Support NeoGAF

[blu]

I think a lot of people would look at an ARM based console as a huge misstep when the entire gaming industry (besides mobile) is x86.
<snip>

Fundamental difference between ARM and x86 is ARM is RISC and x86 is CISC.

Complex Instruction Set Computer (CISC) processors, like the x86, have a rich instruction set capable of doing complex things with a single instruction. Such processors often have significant amounts of internal logic that decode machine instructions to sequences of internal operations (microcode).

RISC architectures, in contrast, have a smaller number of more general purpose instructions, that might be executed with significantly fewer transistors, making the silicon cheaper and more power efficient. Like other RISC architectures, ARM cores have a large number of general-purpose registers and many instructions execute in a single cycle. It has simple addressing modes, where all load/store addresses can be determined from register contents and instruction fields.

Still with me? Both architectures are capable. RISC is better suited for low power consumption CISC is better suited for high power computation. RISC is common for phones/tablets CISC is common for PC/Console gaming. There is a lot of work that needs to be done to convert a game from running on a CISC architecture to RISC. Developers need to write code to make things work the same way on the different chips. This takes time and developers are expensive. When EA is making a game for PS4 they don't have to spend a lot of time to make it work on XBONE or PC because the processor they are running on are fundamentally the same. When they decide whether or not to port to Nintendo or Tablet/Phone they have to decide if spending a bunch of time and/or money to make it run on fundamentally different processor is worth how much money they will make off of the sales. With a low install base this will be Wii U all over again, it won't be financially viable to spend money on writing the code to convert the game. Thats about as simple as a way as i can think to explain it.

Pardon me interfering, but what's your background, again?

 

[theangrypirate]

Pardon me interfering, but what's your background, again?

Computer Science graduate, Professional Software Developer for a major gaming company.

 

[blu]

Computer Science graduate, professional software developer for a major gaming company.

Good. Have you actually seen the instruction sets of amd64 and arm64? Next question: have you written a single line of portable code, as in for anything else but amd64 in your life?

 

[Pokemaniac]

I think a lot of people would look at an ARM based console as a huge misstep when the entire gaming industry (besides mobile) is x86. Also the fact that Nintendo is doing exactly the same thing as the Wii U as far as being behind current gen consoles with respect to power as another bad sign. See below chart:



This is a 30 year trend that is impossible to dismiss.



Fundamental difference between ARM and x86 is ARM is RISC and x86 is CISC.

Complex Instruction Set Computer (CISC) processors, like the x86, have a rich instruction set capable of doing complex things with a single instruction. Such processors often have significant amounts of internal logic that decode machine instructions to sequences of internal operations (microcode).

RISC architectures, in contrast, have a smaller number of more general purpose instructions, that might be executed with significantly fewer transistors, making the silicon cheaper and more power efficient. Like other RISC architectures, ARM cores have a large number of general-purpose registers and many instructions execute in a single cycle. It has simple addressing modes, where all load/store addresses can be determined from register contents and instruction fields.

Still with me? Both architectures are capable. RISC is better suited for low power consumption CISC is better suited for high power computation. RISC is common for phones/tablets CISC is common for PC/Console gaming. There is a lot of work that needs to be done to convert a game from running on a CISC architecture to RISC. Developers need to write code to make things work the same way on the different chips. This takes time and developers are expensive. When EA is making a game for PS4 they don't have to spend a lot of time to make it work on XBONE or PC because the processor they are running on are fundamentally the same. When they decide whether or not to port to Nintendo or Tablet/Phone they have to decide if spending a bunch of time and/or money to make it run on fundamentally different processor is worth how much money they will make off of the sales. With a low install base this will be Wii U all over again, it won't be financially viable to spend money on writing the code to convert the game. Thats about as simple as a way as i can think to explain it.

You might have something resembling a point if games were still written mainly in assembly. They (along with most other software) simply aren't anymore. Nowadays we have these things called compilers that take care of the vast majority of the work involved with porting things between CPU architectures.

 

[ozfunghi]

Thats a good point. There are plenty of reasons to be optimistic about the potential of the Switch IMO. Selling gangbusters right out of the gate is essential to getting third party support and third party support is essential to this thing doing well.

Well, selling gangbusters has a better chance for Switch than WiiU. Remember, this will be the ONLY Nintendo device. That means Nintendo can focus on one device and the software output will be bigger than it was for either WiiU or 3DS. That might get it in people's homes/pockets faster early on, again boosting interest from 3rd parties.

Good. Have you actually seen the instruction sets of amd64 and arm64? Next question: have you written a single line of portable code, as in for anything else but amd64 in your life?

Now i understand why you changed your avatar last year.

 

[jdstorm]

Not a tech guy, but wasnt Nvidia designing the Denver CPU to be able to easily translate between various instruction sets? Did they ever get it working?

Also isnt it more about porting an engine then porting individual games? Ie Say EA makes one Frostbite game run on the NS or Epic makes one Unreal 4 game run on the NS. Would that essentially make most future games using that engine reasonably simple/low effort.

 

[theangrypirate]

Good. Have you actually seen the instruction sets of amd64 and arm64? Next question: have you written a single line of portable code, as in for anything else but amd64 in your life?

Ive written assembly language code for RISC and CISC processors. Ive written tons code in a multitude of high level languages. Java, for example is portable. It runs on any device with a JVM. Games aren't typically written in java because its hard to take advantage of specific hardware when the VM is meant to run anywhere. What point are you trying to get at?

 

[ozfunghi]

Not a tech guy, but wasnt Nvidia designing the Denver CPU to be able to easily translate between various instruction sets? Did they ever get it working?

Also isnt it more about porting an engine then porting individual games? Ie Say EA makes one Frostbite game run on the NS or Epic makes one Unreal 4 game run on the NS. Would that essentially make most future games using that engine reasonably simple/low effort.

I think UE4 has been confirmed to be running on the NS.

 

[Randdalf]

Given how ubiquitous ARM these days is for mobile, I'd be very surprised if the major compilers (Microsoft, Clang, GCC) don't support it at least as well as they do x86. I think it's just a matter of engine and tooling support. Unity runs on both ARM and x86 tablets, I know that. Getting your own code to build on a certain architecture is probably less of an issue than wrangling third party vendors to do the same.

I do believe you'd need to re-vectorise your code since ARM and x86 have different SIMD instruction sets. I actually think No Man's Sky had issues with SIMD on AMD vs Intel processors?

 

[blu]

Ive written assembly language code for RISC and CISC processors.

You didn't answer my first question so I'll have to repeat it: have you seen the instruction sets of amd64 and arm64?

Ive written tons code in a multitude of high level languages. Java, for example is portable. It runs on any device with a JVM. Games aren't typically written in java because its hard to take advantage of specific hardware when the VM is meant to run anywhere. What point are you trying to get at?

So, going outside of your Java experience, let me restate my second question: have you written anything in the C/C++ family (you know, the languages of choice for most performance-sensitive apps, games included) that had to run on anything else but amd64?

 

[Not]

I have no idea how I'm going to afford this, Breath of the Wild, and a 1080p tv.

Someone help

 

[theangrypirate]

You didn't answer my first question so I'll have to repeat it: have you seen the instruction sets of amd64 and arm64?

Yes

So, going outside of your Java experience, let me restate my second question: have you written anything in the C/C++ family (you know, the languages of choice for most performance-sensitive apps, games included) that had to run on anything else but amd64?

Yes

 

[blu]

Yes

Mind explaining the differences between the two to a layman?

Yes

Oh, what was it, if you're at liberty to discuss it?

 

[ggx2ac]

Not a tech guy, but wasnt Nvidia designing the Denver CPU to be able to easily translate between various instruction sets? Did they ever get it working?

They couldn't get Denver to translate both x86 and ARM because of Intel patents blocking x86.

Talked about already in another thread that Denver may be out then because it looks like they are building an ARM CPU from the ground up for the Xavier SoC since it mentions custom 8-core ARM64 chip.

It's likely not going to be in the Switch so still waiting to find out what ARM CPU configuration Nintendo is using.

 

[bomma_man]

I don't know if that's actually true anymore, believe it or not. At least at this time.

The amount of money Rockstar has made off GTA V sales and online DLC alone is unfathomable.

Sure Trevor isn't going to be popping up at any Olympic closing ceremonies any time soon, but I'm thinking strictly in terms of dollars here.

I mean, in raw dollar terms, GTA V is probably the highest grossing piece of entertainment in history right? An incredibly rough calculation, not including sales, currency variation and GTA Online puts it at over $4,000,000,000.00.

 

[Dark Cloud]

I have no idea how I'm going to afford this, Breath of the Wild, and a 1080p tv.

Someone help

I'll give you 5 rupees.

 

[theangrypirate]

Mind explaining the differences between the two to a layman?

I find explaining machine level operations to the layman difficult regardless of the architecture. But here goes:

Arm64 instruction set is very close to microcode (cpu operation). amd64 translates a complex instruction into a 3 or 4 operations or lines of microcode. For example an amd64 CPU can be told to add together two numbers stored in main memory. To do this, the CPU needs to fetch the number from memory address1 (op 1), fetch the number from memory address2 (op 2), add the two numbers (op 3). arm64 instructions are less complex so each instruction is closer to its microcode. For arm64, you use the instruction set to put memory address into registers then execute the addition instruction. AMD -> one line to add two numbers together ARM -> 3 or 4 lines to add two numbers together.

Oh, what wast it, if you're at liberty to discuss it?

My last experience with embedded C was programming MH6211 to work with fire sensors for alarms. It was a learning experience and I'm happy to not do that work anymore

 

[Kilrogg]

Slow down there turbo, nobody said a trend was a law. Looking only at the chart anybody with half a brain can see the sales for a nintendo console over the entire history of their business are worse than the previous generation. The lone exception is the Wii. In statistics/econ you can determine the average slope over all the data points in this chart to predict where the next data point will be. Its called science and if you don't trust thats fine there are plenty of people who don't believe in it. Statistics are math and they have to to with probabilities. This data shows that Nintendo's next console "probably" wont do well. There is still a chance it will be wildly successful like the Wii. I wouldn't bet on it and the people who make money off of these things aren't betting on it either. Thats why Nintendo stocks fell after the announcement.

Sorry if I sounded overly aggressive. It's just that I've seen people use similar graphs elsewhere as some kind of proof that "Nintendo is doomed" and it just feels so lazy.

I believe in statistics. I just believe they should be manipulated with care and be thoroughly analyzed when trying to determine something as complex as whether or not a major consumer electronics device launching across many countries, by a long-standing manufacturer with a history of changing its strategy often in an industry where the market tends to "reset" every few years will be a failure or not.

This is one of those cases where the subject of the analysis has proven wildly unpredictable, alternating between market-leading success and miserable failure in the course of a few years. If Nintendo did the same thing every generation by simply upgrading every metric along the same axis, you'd have a solid case. But when you're dealing with a king of weirdness and unpredictability like Nintendo? In such an unpredictable market as the console market? I wouldn't let data like that graph guide any predictions.

Spoiler

Also, I just noticed the chart is from vg chartzzzzz... Doesn't change the trend, but... Ugh.

 

[MAtgS]

I have no idea how I'm going to afford this, Breath of the Wild, and a 1080p tv.

Someone help

You only got one option:

 

[theangrypirate]

Sorry if I sounded overly aggressive. It's just that I've seen people use similar graphs elsewhere as some kind of proof that "Nintendo is doomed" and it just feels so lazy.

I believe in statistics. I just believe they should be manipulated with care and be thoroughly analyzed when trying to determine something as complex as whether or not a major consumer electronics device launching across many countries, by a long-standing manufacturer with a history of changing its strategy often in an industry where the market tends to "reset" every few years will be a failure or not.

This is one of those cases where the subject of the analysis has proven wildly unpredictable, alternating between market-leading success and miserable failure in the course of a few years. If Nintendo did the same thing every generation by simply upgrading every metric along the same axis, you'd have a solid case. But when you're dealing with a king of weirdness and unpredictability like Nintendo? In such an unpredictable market as the console market? I wouldn't let data like that graph guide any predictions.

Spoiler

Also, I just noticed the chart is from vg chartzzzzz... Doesn't change the trend, but... Ugh.

Im pulling for Nintendo here. I just hope they aren't making the same mistakes as they did with the Wii U. Can't wait til January so we can find out more. I have a feeling the launch window software lineup is going to be extraordinary. That more than anything will sell the Switch. And damnit, I didn't realize that was vg chartz :/

 

[Lord Azrael]

You know what other numbers can't be ignored? Nintendo's handheld sales. It doesn't matter that Nintendo says the Switch is primarily a home console. What matters is what the public perceives it to be. So if people see this as a replacement to the 3DS as well as the Wii U, buyers who traditionally only buy Nintendo's handhelds may also buy the Switch.

 

[Kilrogg]

Im pulling for Nintendo here. I just hope they aren't making the same mistakes as they did with the Wii U. Can't wait til January so we can find out more. I have a feeling the launch window software lineup is going to be extraordinary. That more than anything will sell the Switch. And damnit, I didn't realize that was vg chartz :/

Ever since NPD decided to be stuck up assholes it's become harder and harder to get actual reliable data. And when you don't get actual reliable data, VG Chartz wins by default, sadly.

As for hoping Nintendo doesn't make the same mistakes as before, here's how I see it:

- on the hardware front, pretty much everything I've seen so far, from the actual hardware to the way Nintendo has been communicating on the thing, the Switch is already way better than the Wii U. Like, it's not even a contest. They've got a clear message, a clear defining feature that can't be construed as a gimmick but is simply a way to make gaming more convenient - and convenience always trumps everything else. And while the hardware is once again less powerful than the competitors', it seems in the same ballpark, with exactly the kind of middleware support that was sorely missing from previous systems, and no weird feature that developers will feel forced to tailor their games for. Also, the hybrid concept means the system will get the support of at least some handheld-focused developers. That can only be good.

- on the software front? Well, it's up in the air. Third party support is hard to gauge, and as for first-party games, we've only seen Wii U ports so far. Games like those won't do much for the sales. We'll have to wait and see what kinds of games Nintendo makes for the system past launch. Way I see it, this will be the one key factor to Nintendo's success/failure this gen. Even more so than third-party support. When Nintendo gets its head out of its ass, its systems get killer app after killer app (see Wii Sports, Wii Fit, Mario Kart DS and Wii, NSMB DS and Wii, Nintendogs and Brain Training). They'll need a few of those for the Switch as well. The Wii U and even the 3DS never really got that, beyond maybe Mario Kart. One thing worthy of note for third parties though: with DQ XI making its way to the Switch, and Monster Hunter likely appearing too, the Switch is already very well positioned to succeed in Japan, meaning more support from Japanese developers.

Overall, I think your fears are largely unfounded. I'm sure Nintendo will screw up in a couple areas at least, but their strategy seems already so much more sound this time around than back when the Wii U and the 3DS came out that it's not even funny.

 

[ozfunghi]

You know what other numbers can't be ignored? Nintendo's handheld sales. It doesn't matter that Nintendo says the Switch is primarily a home console. What matters is what the public perceives it to be. So if people see this as a replacement to the 3DS as well as the Wii U, buyers who traditionally only buy Nintendo's handhelds may also buy the Switch.

Yup, i made somewhat the same argument earlier.

 

[theangrypirate]

You know what other numbers can't be ignored? Nintendo's handheld sales. It doesn't matter that Nintendo says the Switch is primarily a home console. What matters is what the public perceives it to be. So if people see this as a replacement to the 3DS as well as the Wii U, buyers who traditionally only buy Nintendo's handhelds may also buy the Switch.

You're absolutely right on that. The argument against your case would be mobile phones now. They are more ubiquitous with kids now than they were 5/6 years ago. Mom and Dad are more likely to buy their kid an iPhone or give them last years model than shell out $300ish on a new toy.

 

[KAL2006]

You know what other numbers can't be ignored? Nintendo's handheld sales. It doesn't matter that Nintendo says the Switch is primarily a home console. What matters is what the public perceives it to be. So if people see this as a replacement to the 3DS as well as the Wii U, buyers who traditionally only buy Nintendo's handhelds may also buy the Switch.

Handheld sales have declined due to smartphones. Also Switch doesn't look to be priced as a handheld which is $200 and lower.

 

[blu]

I find explaining machine level operations to the layman difficult regardless of the architecture. But here goes:

Never underestimate laymen these days - those buggers can be quite smart (teaching some kids myself in my spare time).

Arm64 instruction set is very close to microcode (cpu operation).

On the average, yes. The ratio uops/ops is normally in the 1.x range, for general code. For amd64 it's closer to 2 due to the common use of one memory operand.

amd64 translates a complex instruction into a 3 or 4 operations or lines of microcode. For example an amd64 CPU can be told to add together two numbers stored in main memory.

A pdp11 can be told to add two numbers from memory. An amd64 arithmetic op cannot have two memory operands.

To do this, the CPU needs to fetch the number from memory address1 (op 1), fetch the number from memory address2 (op 2), add the two numbers (op 3). arm64 instructions are less complex so each instruction is closer to its microcode. For arm64, you use the instruction set to put memory address into registers then execute the addition instruction. AMD -> one line to add two numbers together ARM -> 3 or 4 lines to add two numbers together.

Normally it's 2 ops for amd64 and 3 (with PC-relative addressing) for arm64 for adding two numbers. But close enough. You explanation does cover the load/store memory access paradigm of all RISC designs. Of course on modern CISCs you just get an extra uop for a memory operand.

Now let me tell you what you missed (no details, just differences at a glance):

• two- vs three-argument ops
• 16- vs 32-strong GPR files
• SIB vs auto-incremental addressing modes
• non-orthogonal vs orthogonal simd ISA extensions

My last experience with embedded C was programming MH6211 to work with fire sensors for alarms. It was a learning experience and I'm happy to not do that work anymore

See, that's the point I was trying to make (you asked me but I did not answer right away what I was getting at): if you had written any portable C/C++ code for amd64 *and* arm64 you'd have known that the long CISC vs RISC exposition you provided is mostly irrelevant in this day and age - the compilers deal with nuisances like literal pools, PC-relative addressing, and long branches. One cannot even use the strong data alignment as an argument, as it's not valid for arm64. So from the POV of a C/C++ developer, the differences are effectively nil, niente, nada, unless they have to use SIMD intrinsics, where going arm64->amd64 is the much bigger problem than amd64->arm64.

 

[Lord Azrael]

You're absolutely right on that. The argument against your case would be mobile phones now. They are more ubiquitous with kids now than they were 5/6 years ago. Mom and Dad are more likely to buy their kid an iPhone or give them last years model than shell out $300ish on a new toy.

Handheld sales have declined due to smartphones. Also Switch doesn't look to be priced as a handheld which is $200 and lower.

I agree that mobile will negatively impact the Switch. I'm not sure that the Switch will sell as well as the 3DS. But I think it's still a strong reason to believe it will outsell the Wii U.

 

[theangrypirate]

Now let me tell you what you missed (no details, just differences at a glance):

• two- vs three-argument ops
• 16- vs 32-strong GPR files
• SIB vs auto-incremental addressing modes
• non-orthogonal vs orthogonal simd ISA extensions

I doubt the layman would have any idea what we are talking about here... I appreciate your expertise. It's been a while since I've revisited this knowledge.

See, that's the point I was trying to make (you asked me but I did not answer right away what I was getting at): if you had written any portable C/C++ code for amd64 *and* arm64 you'd have known that the long CISC vs RISC exposition you provided is mostly irrelevant in this day and age - the compilers deal with nuisances like literal pools, PC-relative addressing, and long branches. One cannot even use the strong data alignment as an argument, as it's not valid for arm64. So from the POV of a C/C++ developer, the differences are effectively nil, niente, nada, unless they have to use SIMD intrinsics, where going arm64->amd64 is the much bigger problem than amd64->arm64.

Fair point. But there will most certainly be optimization work. My point was Its harder and more expensive to develop for multiple architectures than one. It's not going to be free to port your PS4 game to Switch. Developers will have to determine if the cost is worth the potential sales. If the sales of the Switch are weak, it will make it less attractive to make ports for.

 

[guek]

Kinda crazy we're 2 months and change away from the presentation. It's coming up fast.

 

[blu]

I doubt the layman would have any idea what we are talking about here...

'We'? Unfortunately your explanation of the differences between amd64 and arm64 was yet more of RISC vs CISC fundamentals. See my last paragraph.

 

[sinxtanx]

.

 

[Dark Cloud]

.

This topic is about the January 12th presentation and its turned into this...

 

[The Last One]

I doubt the layman would have any idea what we are talking about here... I appreciate your expertise. It's been a while since I've revisited this knowledge.



Fair point. But there will most certainly be optimization work. My point was Its harder and more expensive to develop for multiple architectures than one. It's not going to be free to port your PS4 game to Switch. Developers will have to determine if the cost is worth the potential sales. If the sales of the Switch are weak, it will make it less attractive to make ports for.

So, you did all that useless retorical gymnastic to state the freaking obvious...

C'mon man... Don't push your bias that much

 

[theangrypirate]

So, you did all that useless retorical gymnastic to state the freaking obvious...

C'mon man... Don't push you bias that much

huh?

 

[theangrypirate]

'We'? Unfortunately your explanation of the differences between amd64 and arm64 was yet more of RISC vs CISC fundamentals. See my last paragraph.

Sorry, I forgot this turned into a dick measuring contest for you.

 

[blu]

Sorry, I forgot this turned into a dick measuring contest for you.

I've just tried showing you how irrelevant the RISC vs CISC argument is to this thread. Sorry if it's been dick measuring for you. My sincere apologies.

 

[theangrypirate]

Im just ready to finally get a proper reveal for this thing. Specuation has these threads going in circles.

How realistic is it to expect a 3D Mario at launch?

 

[Akhe]

I'm retarded trying to follow and understand the conversation between blu and theangrypirate.

Silly me..

 

[theangrypirate]

I feel like a retarded trying to follow the conversation between blue and theangrypirate.

Silly me..

I feel like a retarded getting sucked into that conversation :/

 

[TLZ]

Slow down there turbo, nobody said a trend was a law. Looking only at the chart anybody with half a brain can see the sales for a nintendo console over the entire history of their business are worse than the previous generation. The lone exception is the Wii. In statistics/econ you can determine the average slope over all the data points in this chart to predict where the next data point will be. Its called science and if you don't trust thats fine there are plenty of people who don't believe in it. Statistics are math and they have to to with probabilities. This data shows that Nintendo's next console "probably" wont do well. There is still a chance it will be wildly successful like the Wii. I wouldn't bet on it and the people who make money off of these things aren't betting on it either. Thats why Nintendo stocks fell after the announcement.

Someone already gave you the reasons behind the low sales and you still ignored it. Your analysis will only work if Nintendo continue their mistakes trend. Your analysis assumes they will and that is not science. You should show the history, you show the trend, you give reasons for the slump, then reasons for the wii, and conclude with 2 forecasts; the way of the mistakes trend or the way of the wii.

 

[ggx2ac]

I doubt the layman would have any idea what we are talking about here... I appreciate your expertise. It's been a while since I've revisited this knowledge.

A layman can understand anything as long as you can explain to them things in a way that you would explain it to a 5 year old.

Which in turn usually results in that layman behaving like a 5 year old.

"Why?"

"Why?"

"Why?"

 

[theangrypirate]

Someone already gave you the reasons behind the low sales and you still ignored it. Your analysis will only work if Nintendo continue their mistakes trend. Your analysis assumes they will and that is not science. You should show the history, you show the trend, you give reasons for the slump, then reasons for the wii, and conclude with 2 forecasts; the way of the mistakes trend or the way of the wii.

Its not my analysis, its just looking at what Nintendo has done for their entire console history and applying that trend to their next console. Thats it. I hope they correct the ship and don't continue making mistakes, but history shows us, at least with their consoles, they keep releasing systems that sell worse and worse. The fact that this is hybrid will hopefully break the trend.

 

[Sadist]

Im just ready to finally get a proper reveal for this thing. Specuation has these threads going in circles.

How realistic is it to expect a 3D Mario at launch?

Not very. There was an interview with Miyamoto a few months ago and he talked about "showing something from the new 3D Mario at E3 2017". Mario will most likely launch in November 2017.

 

[theangrypirate]

A layman can understand anything as long as you can explain to them things in a way that you would explain it to a 5 year old.

Which in turn usually results in that layman behaving like a 5 year old.

"Why?"

"Why?"

"Why?"

I was referring to the list of things blu told me I missed in my explanation. It wasn't until I was told my explanation was unsatisfactory did i realized i was being suckered into an argument I had no chance of winning instead of a constructive conversation. I feel bad for derailing. Id love to get this thread back on track somehow...

 

[AuthenticM]

Rockstar could decide to include an extra mission where Mario is fucking Princess Peach and Luigi comes in and beats Mario over the head with a baseball bat in a cutscene. Nintendo still wouldn't say no. Kid friendly image be damned. It's GTA.

Rockstar are the obstacle here. Not Nintendo.

lol no dude

 

[theangrypirate]

Not very. There was an interview with Miyamoto a few months ago and he talked about "showing something from the new 3D Mario at E3 2017". Mario will most likely launch in November 2017.

Damnit. Thats what I'm most excited for. Even though the footage shown in the reveal was "not representative of actual gameplay" It still got me hype. Like, levels of hype I haven't haven't had since N64 was released.

 

[Godzilla's Left Toe]

Not very. There was an interview with Miyamoto a few months ago and he talked about "showing something from the new 3D Mario at E3 2017". Mario will most likely launch in November 2017.

He never gave specifics on what Mario he was talking about in that interview.

 

[Neff]

A trend is not a law. Not only does it completely and conveniently dismisses the Wii as an anomaly, but it also implies that there is an inherent cause in the market and/or within Nintendo that prevents them from breaking it.

Nintendo install base charts are clever optical illusions, only certain people can see the Wii.

 

[BuggyMike]

I think a lot of people would look at an ARM based console as a huge misstep when the entire gaming industry (besides mobile) is x86. Also the fact that Nintendo is doing exactly the same thing as the Wii U as far as being behind current gen consoles with respect to power as another bad sign. See below chart:

Who are these people that feel this way? Again we've only heard good things about the architecture from the people in the "know". ARM doesn't seem to be an issue at all. I think it's the exact opposite of a misstep to go from an outdated IBM CPU architecture, to one of the most widely used CPUs today, especially considering the form factor of the device. You wanted them to pack x86 in a system that doubles as a portable? This seems to be the perfect solution. And once again, they've gotten Nvidia largely involved with the architecture, and on top of that, Nvidia designed the API, which should make for a huge improvement over Nintendo's own in-house tools.

This is a 30 year trend that is impossible to dismiss.

I'm not dismissing it, I literally just explained to you why it happened. And it kept happening because Nintendo wouldn't change their ways. There wasn't going to be an improvement unless Nintendo either caught lightning in a bottle again, or actually adapted to the way things currently are. Consumers don't just run for the hills when they see the name Nintendo, in fact, this past year has shown us just how much the general public love Nintendo, even while they are currently on their worse selling home console to date. The reactions to Pokemon Go and Zelda BoTW are great examples. Nintendo have great potential, they literally just have to get out of their own way, and (so far at least) they seem to be knocking it out of the park. Go and compare the reveal trailers of the Wii U and the Switch. With the reaction that Nintendo has gotten from gamers of all kinds, I don't see how you can claim that this system will sell like the Wii U, this early on.

Still with me?

Well, I read your explanation as to why you think ARM will lead to games being harder to port to. I then read your mini debate with Blu. It seems you're talking RISC vs CISC, but that doesn't seem to be an issue these days concerning ARM and x86? Again I'm not too good with tech but that's what I kinda gathered.

I mean, I'm sorry to be a bit of a jerk, but just understand that it's annoying when people come into these threads and make these bold claims like "the Switch will sell like the Wii U" and "this system will be hard to develop for" when they aren't even fully informed on the things that they are claiming will happen, and their only real argument as to why is "it's Nintendo, that's what's happened in the past, so it'll happen again". It's a weak argument, and as someone put it earlier, it's lazy.

 

[JPickford]

Games are mostly written in C++. Converting from one processor to another is a matter of changing the compiler. There's zero work involved for most coders. Unless you're working at a very low level you don't even need to know what the CPU is.