deal with this regardless, one example being disabling certain visual effects (anti aliasing for one) or lowering the framerate. These solutions were employed by VDDev and Capcom in their 3DS games, unfortunately mostly ignored by other developers (other games with framerate problems don't generally improve anything by disabling 3D).

Developers also have no obligation whatsoever to use the 3D effect in the first place, and they're perfectly free to develop games with it disabled. I'm aware of multiple games that did this, most of which aren't even anywhere close to impressive as other 3DS games that do support 3D. Final Fantasy Explorers for instance is a recent 3DS game that doesn't support 3D, far less impressive than Capcom's efforts with Monster Hunter (which do support 3D, far better framerates and superior visuals). The most technically impressive 3DS games (that rival and in some ways surpass Wii's top work) do actually support 3D, funny that most of the 3DS games that don't support 3D are very unimpressive for the hardware.

 

[sarkwalvein]

Stop derailing the thread with speculative BS, hell, this looks like a gambling house full of know-it-alls (cake taken by spyro and grannyville).
This thread is about an experimental ps1 emulator.
Go create a thread about n3ds master race unused specs speculation and take your shit there. With a tinfoil hat.

 

[heartgold]

What's this about N3DS core running at 804MHz, that's news to me.

Is that the games has access to 3 cores x 268MHz?

 

[granville]

Stop derailing the thread with speculative BS, hell, this looks like a gambling house full of know-it-alls (cake taken by spyro and grannyville).
This thread is about an experimental ps1 emulator.
Go create a thread about n3ds master race unused specs speculation and take your shit there. With a tinfoil hat.

Nothing I said was speculation, they were facts about how GPU's work relative to clock speed and resolution. Tone down the antagonism, hardware performance is quite relevant in a discussion about the viability of emulation...

And master race? I can't speak for anyone else, but I don't own or want a n3ds. I know some people are busy growing boners for it lately, not one of them. I'm of course playfully joking about people liking their New 3DS, always fun to see the fruits of hacking and homebrew (I joined this site during the heyday of original DS hacking/homebrew). But personally a Nintendo system with only one exclusive game (and nothing further announced) holds no interest for me. Let alone a port.

I own an original 3DS, and I haven't bothered to try homebrew. I only ever use it for its intended purpose- playing native DS and 3DS games. And lately i've been mostly PC-centric in gaming/emulation. Or an Android Phone with a good pocket sized bluetooth controller for handheld emulation.

I am interested to see how this PS1 project progresses. I probably won't try it myself when it matures, but the technical merits are always fun to watch regardless.

What's this about N3DS core running at 804MHz, that's news to me.

Is that the games has access to 3 cores x 268MHz?

Been a while since I posted, surprised to see my name pop up in this thread.

Apparently something like that. While that's not usually how multicore processors work from what i'm aware, there seems to be an exception when dealing with the ARM11 inside the N3DS. Last I read, people had noticed performance increases in running single threaded homebrew apps on their New 3DS. Since the homebrew supposedly wasn't accessing the other cores, this initially lead people to think that New 3DS had upped the clock speed or that IPC had been enhanced in some other way.

But more recently I believe the thought is that the New 3DS's ARM11 somehow automates the multithreading process, combining the performance of the three available ARM11 cores together into one.

 

[heartgold]

Been a while since I posted, surprised to see my name pop up in this thread.

Apparently something like that. While that's not usually how multicore processors work from what i'm aware, there seems to be an exception when dealing with the ARM11 inside the N3DS. Last I read, people had noticed performance increases in running single threaded homebrew apps on their New 3DS. Since the homebrew supposedly wasn't accessing the other cores, this initially lead people to think that New 3DS had upped the clock speed or that IPC had been enhanced in some other way.

But more recently I believe the thought is that the New 3DS's ARM11 somehow automates the multithreading process, combining the performance of the three available ARM11 cores together into one.

It it unusual, but that's how it seems to work. Practically there shouldn't be any boost on software running on a single thread as the documented performance is the same per core on the O3ds and N3ds, only the number of cores have increased.

This is very positive, N3DS combining the cores together gives it a better performance than coding for 3 individual cores.

I assume the 4th core is reserved for the OS and 3D eye tracking?

 

[DutchyDutch]

-snip-

Thank you. Watch out though, he's gonna call you a troll..

 

[MyJoyConRunsHot]

Here's where you're wrong - nobody can argue with you because you're consistently not listening to anything anyone is saying unless it fits your narrative.

The CPU is more modern, it is efficient and it does have OoO execution as well as other bells and whistles, but it's running at a low clockspeed because the 3DS is a portable and battery life is more important than raw performance in this context. The fact that the CPU is multicore puts additional obstacles in your way since parallelism is always wasteful - it's not an advantage, it's a necessary evil as we're nearing the end of what can be feasibly done with silicon.

The N3DS could only "eclipse" the Wii when running full-blast in "804MHz mode", or at least at a higher clock than the default 268MHz, that's a fact, I don't really care if you feel like disputing it or not because nothing is at stake.

What you're showing as evidence are low-level factoids and maximum values that have little to do with real-life performance. The console simply does not utilize the CPU to the fullest because it's designed to be portable, that's why the cores are running at clocks lower than standard - it's done on purpose and by design.

Now, if you would kindly stop derailing the thread, it'd be really nice. If you want to discuss this further, take it to a dedicated thread.

Alright, this made my day. Let's keep this fairly short so this thread doesn't morph into another assumption-based preach session, laden with ad hominems.

What you said about multi-core setups only applies to your typical Monster Games fuzzy text ridden, copy-paste jobs like Xenoblade Chronicles, which are best at home on older, relatively less sophisticated CPU from 1999-2000 like the Wii's. It is common knowledge in the hardware community that some CPUs can crunch more numbers per cycle and output more instructions at lower clockspeeds if designed well, this exception can range from quirks such as increased/speedy cache, an efficient bus system(resulting from newer or smartly designed architecture), better pipelines, among other perks like Thumb2 that can contribute to enhance performance. Case in point, the Xbox 360, which runs at a pretty sweet 3.2Ghz vs. the Xbox One at 1.75Ghz. Pretty sure it goes without saying which one is more powerful considering they are separate console generations so you can speculate for yourself. Another example is the PSP which ran at an awesome 333Mhz at the end of it's lifespan vs. the PS2 which ran at a measly 299Mhz in the newer models. While it is most likely attributed to developer laziness, PSP appeared to have a difficult time running PS2 ports, it's original games also didn't hold a candle to what was on PS2 and was frequently likened to Dreamcast instead. Compared to PSP, 3DS handles some ports well for the most part like what we have seen of HW so far, and features games running to the caliber of what can be found on Wii, see Ironfall.

Whether the N3DS's threading enhancements were automated or real or something else entirely, people including developers did witness increased performance when running software. That, coupled with the quad-core can deliver substantial performance benefits. This article highlights the benefits of multi-core setups over lesser core setups:
http://searchitchannel.techtarget.com/feature/Multi-core-processor-benefits-and-trends

The sources I provided were directly connected to real life performance. On occasion it only connects to on paper performance and much of the library points to this, but what we have seen from top tier games, tech demos and emulators so far looks promising. Funnily enough, it was your Dhrytone that is frequently on the chopping block by techies, as is your perpetually cited "mobile vs. fullscale" excuse.

Back off with your pitchfork and let the OP and co. run the thread the way it was intended.

Edit: bit long but gets to the points.

 

[SLiV3R]

Alright, this made my day. Let's keep this fairly short so this thread doesn't morph into another assumption-based preach session, laden with ad hominems.

What you said about multi-core setups only applies to your typical Monster Games fuzzy text ridden, copy-paste jobs like Xenoblade Chronicles, which are best at home on older, relatively less sophisticated CPU from 1999-2000 like the Wii's. It is common knowledge in the hardware community that some CPUs can crunch more numbers per cycle and output more instructions at lower clockspeeds if designed well, this exception can range from quirks such as increased/speedy cache, an efficient bus system(resulting from newer or smartly designed architecture), better pipelines, among other perks like Thumb2 that can contribute to enhance performance. Case in point, the Xbox 360, which runs at a pretty sweet 3.2Ghz vs. the Xbox One at 1.75Ghz. Pretty sure it goes without saying which one is more powerful considering they are separate console generations so you can speculate for yourself. Another example is the PSP which ran at an awesome 333Mhz at the end of it's lifespan vs. the PS2 which ran at a measly 299Mhz in the newer models. While it is most likely attributed to developer laziness, PSP appeared to have a difficult time running PS2 ports, it's original games also didn't hold a candle to what was on PS2 and was frequently likened to Dreamcast instead. Compared to PSP, 3DS handles some ports well for the most part like what we have seen of HW so far, and features games running to the caliber of what can be found on Wii, see Ironfall.

Whether the N3DS's threading enhancements were automated or real or something else entirely, people including developers did witness increased performance when running software. That, coupled with the quad-core can deliver substantial performance benefits. This article highlights the benefits of multi-core setups over lesser core setups:
http://searchitchannel.techtarget.com/feature/Multi-core-processor-benefits-and-trends

The sources I provided were directly connected to real life performance. On occasion it only connects to on paper performance and much of the library points to this, but what we have seen from top tier games, tech demos and emulators so far looks promising. Funnily enough, it was your Dhrytone that is frequently on the chopping block by techies, as is your perpetually cited "mobile vs. fullscale" excuse.

Back off with your pitchfork and let the OP and co. run the thread the way it was intended.

Yeah, that was really short

 

[MyJoyConRunsHot]

Yeah, that was really short

Alright, this made my day. Let's keep this fairly short so this thread doesn't morph into another assumption-based preach session, laden with ad hominems.

 

[Rj.MoG]

Alright, this made my day. Let's keep this short so this thread doesn't morph into another assumption-based preach session, laden with ad hominems.

What you said about multi-core setups only applies to your typical Monster Games fuzzy text ridden, copy-paste jobs like Xenoblade Chronicles, which are best at home on older, relatively less sophisticated CPU from 1999-2000 like the Wii's. It is common knowledge in the hardware community that some CPUs can theoretically crunch more numbers per cycle and output more instructions at lower clockspeeds if designed well, this exception can range from quirks such as increased/speedy cache, an efficient bus system(resulting from newer or smartly designed architecture), better pipelines, among other perks like Thumb2 that can contribute to enhance performance. Case in point, the Xbox 360, which runs at a pretty sweet 3.2Ghz vs. the Xbox One at 1.75Ghz. Pretty sure it goes without saying which one is more powerful considering they are separate console generations so you can speculate for yourself. Another example is the PSP which ran at an awesome 333Mhz at the end of it's lifespan vs. the PS2 which ran at a measly 299Mhz in the newer models. While it is most likely attributed to developer laziness, PSP appeared to have a difficult time running PS2 ports, it's original games also didn't hold a candle to what was on PS2 and was frequently likened to Dreamcast instead. Compared to PSP, 3DS handles some ports well for the most part like what we have seen of HW so far, and features games running to the caliber of what can be found on Wii, see Ironfall.

Whether the N3DS's threading enhancements were automated or real or something else entirely, people including developers did witness increased performance when running software. That, coupled with the quad-core can deliver substantial performance benefits. This article highlights the benefits of multi-core setups over lesser core setups:
http://searchitchannel.techtarget.com/feature/Multi-core-processor-benefits-and-trends

The sources I provided were directly connected to real life performance. On occasion it only connects to on paper performance and much of the library points to this, but what we have seen from top tier games, tech demos and emulators so far looks promising. Funnily enough, it was your Dhrytone that is frequently on the chopping block by techies, as is your perpetually cited "mobile vs. fullscale" excuse.

Back off with your pitchfork and let the OP and co. run the thread the way it was intended.

Tldr. Also please tell, where did you get your degree in Hardware/Electrical Engineering from that makes you such an expert on gate type speed spyro? And you say less sophisticated CPU's without understanding that ARM models use an instruction set that lacks certain SIMD instructions that even the Pentium Pro had.... This isn't to say it's weaker, just very specialized.. It seems you confuse the two. This is the same case with the Xbone vs the Xbox 360. The 360 cpu is in fact stronger in MANY ways than the Xbox 1 CPU. Just not in any way that matters. In ancedote the PSP lacked the bus speed that the PS2 had despite being more powerful in other aspects. The Bus Speed of the PS2 really saved it. TLDR; Stop posting about things you aren't educated on.

 

[DutchyDutch]

Alright, this made my day. Let's keep this fairly short so this thread doesn't morph into another assumption-based preach session, laden with ad hominems.

What you said about multi-core setups only applies to your typical Monster Games fuzzy text ridden, copy-paste jobs like Xenoblade Chronicles, which are best at home on older, relatively less sophisticated CPU from 1999-2000 like the Wii's. It is common knowledge in the hardware community that some CPUs can crunch more numbers per cycle and output more instructions at lower clockspeeds if designed well, this exception can range from quirks such as increased/speedy cache, an efficient bus system(resulting from newer or smartly designed architecture), better pipelines, among other perks like Thumb2 that can contribute to enhance performance. Case in point, the Xbox 360, which runs at a pretty sweet 3.2Ghz vs. the Xbox One at 1.75Ghz. Pretty sure it goes without saying which one is more powerful considering they are separate console generations so you can speculate for yourself. Another example is the PSP which ran at an awesome 333Mhz at the end of it's lifespan vs. the PS2 which ran at a measly 299Mhz in the newer models. While it is most likely attributed to developer laziness, PSP appeared to have a difficult time running PS2 ports, it's original games also didn't hold a candle to what was on PS2 and was frequently likened to Dreamcast instead. Compared to PSP, 3DS handles some ports well for the most part like what we have seen of HW so far, and features games running to the caliber of what can be found on Wii, see Ironfall.

Whether the N3DS's threading enhancements were automated or real or something else entirely, people including developers did witness increased performance when running software. That, coupled with the quad-core can deliver substantial performance benefits. This article highlights the benefits of multi-core setups over lesser core setups:
http://searchitchannel.techtarget.com/feature/Multi-core-processor-benefits-and-trends

The sources I provided were directly connected to real life performance. On occasion it only connects to on paper performance and much of the library points to this, but what we have seen from top tier games, tech demos and emulators so far looks promising. Funnily enough, it was your Dhrytone that is frequently on the chopping block by techies, as is your perpetually cited "mobile vs. fullscale" excuse.

Back off with your pitchfork and let the OP and co. run the thread the way it was intended.

Tl;dr
Also if the 3DS is so much stronger than the wii, why does smash 3DS look worse polygon wise and runs worse (30 fps PKMN) than Brawl?Don't even try telling me that the game wasn't well made or shit like that.