eLua Development

Extending MCU memory with SPI?

Classic

List

Threaded

38 messages Options

BogdanM

Re: Extending MCU memory with SPI?

On Wed, Jun 6, 2012 at 1:21 PM, pito <[hidden email]> wrote:
>> It doesn't actually say that it executes from SPI
>> flash. It states it runs
>> a virtualized CPU with the emulator fetching
>> virtual instructions from SPI
>> flash.
> Paul, it is a common understanding here we do not plan or discuss to
> run a code directly out of SPI.

Why is that a common understanding? SPIFI allows you do to exactly
this, if I understood correctly.

Best,
Bogdan

> It is always a kind of
> emulation/virtualisation..
>
> --
> - - Reklama - - - - - - - - - - - - - -
> Maturity, přijímací zkoušky, státnice, diplomky... Vše o závěru
> studia na středních a vysokých školách čtěte na portálu VOLNÝ.cz na
> http://bit.ly/LfImCR
>
> _______________________________________________
> eLua-dev mailing list
> [hidden email]
> https://lists.berlios.de/mailman/listinfo/elua-dev

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Paul Curtis

Re: Extending MCU memory with SPI?

In reply to this post by Pito

Hi,

> -----Original Message-----
> From: pito [mailto:[hidden email]]
> Sent: 06 June 2012 11:22
> To: [hidden email]; [hidden email]
> Subject: Re: [eLua-dev] Extending MCU memory with SPI?
>
> > It doesn't actually say that it executes from SPI flash. It states it
> > runs a virtualized CPU with the emulator fetching virtual instructions
> > from SPI flash.
>
> Paul, it is a common understanding here we do not plan or discuss to run a
> code directly out of SPI. It is always a kind of
> emulation/virtualisation..

...but we were actually talking about running code out of SPI, that's
exactly what has been endowed on the LPC1800 microcontrollers.

--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
SolderCore running Defender... http://www.vimeo.com/25709426

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Pito

Re: Extending MCU memory with SPI?

In reply to this post by Paul Curtis

FYI - :)
http://dmitry.co/index.php?p=./04.Thoughts/08. uM0
http://dmitry.co/index.php?p=./04.Thoughts/07. Linux on 8bit

A friend of mine ported the above linux stuff to pic32 so it is much
faster. There is a LOT of people who do not chase the speed, but
deal more with such crazy challenges as emulation/virtualisation. It
is a pity the compiler developers do not support the 18037 standard
:) broadly..

----- PŮVODNÍ ZPRÁVA -----
Od: "Paul Curtis" <[hidden email]>
Komu: "'eLua Users and Development List (www.eluaproject.net)'"
<[hidden email]>
Předmět: Re: [eLua-dev] Extending MCU memory with SPI?
Datum: 6.6.2012 - 12:14:09

> > http://www.trioflex.ee/index.php/services/portfolio/6-comp2hdmi
> > >
> > on the bottom of that PCB there is CortexM0 that
> > is executing CortexM0
> > > code from SPI flash...
> >
> > updates to those images are done via web browser
> > access from WII console,
> > > point to update page, start flash movie.. the
> > sound is decoded in real
> > > time while the M0 is generating onscreen
> > graphics in background.. updates
> > > are written to the SPI flash, and then new code
> > is executing from spi
> > > flash
>
> It doesn't actually say that it executes from SPI
> flash. It states it runs
> a virtualized CPU with the emulator fetching
> virtual instructions from SPI
> flash. At no point is hard Cortex-M0 running its
> own instructions from SPI
> flash. Note that the blurb is actually quite
> loose because it says it
> fetches ARM instructions for a Thumb-2
> processor--in general, neither the
> ARM instruction set nor the "full" Thumb-2
> instruction set will run on a
> Cortex-M0 device because they are different
> architectures. (Ok, the Thumb
> subset will run on a Thumb-2-capable processor,
> please don't be pedantic).
>
> Well, that's what I read from the article.
>
> --
> Paul Curtis, Rowley Associates Ltd
> http://www.rowley.co.uk
> SolderCore running Defender...
> http://www.vimeo.com/25709426
>
> _______________________________________________
> eLua-dev mailing list
> [hidden email]
> https://lists.berlios.de/mailman/listinfo/elua-dev
>

--
- - Reklama - - - - - - - - - - - - - -
Maturity, přijímací zkoušky, státnice, diplomky... Vše o závěru
studia na středních a vysokých školách čtěte na portálu VOLNÝ.cz na
http://bit.ly/LfImCR

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Paul Curtis

Re: Extending MCU memory with SPI?

In reply to this post by BogdanM

> On Wed, Jun 6, 2012 at 1:21 PM, pito <[hidden email]> wrote:
> >> It doesn't actually say that it executes from SPI flash. It states
> >> it runs a virtualized CPU with the emulator fetching virtual
> >> instructions from SPI flash.
> > Paul, it is a common understanding here we do not plan or discuss to
> > run a code directly out of SPI.
>
> Why is that a common understanding? SPIFI allows you do to exactly this,
> if I understood correctly.

Indeed, that is exactly what SPIFI allows you to do.

So, let's benchmark an LPC1800 at 180 MHz... (CrossWorks+gcc4.6, Hitex
LPC1850 board)

From internal SRAM:

Microseconds for one run through Dhrystone: 3.5
Dhrystones per Second: 285714.3

From external NOR flash (16-bit interface; 14 wait states):

Microseconds for one run through Dhrystone: 71.0
Dhrystones per Second: 14084.5

From external SPIFI flash:

Microseconds for one run through Dhrystone: 101.0
Dhrystones per Second: 9901.0

We haven't configured to run a program from external RAM--I guess the wait
states would come down and performance would go up...

Regards,

--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
SolderCore running Defender... http://www.vimeo.com/25709426

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Pito

Re: Extending MCU memory with SPI?

In reply to this post by Pito

http://jaromir.xf.cz/old/arm_emu/arm_emu.html

----- PŮVODNÍ ZPRÁVA -----
Od: "pito" <[hidden email]>
Komu: [hidden email], [hidden email]
Předmět: Re: [eLua-dev] Extending MCU memory with SPI?
Datum: 6.6.2012 - 12:47:33

> FYI - :)
> http://dmitry.co/index.php?p=./04.Thoughts/08. uM0
> http://dmitry.co/index.php?p=./04.Thoughts/07.
> Linux on 8bit
>
> A friend of mine ported the above linux stuff to
> pic32 so it is much
> faster. There is a LOT of people who do not chase
> the speed, but
> deal more with such crazy challenges as
> emulation/virtualisation. It
> is a pity the compiler developers do not support
> the 18037 standard
> :) broadly..
>
> ----- PŮVODNÍ ZPRÁVA -----
> Od: "Paul Curtis" <[hidden email]>
> Komu: "'eLua Users and Development List
> (www.eluaproject.net)'"
> <[hidden email]>
> Předmět: Re: [eLua-dev] Extending MCU memory with
> SPI?
> Datum: 6.6.2012 - 12:14:09
>
> > > http://www.trioflex.ee/index.php/services/portfolio/6-comp2hdmi
> > > > > > > > >
> > > on the bottom of that PCB there is CortexM0
> > > that
> > > > > is executing CortexM0
> > > > code from SPI flash...
> > >
> > > updates to those images are done via web
> > > browser
> > > > > access from WII console,
> > > > point to update page, start flash movie..
> > > > the
> > > > > > sound is decoded in real
> > > > time while the M0 is generating onscreen
> > > graphics in background.. updates
> > > > are written to the SPI flash, and then new
> > > > code
> > > > > > is executing from spi
> > > > flash
> >
> > It doesn't actually say that it executes from
> > SPI
> > > flash. It states it runs
> > a virtualized CPU with the emulator fetching
> > virtual instructions from SPI
> > flash. At no point is hard Cortex-M0 running
> > its
> > > own instructions from SPI
> > flash. Note that the blurb is actually quite
> > loose because it says it
> > fetches ARM instructions for a Thumb-2
> > processor--in general, neither the
> > ARM instruction set nor the "full" Thumb-2
> > instruction set will run on a
> > Cortex-M0 device because they are different
> > architectures. (Ok, the Thumb
> > subset will run on a Thumb-2-capable processor,
> > please don't be pedantic).
> >
> > Well, that's what I read from the article.
> >
> > --
> > Paul Curtis, Rowley Associates Ltd
> > http://www.rowley.co.uk
> > SolderCore running Defender...
> > http://www.vimeo.com/25709426
> >
> > _______________________________________________
> > eLua-dev mailing list
> > [hidden email]
> > https://lists.berlios.de/mailman/listinfo/elua-dev
> > >
>
> --
> - - Reklama - - - - - - - - - - - - - -
> Maturity, přijímací zkoušky, státnice, diplomky...
> Vše o závěru
> studia na středních a vysokých školách čtěte na
> portálu VOLNÝ.cz na
> http://bit.ly/LfImCR
>
>

Pito

Re: Extending MCU memory with SPI?

In reply to this post by Paul Curtis

> > Why is that a common understanding? SPIFI allows
> > you do to exactly this,
> > > if I understood correctly.
Does SPIFI map RAM area as well?? Or flash only?
Is the SPIFI a widely available today?
p.

--
- - Reklama - - - - - - - - - - - - - -
Maturity, přijímací zkoušky, státnice, diplomky... Vše o závěru
studia na středních a vysokých školách čtěte na portálu VOLNÝ.cz na
http://bit.ly/LfImCR

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

BogdanM

Re: Extending MCU memory with SPI?

In reply to this post by Pito

On Wed, Jun 6, 2012 at 1:47 PM, pito <[hidden email]> wrote:
> FYI - :)
> http://dmitry.co/index.php?p=./04.Thoughts/08. uM0
> http://dmitry.co/index.php?p=./04.Thoughts/07. Linux on 8bit

I believe we're talking about different things here. This is
emulation, a completely different topic. The AVR booting Linux thing
is spectacular, no doubt about that, but as far as I'm concerned it
has virtually no practical uses. Running machine code from external
serial memories might be more useful.

Best,
Bogdan

>
> A friend of mine ported the above linux stuff to pic32 so it is much
> faster. There is a LOT of people who do not chase the speed, but
> deal more with such crazy challenges as emulation/virtualisation. It
> is a pity the compiler developers do not support the 18037 standard
> :) broadly..
>
> ----- PŮVODNÍ ZPRÁVA -----
> Od: "Paul Curtis" <[hidden email]>
> Komu: "'eLua Users and Development List (www.eluaproject.net)'"
> <[hidden email]>
> Předmět: Re: [eLua-dev] Extending MCU memory with SPI?
> Datum: 6.6.2012 - 12:14:09
>
>> > http://www.trioflex.ee/index.php/services/portfolio/6-comp2hdmi
>> > >
>> > on the bottom of that PCB there is CortexM0 that
>> > is executing CortexM0
>> > > code from SPI flash...
>> >
>> > updates to those images are done via web browser
>> > access from WII console,
>> > > point to update page, start flash movie.. the
>> > sound is decoded in real
>> > > time while the M0 is generating onscreen
>> > graphics in background.. updates
>> > > are written to the SPI flash, and then new code
>> > is executing from spi
>> > > flash
>>
>> It doesn't actually say that it executes from SPI
>> flash. It states it runs
>> a virtualized CPU with the emulator fetching
>> virtual instructions from SPI
>> flash. At no point is hard Cortex-M0 running its
>> own instructions from SPI
>> flash. Note that the blurb is actually quite
>> loose because it says it
>> fetches ARM instructions for a Thumb-2
>> processor--in general, neither the
>> ARM instruction set nor the "full" Thumb-2
>> instruction set will run on a
>> Cortex-M0 device because they are different
>> architectures. (Ok, the Thumb
>> subset will run on a Thumb-2-capable processor,
>> please don't be pedantic).
>>
>> Well, that's what I read from the article.
>>
>> --
>> Paul Curtis, Rowley Associates Ltd
>> http://www.rowley.co.uk
>> SolderCore running Defender...
>> http://www.vimeo.com/25709426
>>
>> _______________________________________________
>> eLua-dev mailing list
>> [hidden email]
>> https://lists.berlios.de/mailman/listinfo/elua-dev
>>
>
> --
> - - Reklama - - - - - - - - - - - - - -
> Maturity, přijímací zkoušky, státnice, diplomky... Vše o závěru
> studia na středních a vysokých školách čtěte na portálu VOLNÝ.cz na
> http://bit.ly/LfImCR
>
> _______________________________________________
> eLua-dev mailing list
> [hidden email]
> https://lists.berlios.de/mailman/listinfo/elua-dev

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

BogdanM

Re: Extending MCU memory with SPI?

In reply to this post by Pito

On Wed, Jun 6, 2012 at 2:04 PM, pito <[hidden email]> wrote:
>> > Why is that a common understanding? SPIFI allows
>> > you do to exactly this,
>> > > if I understood correctly.
> Does SPIFI map RAM area as well?? Or flash only?
> Is the SPIFI a widely available today?

Read Paul's messages on this thread. It's not widely available, NXP
has an implementation (maybe there are others, I don't know). About
RAM: probably not. Even if it would support it, serial RAM chips are
extremely limited in capacity when compared to serial Flash chips.

Best,
Bogdan

> p.
>
> --
> - - Reklama - - - - - - - - - - - - - -
> Maturity, přijímací zkoušky, státnice, diplomky... Vše o závěru
> studia na středních a vysokých školách čtěte na portálu VOLNÝ.cz na
> http://bit.ly/LfImCR
>
_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Pito

Re: Extending MCU memory with SPI?

So SPIFI does not help much to eLua, then..

----- PŮVODNÍ ZPRÁVA -----
Od: "Bogdan Marinescu" <[hidden email]>
Komu: "pito" <[hidden email]>
Předmět: Re: [eLua-dev] Extending MCU memory with SPI?
Datum: 6.6.2012 - 13:11:17

> On Wed, Jun 6, 2012 at 2:04 PM, pito
> <[hidden email]> wrote:
> >> > Why is that a common understanding? SPIFI
> >> > allows
> >> > >> > you do to exactly this,
> >> > > if I understood correctly.
> > Does SPIFI map RAM area as well?? Or flash only?
> > Is the SPIFI a widely available today?
>
> Read Paul's messages on this thread. It's not
> widely available, NXP
> has an implementation (maybe there are others, I
> don't know). About
> RAM: probably not. Even if it would support it,
> serial RAM chips are
> extremely limited in capacity when compared to
> serial Flash chips.
>
> Best,
> Bogdan
>
> > p.
> >
> > --
> > - - Reklama - - - - - - - - - - - - - -
> > Maturity, přijímací zkoušky, státnice,
> > diplomky... Vše o závěru
> > > studia na středních a vysokých školách čtěte na
> > portálu VOLNÝ.cz na
> > > http://bit.ly/LfImCR
> >
>

Paul Curtis

Re: Extending MCU memory with SPI?

> So SPIFI does not help much to eLua, then..

You are clearly a man that has no imagination. :-)

SPIFI has a great advantage that you can execute out of it, masses of it,
but rather slowly. The fact is you could JIT into SPIFI and then run faster
than running interpreted bytecode, if eLua supported it. Or, you can write
bytecode to the flash and run it by "just reading through a regular pointer"
giving you a huge amount of code storage.

It also is directly addressable to add a huge amount of read-only storage
(something that I would welcome) in a fairly fast device. A ROM file system
in SPIFI flash has no place in a general embedded system?

Right. Does that help eLua? Someone with an imagination, please...

--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
SolderCore running Defender... http://www.vimeo.com/25709426

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Pito

Re: Extending MCU memory with SPI?

So how to run for example LIFE (150,150) on an LPCxyz with 4Mbyte
SPIFI flash memory?
p.

----- PŮVODNÍ ZPRÁVA -----
Od: "Paul Curtis" <[hidden email]>
Komu: "'pito'" <[hidden email]>, [hidden email],
[hidden email]
Předmět: RE: [eLua-dev] Extending MCU memory with SPI?
Datum: 6.6.2012 - 13:21:29

> > So SPIFI does not help much to eLua, then..
>
> You are clearly a man that has no imagination.
> :-)
>
> SPIFI has a great advantage that you can execute
> out of it, masses of it,
> but rather slowly. The fact is you could JIT into
> SPIFI and then run faster
> than running interpreted bytecode, if eLua
> supported it. Or, you can write
> bytecode to the flash and run it by "just reading
> through a regular pointer"
> giving you a huge amount of code storage.
>
> It also is directly addressable to add a huge
> amount of read-only storage
> (something that I would welcome) in a fairly fast
> device. A ROM file system
> in SPIFI flash has no place in a general embedded
> system?
>
> Right. Does that help eLua? Someone with an
> imagination, please...
>
> --
> Paul Curtis, Rowley Associates Ltd
> http://www.rowley.co.uk
> SolderCore running Defender...
> http://www.vimeo.com/25709426
>
>
>

BogdanM

Re: Extending MCU memory with SPI?

In reply to this post by Paul Curtis

On Wed, Jun 6, 2012 at 2:21 PM, Paul Curtis <[hidden email]> wrote:
>> So SPIFI does not help much to eLua, then..
>
> You are clearly a man that has no imagination. :-)
>
> SPIFI has a great advantage that you can execute out of it, masses of it,
> but rather slowly. The fact is you could JIT into SPIFI and then run faster
> than running interpreted bytecode, if eLua supported it. Or, you can write
> bytecode to the flash and run it by "just reading through a regular pointer"
> giving you a huge amount of code storage.

:) You beat me to it. eLua recently got a patch that allows it to do
exactly this: execute bytecode "in place", without loading it to RAM.
Also, it can keep strings partially in a ROM. Both features need a
memory that is directly accesible by the CPU, so this is a clear win
for SPIFI. Also: multiple ROM file systems, R/W flash file systems
(when we get to that), Lua modules for persistent storage, on-chip
firmware backups, multiboot :), all kinds of crazy stuff.

Best,
Bogdan

>
> It also is directly addressable to add a huge amount of read-only storage
> (something that I would welcome) in a fairly fast device. A ROM file system
> in SPIFI flash has no place in a general embedded system?
>
> Right. Does that help eLua? Someone with an imagination, please...
>
> --
> Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
> SolderCore running Defender... http://www.vimeo.com/25709426
>
>

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

BogdanM

Re: Extending MCU memory with SPI?

In reply to this post by Pito

On Wed, Jun 6, 2012 at 2:30 PM, pito <[hidden email]> wrote:
> So how to run for example LIFE (150,150) on an LPCxyz with 4Mbyte
> SPIFI flash memory?

You probably won't be able to do that. But, as mentioned in my
previous message, having a huge serial Flash memory mapped on your bus
can't be a bad thing :)

Best,
Bogdan

> p.
>
> ----- PŮVODNÍ ZPRÁVA -----
> Od: "Paul Curtis" <[hidden email]>
> Komu: "'pito'" <[hidden email]>, [hidden email],
> [hidden email]
> Předmět: RE: [eLua-dev] Extending MCU memory with SPI?
> Datum: 6.6.2012 - 13:21:29
>
>> > So SPIFI does not help much to eLua, then..
>>
>> You are clearly a man that has no imagination.
>> :-)
>>
>> SPIFI has a great advantage that you can execute
>> out of it, masses of it,
>> but rather slowly. The fact is you could JIT into
>> SPIFI and then run faster
>> than running interpreted bytecode, if eLua
>> supported it. Or, you can write
>> bytecode to the flash and run it by "just reading
>> through a regular pointer"
>> giving you a huge amount of code storage.
>>
>> It also is directly addressable to add a huge
>> amount of read-only storage
>> (something that I would welcome) in a fairly fast
>> device. A ROM file system
>> in SPIFI flash has no place in a general embedded
>> system?
>>
>> Right. Does that help eLua? Someone with an
>> imagination, please...
>>
>> --
>> Paul Curtis, Rowley Associates Ltd
>> http://www.rowley.co.uk
>> SolderCore running Defender...
>> http://www.vimeo.com/25709426
>>
>>
>>
>
> --
> - - Reklama - - - - - - - - - - - - - -
> Maturity, přijímací zkoušky, státnice, diplomky... Vše o závěru
> studia na středních a vysokých školách čtěte na portálu VOLNÝ.cz na
> http://bit.ly/LfImCR
>

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Paul Curtis

Re: Extending MCU memory with SPI?

In reply to this post by Pito

> So how to run for example LIFE (150,150) on an LPCxyz with 4Mbyte SPIFI
> flash memory?

150x150 bits fits in 4k of RAM to compute the next generation (2x2k blocks),
so you don't need SPIFI, just a compact implementation of Life. You are
clearly playing Devil's advocate. :-)

At present, SPIFI is directly-addressable predominantly-read-only mass
storage on LPC1800 microcontrollers. I do not believe there is a Quad-SPI
SRAM/MRAM/FRAM because there is no market for such a device, and even if
such a device did exist, it would be terribly costly and offer severely
limited storage compared to flash devices.

The OP of this requested how to extend memory using SPI devices. I think
we've clearly demonstrated that, in general, it is hard to do, but there are
microcontrollers where it is possible to undertake without much upheaval.

I think I'm going to let it rest here, and you eLuites can do what you want,
I really have no long-lived interest in this. :-)

--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
SolderCore running Defender... http://www.vimeo.com/25709426

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Pito

Re: Extending MCU memory with SPI?

In reply to this post by BogdanM

> You probably won't be able to do that. But, as
> mentioned in my
> previous message, having a huge serial Flash
> memory mapped on your bus
> can't be a bad thing :)
Yes, of course. As we discussed some time back the external ram for
f4discovery - I've got a few 2Mx16 10ns srams, but I discovered
afterwards the 100pin version of the stm32f4 is not best for that,
though. The FSMC on 144pin version maps all addresses straight, so
that would be a better candidate.. :(
p.

--
- - Reklama - - - - - - - - - - - - - -
Maturity, přijímací zkoušky, státnice, diplomky... Vše o závěru
studia na středních a vysokých školách čtěte na portálu VOLNÝ.cz na
http://bit.ly/LfImCR

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Pito

Re: Extending MCU memory with SPI?

In reply to this post by Paul Curtis

> I do not
> believe there is a Quad-SPI
> SRAM/MRAM/FRAM because there is no market for such
> a device, and even if
> such a device did exist, it would be terribly
> costly ..
Just FYI - eLua is not only mcu-based gadget which suffers ram, with
retrobsd we needed a swapdevice so I created a 4Mx8 'Octet-SPI-like"
device based on a xilinx low cost cpld (44pin, $1.5) and cypress
4Mx8 55ns sram ($20). You need only 11wires to connect this ramdisk
device to any MCU, the r/w transfer rate (pic32mx @80MHz) is
~7MBytes/sec when accessing a 512bytes large block, an int32 could
be accessed in ~700ns. When using 10ns sram you might be 2x faster.
http://retrobsd.org/topic/swapdisk-v1-1-is-4mbx8-and-alive/

p/

--
- - Reklama - - - - - - - - - - - - - -
Maturity, přijímací zkoušky, státnice, diplomky... Vše o závěru
studia na středních a vysokých školách čtěte na portálu VOLNÝ.cz na
http://bit.ly/LfImCR

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Paul Curtis

Re: Extending MCU memory with SPI?

Hi,

> > I do not
> > believe there is a Quad-SPI
> > SRAM/MRAM/FRAM because there is no market for such a device, and even
> > if such a device did exist, it would be terribly costly ..
> Just FYI - eLua is not only mcu-based gadget which suffers ram, with
> retrobsd we needed a swapdevice so I created a 4Mx8 'Octet-SPI-like"
> device based on a xilinx low cost cpld (44pin, $1.5) and cypress
> 4Mx8 55ns sram ($20). You need only 11wires to connect this ramdisk device
> to any MCU, the r/w transfer rate (pic32mx @80MHz) is ~7MBytes/sec when
> accessing a 512bytes large block, an int32 could be accessed in ~700ns.
> When using 10ns sram you might be 2x faster.
> http://retrobsd.org/topic/swapdisk-v1-1-is-4mbx8-and-alive/

Nice! :-) I'm disappointed with the EBI of the LM3S9B92, and the 9D92
isn't much better really...

--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
SolderCore running Defender... http://www.vimeo.com/25709426

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev

Pito

Re: Extending MCU memory with SPI?

> Nice! :-) I'm disappointed with the EBI of the
> LM3S9B92, and the 9D92
> isn't much better really...

We have spent tremendous amount time to develop something similar to
EPI on Stellaris a year ago

http://retrobsd.org/topic/static-ram-for-swap-space/page/2/

even not knowing the EPI actually does exist.. :) (see #681
message)

Not succesfull yet with sdrams. We abandoned the sdram EPI stuff,
the only working gadget is the srams swapdevice with xilinx today..
p.

--
- - Reklama - - - - - - - - - - - - - -
Maturity, přijímací zkoušky, státnice, diplomky... Vše o závěru
studia na středních a vysokých školách čtěte na portálu VOLNÝ.cz na
http://bit.ly/LfImCR

_______________________________________________
eLua-dev mailing list
[hidden email]
https://lists.berlios.de/mailman/listinfo/elua-dev