Rpi 2.0 in the works?

[scooterjes]

Hopefully this isn't the wrong place to post this question but is there a 2.0 version of the Rpi in the works? I would really like to be able to do a little I do understand it will cost more(better usually does) I don't want to make this a "what would my dream Rpi be?" thread I am sure there is at least 1 or 1000 already. I am considering getting a slightly faster computer than the Rpi maybe one of those micro ATX Atom based boards or the like.

If this is the wrong place move it to the right place.

[Wendo]

I would _very_ much doubt there is a 2.0 in the works, simply because the Model A hasn't even released yet, and the education version hasn't released yet.

Since the stated goal of the foundation is to get school kids into programming, one has to imagine they won't be working on a new version until they have at least begun to sell version 1 to schools

[jamesh]

There certainly isn't a faster Raspi in the works - sorry.

[W. H. Heydt]

Depends on what you mean by a "Pi 2.0". If you accept have 0 ohm "resistors" in place of polyfuses F1 and F2, then the "Pi 2.0" is already shipping. Otherwise...no.

[Jim JKla]

You could argue that the current model is the Mk2 and it is the Mk1 the model "A" that we are waiting for there are some discussions on what people would like to see in a Model "C" but as it stands this is nothing more than a wish list.

By default a "B" with a Gertboard is probably the nearest to a "C" as you will get for the time being.

[jamesh]

I look on the minor changes to the board currently out there as just a revision change (i.e. not noticeable to the majority) - the basic functionality/capability of the board hasn't changed.
Even some upcoming stuff is just a revision change to my eyes. Just like the BBC micro, which looked the same on the outside but had different board revision inside.

A different mk number, I think, would involve major changes - SoC etc.

[hojnikb]

Depends on what you mean by a "Pi 2.0". If you accept have 0 ohm "resistors" in place of polyfuses F1 and F2, then the "Pi 2.0" is already shipping. Otherwise...no.

Can you explain more what this means ?

[RaTTuS]

Depends on what you mean by a "Pi 2.0". If you accept have 0 ohm "resistors" in place of polyfuses F1 and F2, then the "Pi 2.0" is already shipping. Otherwise...no.

Can you explain more what this means ?

latest versions are shipped with 0ohm resistors instead of polyfuses helping some power issues

[W. H. Heydt]

Depends on what you mean by a "Pi 2.0". If you accept have 0 ohm "resistors" in place of polyfuses F1 and F2, then the "Pi 2.0" is already shipping. Otherwise...no.

Can you explain more what this means ?

Rattus gave the data. But to spring off what Jamesh posted as well...

The change from the polyfuses that limit the current draw on the USB ports, but have the side effect of dropping the voltage to the USB ports significantly as well to "0 ohm" resistors...basically a package replacement to get rid of the polyfuses...has some knock on effects.

Besides effectively raising the voltage on the USB ports and making more (but still very limited) current available has mitigated a number of problems people have had with keyboards, mice, and (especially) wireless network dongles.

As soon as I get one of the new boards, I plan to test it with a USB-to-PS/2 adapter--another device that has problems with lowered voltage and limited current from the polyfuses--that has previously failed to work. This is me liking the change for selfish reasons. IF the USB-to-PS/2 adapter works with the board revision, then I can hook up a Pi completely through my KVM switch.

And, no, I don't consider the change to be a "Ver. 2.0". It is--at most--"Ver. 1.0.1", or possibly even "Ver. 0.8.1".

[liz]

Definitive reply to OP: No, there isn't! We've got a whole platform that's currently under development (the current Raspberry Pi) to finish and polish before we even start to consider future versions.

[gritz]

And, no, I don't consider the change to be a "Ver. 2.0". It is--at most--"Ver. 1.0.1", or possibly even "Ver. 0.8.1".

This prompted a proper "laugh out loud" moment because of a current circumstance here...

Six years ago I bought an early bird licence on a piece of beta audio / MIDI software. I think it was at about V0.92.

It's still in beta and the 0.9726 update was released just yesterday! I'm worried we're going to hit the resolution limit of 32bit float and the developers will run out of decimal places before the full release happens. That said it's plenty stable enough for live use and bugs get squished promptly. I've even had a couple of feature requests implemented. TBH I couldn't live without it.

Sorry. Offtopic again...

[Jim Manley]

I'm worried we're going to hit the resolution limit of 32bit float and the developers will run out of decimal places before the full release happens.

Could I interest you in the wonderful new concept called "double"? You get 64 bits of floating-point goodness and it doesn't even require a hardware or Raspbian upgrade! If that's still not good enough for you, then let me call your attention to the dessert portion of the menu, where we are offering the sumptuous quadruple, a whole 128 bits of floating point Nirvana!

As for a Model C or Model B 2.0, as Tony Soprano would say, "Fuhgeddaboudid!" We haven't even scratched the surface of the VideoCore IV GPU, which is the real meat of the Pi, not the essentially traffic cop ARM CPU that should mostly be relied upon to get data between the GPU and the Ethernet and USB ports. We are in a similar situation that developers for the Playstation 3 were in when it first shipped - an incredible graphics powerhouse (the multi-core, multi-GHz Cell processor in the case of the PS3) that requires a significant amount of work to get useful output. When people see what 24 gigaFLOPS and 40 million shaded polygons per second look like, then they will realize what amazing capabilities the GPU in the Pi really has. The Quake 3 demo is just a taste of great things to come on the Pi in the way of blockbuster 3-D graphics which a number of us are working very hard on right now.

[JoeDaStudd]

Personally I think the RasPi should stay the way it is specifically with the same SoC.
It takes time to optimize software and make the most of hardware.
Look at the PS2, if you compare the original games (eg GTA 3) to the later games (eg GTA SA), you can see the massive improvement which is gained as the developers made better use of the system.
Raspberry Pi is doing a similar think only much quicker, look at the original (OS) images and compare them to the latest images. The same tasks are faster, less buggy and more fluid, that's with only a small number of RasPi's (actually shipped and in peoples hands) and in about 3 months.
Just imagine what we could have by the end of the year when far more people have a RasPi and another 4 months of development time.

[Jim JKla]

No not many RPi's out there only going to be about a half million some time next month.

last unoffical count was 470,000.

[W. H. Heydt]

Personally I think the RasPi should stay the way it is specifically with the same SoC.
It takes time to optimize software and make the most of hardware.
Look at the PS2, if you compare the original games (eg GTA 3) to the later games (eg GTA SA), you can see the massive improvement which is gained as the developers made better use of the system.
Raspberry Pi is doing a similar think only much quicker, look at the original (OS) images and compare them to the latest images. The same tasks are faster, less buggy and more fluid, that's with only a small number of RasPi's (actually shipped and in peoples hands) and in about 3 months.
Just imagine what we could have by the end of the year when far more people have a RasPi and another 4 months of development time.

That's the sort of debate I've been reading about for more than a decade. On the one hand, you have static hardware (e.g. game consoles) that developers try to wring every last bit of performance out of. The reason they do it that way is because the hardware isn't changed...until a new model is brought out and the cycle starts all over again. On the other hand, you have PCs, which are constantly and incrementally improving, there game companies have decide whether to write their games for the spread of hardware is currently in use, or what is currently available, or what they expect to be available when the game ships.

Both approaches have their virtues and faults, neither is perfect nor disasterous.

The Pi is probably going to resemble the console model more than the PC model. The hardware will be (relatively) static, though open (which consoles aren't) for longish periods of time--say on the order 3 years, at a guess. Given how new the Pi is, there is a lot of performance to be wrung out of the currently available hardware. We've already seen some of that in the released work using hardware floating point and USB improvements. I think we'll see a lot more when a hardware GPU version of X is ready for general consumption.

At the same time, given the way hardware prices change, both over time and with respect to ordering/production volumes (one will get a lot more attention from a supplier by ordering a million units than one will by ordering ten thousand....), it is reasonable to expect that there will be changes to the Pi hardware in directions to increase capability. (It *might* go in the direction of price reductions, but the problem there is that there usually a "floor" below which it is simply not economic to make something.) This is why I think that, in 2 to 3 years, we will see the Pi migrate to 512MB memory and, depending on Broadcom's plans, the possibility of a default clock speed increase to--potentially--as much as 1 GHz in the same time frame.

In a few years. probably no more than five or six, the Model A & B Pis will probably be sen as being rather "long in the tooth" and replacement with a more powerful system is likely to be in the cards. At THAT point, I would expect to see a dual- or quad-core design with (probably) 1GB memory module...and all at the same price point. THAT will be the "true" Rpi 2.0.

[gritz]

^^^ What Jim Manley said. ^^^

The GPU (which is the workhorse in the Pi) is very under-utilised at the moment. far better IMO to get all the software working optimally with the existing hardware than to do the x86 / x64 thing of hiding software kludges by tossing hardware resources at 'em, because that's just plain lazy!

Let's get this one "finished" first.