techpaul
Posts: 1512
Joined: Sat Jul 14, 2012 6:40 pm
Location: Reading, UK
Contact: Website

Re: Minor improvements for a model C

Mon Aug 13, 2012 9:37 pm

Lob0426 wrote: Some Data that should be closely monitored, is the number of B models versus A models that educators buy. Me and Vindicator, and others, believe that the B model will be the model of choice for most schools.
Well I am being 'nagged' to do the building bits and software introductions for GPIO only for Model B from a teacher in ICT and computing.

Model B is only practical as it can be run headless which means less cabling and avoids the VGA problems as schools have VGA only monitors (and will have for up to 5 years on replacement cycle). This also avoids USB keyboard and mouse issues.

More improtantly it is quicker to setup a headless unit than a fully wired up system, which is very important as smaller prebuilt assemblies can be used, quicker to setup and breakdown means less bits go missing.

The time factor is important as in a 45 minute to 60 minute, you don't want 10-15 minutes at both ends of the lesson lost in setup and breakdown.
Just another techie on the net - For GPIO boards see http:///www.facebook.com/pcservicesreading
or http://www.pcserviceselectronics.co.uk/pi/

User avatar
jackokring
Posts: 816
Joined: Tue Jul 31, 2012 8:27 am
Location: London, UK
Contact: ICQ

Re: Minor improvements for a model C

Tue Aug 14, 2012 3:11 am

Eventual additions to the memory chip on top of the broadcomm chip.
I've been thinking that any technology which improves memory chips should be added as and when the wafer design is improved. There therefore has to be a way arranged now to negotiate use of memory bound extensions to the memory chip. I propose that as any memory block engaged in extra functionality can be paged out using virtual memory, and that any kernel update has to work on old memory designs without problems. Then the extra functionality has to be paged in to start with, and a detect process change in memory has to be run by the kernel at boot. If the extras present change the memory contents and so are detected, they are paged out using one control memory page, and will await activation by a utility library, which may be emulated on pre-wafer boost technology. The other option is to hard boot them paged out, and switch them on somehow later, such that old kernels do not complain about changes in memory. Can a magic number be located in the kernel so the memory looks for the go?

If all people go on about is I wait to by model C, then model B becomes un-profited, and has a high wastage. Where as getting more chip function using less power, and maybe having an on for some date after the production date of wafer 2. The interface needs to be made so as chip 2 can be matured too.

Cheers Jacko
Pi[NFA]=B256R0USB CL4SD8GB Raspbian Stock.
Pi[Work]=A+256 CL4SD8GB Raspbian Stock.
My favourite constant 1.65056745028

User avatar
Vindicator
Posts: 314
Joined: Sat Sep 17, 2011 11:10 pm
Location: Susanville Ca USA
Contact: Website

Re: Minor improvements for a model C

Tue Aug 14, 2012 4:36 am

I agree @Jamesh there are far to many people trying to make specific quotes out of general statements (if that is at all what your suggesting) and this does not really have anything to do with thread other than the mention of the power problems that have arisen from the choice of power source for the system.

Here's my point, the device (Raspberrypi) serves the purpose and intent of the foundation and revising it for the whims of the general public or hobbyist out there may cause more harm than it solves, what do you really want from a $35 device.

For my projects the A model would do nicely and be far less work than removing unnecessary components.

I would vote for a completely separate release (if this was at all plausible in the first place) only aimed at the hobbyist/general public and they could just pay the additional prices as a profitable venture, rather than asking a non profit venture to create a new revision of the device for their needs.(this is not the path that was chosen by the foundation so please do not misrepresent this statement)

Most of the features being requested in numerous threads besides this one, are available in other much more expensive devices, but they are available(except maybe usb 3.0)

The continuous unnecessary quoting is just spoiling any fun that could have been generated by this thread and any other thread you may be posting in and quoting in is really just chasing your opponents from the thread rather than producing a constructive discussion.
If you are more worried about ,spelling, punctuation or grammar you have probably already missed the point so please just move on.

User avatar
jackokring
Posts: 816
Joined: Tue Jul 31, 2012 8:27 am
Location: London, UK
Contact: ICQ

Re: Minor improvements for a model C

Tue Aug 14, 2012 5:00 pm

jackokring wrote:Eventual additions to the memory chip on top of the broadcomm chip.
The interface needs to be made so as chip 2 can be matured too.
A few rough paper calculations, imply that with one 32 bit float multiplier per 4KB of memory, would achieve about 40+ Gflops. If it's a smaller 16 bit multiplier with an assembly pipe to make 32 bit results, then 10+ Gflops of extra compute could be provided instead of doubling the memory.

Another reason for the memory chip as a reason for the location is the high volume of memory chip production, and design regularity.
Pi[NFA]=B256R0USB CL4SD8GB Raspbian Stock.
Pi[Work]=A+256 CL4SD8GB Raspbian Stock.
My favourite constant 1.65056745028

W. H. Heydt
Posts: 9996
Joined: Fri Mar 09, 2012 7:36 pm
Location: Vallejo, CA (US)

Re: Minor improvements for a model C

Wed Aug 15, 2012 12:10 am

jackokring wrote:
jackokring wrote:Eventual additions to the memory chip on top of the broadcomm chip.
The interface needs to be made so as chip 2 can be matured too.
A few rough paper calculations, imply that with one 32 bit float multiplier per 4KB of memory, would achieve about 40+ Gflops. If it's a smaller 16 bit multiplier with an assembly pipe to make 32 bit results, then 10+ Gflops of extra compute could be provided instead of doubling the memory.

Another reason for the memory chip as a reason for the location is the high volume of memory chip production, and design regularity.
I'm missing something here.... The registers are in the CPU or GPU. The memory is allocated to either the CPU or the GPU.

The PCB design, functionally, dictates that any processor chip (GPU+CPU) must have the same pinouts and very, very similar voltage requirement as the BCM2835, and not too much difference in current requirements.

The memory POP module has to have pinouts compatible with the processor package.

From what has been stated in various threads, this means that, at present, we are limited to the BCM2835. Whether or not there are higher clocked versions hasn't been determined (I can't tell from Broadcom's website...I simply haven't found enough specs to tell). It has been stated that a 512MB memory package exists--but it's too expensive to use and keep the Pi's price point. That is something that has the potential to change...the module price may come down, the cost margins may improve (due to the volume of demand), or--least likely of all--a higher priced Model "C" might be introduced.

Since this thread is concerned with MINOR improvements to make a "Model C", I could see three, or possibly four changes that could be realistically considered (in no particular order):

1. Faster model of BCM2835 (if it exists).
2. The 512MB memory module
3. A better SD card connector.
4. Some sort of upgrade to the polyfuses to handle MODEST additional current to both the system generally and the USB ports in particular (for example, raising the allowed current to each USB port to 200mA from the present 140mA, and total system current to a max of 1A).

Bear in mind that there are--apparently--issues with the polyfuses at present, and those issues are being looked at.

I believe that all four of those upgrades is consistent with the goals for the Pi, in that they would make the Pi more robust (the SD connector), easier to program (more memory and faster processor), or less finicky to work with (the SD connector and the polyfuse/power revision).

The only reason I wouldn't just suggest those items for a revised Model "B" is that they would drive the cost too high to hold the current price point, since the price point is a deliberate choice on the part of the Foundation.

On the other hand, if one were given at least a semi-free hand, after going for the above changes, I would start with replacing the voltage regulators with switching regulators, drop the composite video out connector in favor of an audio in connector, and add some form of system bus connector that would provide the ability to attach an expansion card that could carry all the other connectors people keep asking for (say, more USB ports, VGA, SATA, and so on).

User avatar
Lob0426
Posts: 2198
Joined: Fri Aug 05, 2011 4:30 pm
Location: Susanville CA.
Contact: Website

Re: Minor improvements for a model C

Wed Aug 15, 2012 1:54 am

There is no "pin" (BGA) compatible processor to the BCM2835. The closest thing to it is the BCM11311 which is not compatible. It does have the VideoCoreIV GPU but it is dual core and v7. So even the software would not be compatible. Of course the Os's could be recompiled with the newer instruction set. It has several other features, one such is a dedicated Ethernet port. Also is able to except the larger (size) Memory modules.
There are 512MB memory modules available for the BCM2835, but they are expensive, so until they drop in price you will have to wait. The A Model was going to have a 128MB module. It was found that it was better to just source the 256MB module for both rather than by two different modules. So we come out good here.

Not much hope of a better SD card connector coming along unless the ones that have been sourced are no longer available. Then new ones would have to be sourced at a similar price.

The poly fuses are part of the parts sourcing also. So there is not much chance of a change there either. The current F3 is only 750ma. The USB (F1/F2) are 140ma (these are the rated trip current) So 280ma can be used by the USB alone leaving 470ma for the rest of the board. The original power schematics called for a 1.1A (1100ma) F3, even at that you would be limited to about 300ma per USB fuse. But now you have to replace F1, F2 and F3. The price difference is probably only cents apiece, but the $35 (or $25) is contractual. Besides these fuses are easy to get and to replace yourself. It will cost you about a $1.25 and ten minutes to do it. Most of the time is waiting for the iron to heat up.

Farnell and RS are not going to take even cents out of their pockets to build a better A or B model. This would require a new contract with the Foundation. So the price would have to go up. I have to say neither would most of us, after taking the time to research the necessary changes. The Foundation is not going to agree to that. I do not think the Foundation is going to agree to a larger Memory module unless it is in the same price range as the 256MB module. Again it is a contracted deal. The Foundation would have to negotiate a seperate deal for a "slightly upgraded" device. Why waste the (negotiation) time on a slightly upgraded device. They may as well wait and negotiate for a "C Model". The only other avenue would be if Farnell or RS approached the Foundation and asked to sell an upgraded unit above the contracted price per unit. At this point a deal may be made for an upgraded device. Most likely they would bring the fuses up and install a 512MB memory module. The price would probably be in the area of $50. Theyt are going to want to get more profit margin into an upgraded version than they are getting for the "standard" models. The Foundation is only going to do this after the current, huge backlog of the B model is cleared up by both of the retailers. At the current rate I do not expect this to happen until well after Christmas time. The other possibility is a board revision to cure the 1v8 rail error, so far that is a dead end. The board works, with the error, just fine for the majority of people.

I am no trying to dampen all of OUR hopes! I just needed to put a realistic light onto the discussion of a slightly upgraded board. We really have a better chance of a C Model in the future than an upgraded model anytime soon.

We can upgrade several things on the board ourselves. Many of us are already "modding" (upgrading) features that we are not happy with on the Raspberry Pi B model. Unfortunately The Memory Module is not one of them!

So what are the realistic hopes for a C Model?
Upgraded processor; not likely for at least several years. After that likely.
Upgraded Memory; Likely future addition.
Power connector change; very unlikely. Working fine, maybe not the power supplies connected to it.
Upgraded USB Polyfuses; Maybe. Most likely higher trips for a wider range of devices.
USB 3.0; not even on the radar for an under $50 device.
Switching regulators; Very unlikely.
SATA, PATA or other mass storage connector; Not even on the radar for an under $50 device.
VGA connector; NOPE.
Upgraded analog or digital sound; likely.
Audio in; likely.

The caveat is that most of this is only "likely" in a couple of years. Remember it took six years for the RasPi to materialise. A C model will be faster due to a shorter learning curve overall.
The number of ARM devices that are in design right now, due in part to the Raspberry Pi, may make a C Model completely unneccesary for the Foundation!
512MB version 2.0 as WordPress Server
Motorola Lapdock with Pi2B
Modded Rev 1.0 with pin headers at USB

http://rich1.dyndns.tv/
(RS)Allied ships old stock to reward its Customers for long wait!

User avatar
reiuyi
Posts: 165
Joined: Sun Oct 09, 2011 4:59 pm
Contact: Website

Re: Minor improvements for a model C

Wed Aug 15, 2012 8:44 am

@Lob0426
Finally there is a person which is realistic about possible changes and the lack of motivation to do so.

To anyone in this thread:
Why do you seek a "better/faster/bigger" raspberry pi? There are so many other demo boards out there that meet your exact specifications within the 50-150$ range. I do not understand at all why you wish to fragment the raspberry platform that can be beautiful if tech users weren't so addicted to change.

To me personally it still sometimes sounds like these kinds of threads are only made so that an ideal HTPC is put together. If you want a htpc, just go out and buy an Apple TV, Roku, WDTV live, Boxee or whatever else is on the market these days.

Have fun with your raspberry pi, that is all :D

itimpi
Posts: 1083
Joined: Sun Sep 25, 2011 11:44 am
Location: Potters Bar, United Kingdom
Contact: Website

Re: Minor improvements for a model C

Wed Aug 15, 2012 9:04 am

I have heard Eben describe the Pi as a breakout board for the bcm2835 soc that is used as the most important chip on the Pi. Getting lots of capability into a single chip is what has kept the price down. As such any ideas for enhancement are constrained by the capabilities that the chip has built into it. Adding anything else is likely to significantly increase the price. This is shown as simply adding the ethernet chip to produce the Model B added $10 to the price compared to the Model A.

That is why there are things like the RCA video output - the chip has support built in. It is also why there is no concept of VGA output as the chip does not support it. Many other ideas for enhancement that get bought up fall into this category.

The other point that often seems to get forgotten is that the Foundation is a charity with specific aims. Making changes that do not support those aims is outside the Foundations remit despite how much many people using the Pi in other areas of use might like some capability to be added.

The most likely change I can see happening is simply increasing the PolyFuse ratings. This is not a significant change from a functional point of view and does not require any board change. It may well help with getting a wider variety of mice and keyboards (and WiFi dongles/USB sticks) working without needing a powered hub - all of which seems in line with the Foundations aims.

alexchamberlain
Posts: 121
Joined: Thu Jun 14, 2012 11:20 am
Location: Leamington Spa, UK
Contact: Website

Re: Minor improvements for a model C

Wed Aug 15, 2012 9:22 am

One solution to the USB hardware problems is for the community to develop a stackable USB hub. Such a hub could be externally powered, power the Raspberry Pi, and by a hub at the same time.

It could include an ethernet interface, but I would say no to this suggestion as it could be another expansion board.

What do people think?

The community need to take on some of these challenges themselves and develop the expansion boards. Leave the foundation to do what it set out to do. Teach kids to program by providing a simple and solid base.
Developer of piimg, a utility for working with RPi images.

User avatar
jackokring
Posts: 816
Joined: Tue Jul 31, 2012 8:27 am
Location: London, UK
Contact: ICQ

Re: Minor improvements for a model C

Wed Aug 15, 2012 12:41 pm

{quote]
Eventual additions to the memory chip on top of the broadcomm chip.
The interface needs to be made so as chip 2 can be matured too.

A few rough paper calculations, imply that with one 32 bit float multiplier per 4KB of memory, would achieve about 40+ Gflops. If it's a smaller 16 bit multiplier with an assembly pipe to make 32 bit results, then 10+ Gflops of extra compute could be provided instead of doubling the memory.

Another reason for the memory chip as a reason for the location is the high volume of memory chip production, and design regularity.
I'm missing something here.... The registers are in the CPU or GPU. The memory is allocated to either the CPU or the GPU. [/quote]

True at present yes. For the totally boring uses, I suppose you'd never even know people who make memory chips can improve designs and have to make new models to remain competitive.
The PCB design, functionally, dictates that any processor chip (GPU+CPU) must have the same pinouts and very, very similar voltage requirement as the BCM2835, and not too much difference in current requirements.

The memory POP module has to have pinouts compatible with the processor package.

True and possible.
Pi[NFA]=B256R0USB CL4SD8GB Raspbian Stock.
Pi[Work]=A+256 CL4SD8GB Raspbian Stock.
My favourite constant 1.65056745028

User avatar
RaTTuS
Posts: 10242
Joined: Tue Nov 29, 2011 11:12 am
Location: North West UK

Re: Minor improvements for a model C

Wed Aug 15, 2012 12:51 pm

I think it has been said that they can go up to 512Mb RAM with just a part upgrade - anything else needs a redesign
How To ask Questions :- http://www.catb.org/esr/faqs/smart-questions.html
WARNING - some parts of this post may be erroneous YMMV

1QC43qbL5FySu2Pi51vGqKqxy3UiJgukSX
Covfefe

User avatar
jackokring
Posts: 816
Joined: Tue Jul 31, 2012 8:27 am
Location: London, UK
Contact: ICQ

Re: Minor improvements for a model C

Wed Aug 15, 2012 12:56 pm

RaTTuS wrote:I think it has been said that they can go up to 512Mb RAM with just a part upgrade - anything else needs a redesign
Yep, I've only suggested that a memory module, could be replace as and when silicon gets made anyway, which required no Pi redisign, and would fit on the bcm just in exactly the same way. I suggested it needed a software interface for certain features to become available and so I proposed one. Maybe this is not model C specific, but educational non the less. So a likely possible path is

256MB->512MB->512MB/20Gf->512MB/40Gf(independent 512MB data server)...... hit silicon wall.
Pi[NFA]=B256R0USB CL4SD8GB Raspbian Stock.
Pi[Work]=A+256 CL4SD8GB Raspbian Stock.
My favourite constant 1.65056745028

User avatar
MadHorseman
Posts: 83
Joined: Wed Mar 21, 2012 9:48 am
Location: Almadena, W Algarve, Portugal
Contact: Website

Re: Minor improvements for a model C

Wed Aug 15, 2012 1:15 pm

I think it should look much more impressive, like this:-

http://jjbunn.wordpress.com/2008/09/03/ibm-3090/

plus the error messages have obviously come from the same source.

gritz
Posts: 449
Joined: Sat Jan 28, 2012 2:33 am

Re: Minor improvements for a model C

Wed Aug 15, 2012 3:01 pm

*Sigh*

Even a simple RAM increase would break backward compatibility - look at all the hassle Android is having with platform fragmentation. The Pi OS and driver package is barely out of the womb, there are pressing issues of basic functionality and all some people can do is cry for upgrades. Geeks, eh? :lol:

I vote for the Foundation just getting the thing working as it should. If the platform is to have any kind of longevity as anything other than a cheap board for schoolkids then it simply makes good business sense to do so - the wider community is powerless to properly address issues that are "under the hood", or hardware based. The initial 10k "development" run has been exceeded by way more than an order of magnitude, after all.

User avatar
Lob0426
Posts: 2198
Joined: Fri Aug 05, 2011 4:30 pm
Location: Susanville CA.
Contact: Website

Re: Minor improvements for a model C

Wed Aug 15, 2012 4:01 pm

The Memory Module would simply be sourcing the new modules for future production. It would not cause any backward compatibility issues at all. It would create more resources to use and abuse. Changing the SoC would create compatibility issues.
Broadcom could create a compatible upgrade to the SoC. But why would it bother. They already have an upgraded SoC that uses VideoCoreIV technology. V6 instructions are old. And the ARM9 is old hat also. The popularity of the Raspberry Pi might cause a new processor in the lineup but it would most likely not be "pin compatible". So a board redesign will be necessary. If one of us had about a half million dollars or so and could get an in with Broadcom we might be able to come up with an upgraded board. Or donate that half million to the Foundation to motivate a new board! I do believe they will eventually take up the task of a new board, but it is just too soon for it too happen yet. :(
512MB version 2.0 as WordPress Server
Motorola Lapdock with Pi2B
Modded Rev 1.0 with pin headers at USB

http://rich1.dyndns.tv/
(RS)Allied ships old stock to reward its Customers for long wait!

gritz
Posts: 449
Joined: Sat Jan 28, 2012 2:33 am

Re: Minor improvements for a model C

Wed Aug 15, 2012 4:25 pm

Lob0426 wrote:The Memory Module would simply be sourcing the new modules for future production. It would not cause any backward compatibility issues at all. It would create more resources to use and abuse. Changing the SoC would create compatibility issues.
Not so. Doubling the memory would doubtless cause a fork in software development, with the risk of the "have nots" being left behind. There seems no point in fragmenting the platform at such an early stage. Focus!

User avatar
Lob0426
Posts: 2198
Joined: Fri Aug 05, 2011 4:30 pm
Location: Susanville CA.
Contact: Website

Re: Minor improvements for a model C

Wed Aug 15, 2012 5:39 pm

gritz wrote:
Lob0426 wrote:The Memory Module would simply be sourcing the new modules for future production. It would not cause any backward compatibility issues at all. It would create more resources to use and abuse. Changing the SoC would create compatibility issues.
Not so. Doubling the memory would doubtless cause a fork in software development, with the risk of the "have nots" being left behind. There seems no point in fragmenting the platform at such an early stage. Focus!
There undoubtably would be those that would program for the larger memory package alone. But most would design to fit both platforms memory. This is an issue that could be handled by managing your swap settings properly. Anybody that programs for the educational release will optimize for the 256MB package. The ones that really want more memory will be the home theater and desktop replacement crowd. Their software would be optimized for a larger PoP if it were available. As to "fragmenting" the platform. It is already pretty fragmented, programming, HTPC, Desktop replacement, all around smart controller, router, automation and so on. There are quite a number of specialist groups here in the forums. So I have to disagree that a larger memory package would endanger the platform!
512MB version 2.0 as WordPress Server
Motorola Lapdock with Pi2B
Modded Rev 1.0 with pin headers at USB

http://rich1.dyndns.tv/
(RS)Allied ships old stock to reward its Customers for long wait!

User avatar
jackokring
Posts: 816
Joined: Tue Jul 31, 2012 8:27 am
Location: London, UK
Contact: ICQ

Re: Minor improvements for a model C

Wed Aug 15, 2012 5:44 pm

gritz wrote:
Lob0426 wrote:The Memory Module would simply be sourcing the new modules for future production. It would not cause any backward compatibility issues at all. It would create more resources to use and abuse. Changing the SoC would create compatibility issues.
Not so. Doubling the memory would doubtless cause a fork in software development, with the risk of the "have nots" being left behind. There seems no point in fragmenting the platform at such an early stage. Focus!
This fork is minor. It's a less complex fork than the 16k/48k split the ZX Spectrum had for example. This can be solved by non critical array decimation. 256MB is loads.

Cheers Jacko
Pi[NFA]=B256R0USB CL4SD8GB Raspbian Stock.
Pi[Work]=A+256 CL4SD8GB Raspbian Stock.
My favourite constant 1.65056745028

gritz
Posts: 449
Joined: Sat Jan 28, 2012 2:33 am

Re: Minor improvements for a model C

Wed Aug 15, 2012 5:51 pm

Mmmmm, instigating swap (and on an SD card too). Bigger! faster! Still in beta! Profligacy is a fine example to set the next generation of coders, I'm sure.

Imagine I've just bought a new car. When it turns up I find that it doesn't always go in the direction that the steering wheel is pointed. Also, the boot has a tendency to spring open. I've lost my shopping twice now. And my dog. Apparently Rover wasn't compatible. Surely a Jack Russell is compatible with everything? It were dead cheap though (the car, not the dog), so I can't complain. It's cost me a bit in diesel, mind - some just don't seem to work. You'd think it'd all be the same...

I'm hoping someone at the owner's club will fix the boot and steering for free (and in their spare time). Still, it's only details, innit? TBH I'd rather have a bigger engine and wider wheels. Vroom! Vroom!

W. H. Heydt
Posts: 9996
Joined: Fri Mar 09, 2012 7:36 pm
Location: Vallejo, CA (US)

Re: Minor improvements for a model C

Wed Aug 15, 2012 10:22 pm

Well...if people are going to get into software issues, it would probably be a good idea do do a "wait and see" to find out what happens when hardware accelerated X is achieved. That *might* even ease up some of the memory constraints (by offloading software routines in memory to GPU calls).

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 22063
Joined: Sat Jul 30, 2011 7:41 pm

Re: Minor improvements for a model C

Thu Aug 16, 2012 2:56 am

RAM costs drop all the time, and smaller sizes get deprecated - as I posted in anther thread, I can see a time where the 256MB package is more expensive than the 512MB package. The only sensible thing to do from a cost POV is then move to 512MB. A more interesting choice is if the packages are very similar in price - what would you do? Stick with 256 even though 512 is only $0.10 more? Remember the Raspi is all about pragmatism - it's fed most of the design choices.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed. Here's an example...
"My grief counseller just died, luckily, he was so good, I didn't care."

User avatar
jackokring
Posts: 816
Joined: Tue Jul 31, 2012 8:27 am
Location: London, UK
Contact: ICQ

Re: Minor improvements for a model C

Thu Aug 16, 2012 3:21 am

jamesh wrote:RAM costs drop all the time, and smaller sizes get deprecated - as I posted in anther thread, I can see a time where the 256MB package is more expensive than the 512MB package. The only sensible thing to do from a cost POV is then move to 512MB. A more interesting choice is if the packages are very similar in price - what would you do? Stick with 256 even though 512 is only $0.10 more? Remember the Raspi is all about pragmatism - it's fed most of the design choices.
I'd go with 256MB... and use the other silicon area as compute resources.... :D
Maybe the best solution for some is a way to get one Pi to work as a memory server for another. So the best thru put interface is the .... ? Maybe a distro just for Pi duo expansion services, and a tiny cat5 xover? Or just 2 commands "master" and "slave" within raspbian, a small cat5, and some large memory homework egg-samples

Cheers Jacko
Pi[NFA]=B256R0USB CL4SD8GB Raspbian Stock.
Pi[Work]=A+256 CL4SD8GB Raspbian Stock.
My favourite constant 1.65056745028

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 22063
Joined: Sat Jul 30, 2011 7:41 pm

Re: Minor improvements for a model C

Thu Aug 16, 2012 5:28 am

jackokring wrote:
jamesh wrote:RAM costs drop all the time, and smaller sizes get deprecated - as I posted in anther thread, I can see a time where the 256MB package is more expensive than the 512MB package. The only sensible thing to do from a cost POV is then move to 512MB. A more interesting choice is if the packages are very similar in price - what would you do? Stick with 256 even though 512 is only $0.10 more? Remember the Raspi is all about pragmatism - it's fed most of the design choices.
I'd go with 256MB... and use the other silicon area as compute resources.... :D
Maybe the best solution for some is a way to get one Pi to work as a memory server for another. So the best thru put interface is the .... ? Maybe a distro just for Pi duo expansion services, and a tiny cat5 xover? Or just 2 commands "master" and "slave" within raspbian, a small cat5, and some large memory homework egg-samples

Cheers Jacko
Doesn't work like that. The PoP Ram is just soldered on top of the SoC. So going with larger memory just means a higher capacity PoP package. Nothing to do with the CPU which remains the same BRCM2835
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed. Here's an example...
"My grief counseller just died, luckily, he was so good, I didn't care."

W. H. Heydt
Posts: 9996
Joined: Fri Mar 09, 2012 7:36 pm
Location: Vallejo, CA (US)

Re: Minor improvements for a model C

Thu Aug 16, 2012 6:07 am

jamesh wrote:RAM costs drop all the time, and smaller sizes get deprecated - as I posted in anther thread, I can see a time where the 256MB package is more expensive than the 512MB package. The only sensible thing to do from a cost POV is then move to 512MB. A more interesting choice is if the packages are very similar in price - what would you do? Stick with 256 even though 512 is only $0.10 more? Remember the Raspi is all about pragmatism - it's fed most of the design choices.
Agreed...and that cost trend is why I have been pointing out the future practicality of going to 512MB. I wouldn't be surprised to see the move in a couple of years (electronics changes move pretty fast).

If it were me, once the memory module price is functionally the same, I'd line up contracts for the larger module, but not announce the change until the boards with more memory were ready to go out the door in quantity. In the mean time, run the inventory levels as low as possible...because once a 512MB board is known to be in the works, a lot of people will wait for it, cutting off the market for the 256MB boards.

It's not like type of problem hasn't happened before...

grmbbsr
Posts: 1
Joined: Sat Aug 18, 2012 2:39 am

Re: Minor improvements for a model C

Sat Aug 18, 2012 3:27 am

I feel there is tremendous potential for improvement in future versions. It can have extra ports. Port for variable RAM. Port for Graphics card. Port for Sound card. Port for Hard Disk. Port for CD/DVD. SIM Card Slot. A/V input/output Slot. More USB port. VGA, COM, LPT ports. Processor Slot. Then only it would be complete package

Return to “General discussion”