New microprocessor architecture

"rickman" <gnuarm@gmail.com> wrote in message 
news:1173118005.268771.265510@8g2000cwh.googlegroups.com...
On Mar 5, 4:44 am, torb...@app-0.diku.dk (Torben &#4294967295;gidius Mogensen)
wrote:
> "rickman" <gnu...@gmail.com> writes:
> > On Mar 4, 6:31 am, already5cho...@yahoo.com wrote:

>> I don't think patents or not makes much of a difference.  What you
>> need is a convincing proof of concept -- few are willing to invest in
>> something completely new unless they can be sure it will be
>> worthwhile, and just having a paper design of an architecture with no
>> implementation, support chips, compilers, operating systems, etc. will
>> require massive investments to make commercially useful.

Indeed. Selling a new unproven idea is much harder than an improvement
of an existing idea that complements what a company is already doing.
If you made a better branch predictor you would find many interested 
parties.
Dataflow architectures are outside mainstream and will require a lot of
effort to show that they compete with conventional hardware and software.
Most companies prefer solutions not unsolved (or unsolveable!) problems.

So I think it is pretty unlikely someone just takes the idea and runs with 
it -
having a patent on it or not is pretty much irrelevant as the patent on its
own is only a tiny part of a complete working system. I think the best way
is to set up your own company and do it yourself. Once you've got it all
working you can start thinking about selling the proven technology to other
companies.

> I don't see why you are using ARM as an example.  That is my example
> of how they are preventing the world from copying their famously
> popular instruction set as compared to Intel who allowed their 8051
> and x386 ISAs to be copied freely.  As a result Intel has lots of
> direct competition for sockets and ARM is looking like the 32 bit 8051
> except that they make money off of every chip built.

The difference between ARM and x86 is that ARM doesn't have fabs,
let alone ones that are the best in the world, so it can only sell IP. The
patents are needed to protect this IP. In addition, an ARM design takes
far less effort to copy than an x86 design, so the risk of someone making
a competitive clone without paying ARM is much higher. Companies
have tried making x86 CPUs but nobody had enough resources to
compete with Intel/AMD, so no patent protection is required for x86.

>The only real
> barrier to potential customers developing or using open source ARM
> designs is the patent ARM has on an essential circuit I am told.  When
> that runs out in another 10 or so years, ARM7 will be freely copied if
> anyone still has interest in it.  Maybe that is why they have come out
> with the Cortex M3, to start a new patent clock ticking?

Not really. ARM7 is 12 years old and already at the end of its useful
life, so it was becoming urgent to design something better. It's always
better to obsolete your own products rather than let a competitor do it!

Incidentally, ARM2 is now over 20 years old, so the earliest patents
must be close to running out. Will anyone bother cloning an ARM2?
I think it's very unlikely - ARM2 was obsoleted by ARM7 in 1995.
So few will be interested in Thumb-1 when the Thumb patents run out.

Of course new patents apply to the Cortex architecture to protect the
investment, but it's the technology that attracts ARM's customers in
the first place, not whether it is patented or not.

Wilco 

On Mar 5, 3:46 pm, "Wilco Dijkstra" <Wilco_dot_Dijks...@ntlworld.com>
wrote:
> "rickman" <gnu...@gmail.com> wrote in message
>
> news:1173118005.268771.265510@8g2000cwh.googlegroups.com...
> On Mar 5, 4:44 am, torb...@app-0.diku.dk (Torben =C6gidius Mogensen)
> wrote:
>
> > "rickman" <gnu...@gmail.com> writes:
> So I think it is pretty unlikely someone just takes the idea and runs with
> it -
> having a patent on it or not is pretty much irrelevant as the patent on i=
ts
> own is only a tiny part of a complete working system. I think the best way
> is to set up your own company and do it yourself. Once you've got it all
> working you can start thinking about selling the proven technology to oth=
er
> companies.

I'm not sure I understand what you are saying.  If you are saying that
there is much more than just patent to makeing money on an invention,
then we agree.  If you are saying you don't need a patent, we disagree
totally.  The patent is the absolute requirement on a technical
invention if you don't want people to copy your work and make money
without giving you a cut.


> > I don't see why you are using ARM as an example.  That is my example
> > of how they are preventing the world from copying their famously
> > popular instruction set as compared to Intel who allowed their 8051
> > and x386 ISAs to be copied freely.  As a result Intel has lots of
> > direct competition for sockets and ARM is looking like the 32 bit 8051
> > except that they make money off of every chip built.
>
> The difference between ARM and x86 is that ARM doesn't have fabs,
> let alone ones that are the best in the world, so it can only sell IP. The
> patents are needed to protect this IP.

I disagree.  They could protect the IP with any number of mechanisms
such as encrypted libraries, non-disclosure agreements, etc.  But the
invention could be duplicated by simply reading the ISA documentation
and designing your own... if there were no patents involved.  My
understanding is that ARM has patented some feature of the way they
handle interrupts (not sure I am getting this right) that is closely
tied to the instruction set.  So to implement the instruction set you
*have* to violate the patent.

If ARM did not have a patent, their chips would go the path of the
8051 and be even more widely popular than they are now, but like Intel
gets nothing from the 8051 clones, ARM would get nothing.


> In addition, an ARM design takes
> far less effort to copy than an x86 design, so the risk of someone making
> a competitive clone without paying ARM is much higher. Companies
> have tried making x86 CPUs but nobody had enough resources to
> compete with Intel/AMD, so no patent protection is required for x86.

It has nothing to do with fabs or the size of the company.  BTW, did
you try to slip one past us by saying Intel/AMD as if they were one
company???  Doesn't AMD *compete* with Intel in the x86 market???
Sure the others may have gone by the wayside as happens in so many
electronics sectors, but AMD is eating Intel's lunch and it is all
because there were no patents that could be used to stop AMD.  Oh,
Intel tried!  They spent several years in court and AMD had to
redesign the microcode for their version of the 486.  But from then on
AMD had free reign to make work-alike chips with varying periods of
boom and bust.  The last few years have seen AMD nibbling away at
Intel market share and now they are starting to be a force in the
server market.  Too bad Intel didn't have a patent to protect their
archetecture.


> >The only real
> > barrier to potential customers developing or using open source ARM
> > designs is the patent ARM has on an essential circuit I am told.  When
> > that runs out in another 10 or so years, ARM7 will be freely copied if
> > anyone still has interest in it.  Maybe that is why they have come out
> > with the Cortex M3, to start a new patent clock ticking?
>
> Not really. ARM7 is 12 years old and already at the end of its useful
> life, so it was becoming urgent to design something better. It's always
> better to obsolete your own products rather than let a competitor do it!

Funny, when I talk to the vendors of ARM MCUs they tell me that they
have great new ARM7 parts coming out nearly every day.  So far only
Luminary has a CM3 chip line out and they can't compete with the ARM7
except in limited areas.  One way their parts are very uncompetitive
is in power consumption.  But we'll have to wait for others to produce
the CM3 in more modern technologies.


> Incidentally, ARM2 is now over 20 years old, so the earliest patents
> must be close to running out. Will anyone bother cloning an ARM2?
> I think it's very unlikely - ARM2 was obsoleted by ARM7 in 1995.
> So few will be interested in Thumb-1 when the Thumb patents run out.

That is the point of continuing to reinvent the archtecture and aquire
new patents... to continue to lock out competition from using the ISA
that you made popular.


> Of course new patents apply to the Cortex architecture to protect the
> investment, but it's the technology that attracts ARM's customers in
> the first place, not whether it is patented or not.

You just don't understand the business issues.  Sure, customers don't
care about your patents.  But that is not what they are for.  A patent
protects your inventions against being copied by your competition.
That means they have to do their own work and not ride on your
coattails.  Then your customers say with you to get the inventions if
they are better than the competitor's inventions (along with all the
other stuff that goes with MCUs such as support, knowledge base,
tools, etc).

This all seems pretty simple and obvious to me.  Are you pulling my
leg?

rickman wrote:
> On Mar 5, 3:46 pm, "Wilco Dijkstra" <Wilco_dot_Dijks...@ntlworld.com>
> wrote:
> 
>>
>>Not really. ARM7 is 12 years old and already at the end of its useful
>>life, so it was becoming urgent to design something better. It's always
>>better to obsolete your own products rather than let a competitor do it!
> 
> 
> Funny, when I talk to the vendors of ARM MCUs they tell me that they
> have great new ARM7 parts coming out nearly every day.  So far only
> Luminary has a CM3 chip line out and they can't compete with the ARM7
> except in limited areas.  One way their parts are very uncompetitive
> is in power consumption.  But we'll have to wait for others to produce
> the CM3 in more modern technologies.

The 32 bit market is not all going ARMs way, and the protection
on cores is less of a road block than it was, and getting less over 
time. (see below)


> 
>>Of course new patents apply to the Cortex architecture to protect the
>>investment, but it's the technology that attracts ARM's customers in
>>the first place, not whether it is patented or not.
> 
> 
> You just don't understand the business issues.  Sure, customers don't
> care about your patents.  But that is not what they are for.  A patent
> protects your inventions against being copied by your competition.

Not quite : A patent is a license to litigate.
Yes, it gives some protection, but there are other ways to compete
than a slavish-clone, and it is no help then.

> That means they have to do their own work and not ride on your
> coattails.  Then your customers say with you to get the inventions if
> they are better than the competitor's inventions (along with all the
> other stuff that goes with MCUs such as support, knowledge base,
> tools, etc).

and that 'other stuff' is getting easier to clone these days.

Freescale are about the spin their Coldfire-V1 core,(and did a new 
Starcore recently), Zilog have Zneo, MIPS refuses to go away, Atmel have 
AVR32, and DSP devices from Analog Devices, and TI continue to hold the 
high ground on speed.

Lattice have an open source 32 bit CPU - so the core itself
is less and less of a hold over customers.

Of course, a core-vendor will claim the core is everything :)

-jg

On Mon, 05 Mar 2007 10:06:45 -0800, rickman wrote:
> On Mar 5, 4:44 am, torb...@app-0.diku.dk (Torben �gidius Mogensen) wrote:
>> "rickman" <gnu...@gmail.com> writes:
>> > On Mar 4, 6:31 am, already5cho...@yahoo.com wrote:
>> >> If you really want people to use it giving up on patent would be a
>> >> good starting point.
>>
>> > Yeah, I bet that is the way that ARM got people to use their
>> > architecture, giving up their patents!
>>
>> ARM had a proof of concept in the Archimedes computer from Acorn, which
>> persuaded Apple to invest in the CPU architecture (intending to use it
>> for their Newton PDA and, romours tell, for a home computer that never
>> saw the light of day).
> 
> Perhaps you missed my point.  I am not saying that a Patent is everything.
>  Read my original post here.  I am saying that a Patent is required to
> prevent others from stealing your invention, ...

This is not entirely true. A patent _and more money than the other guy_
are required to prevent others from stealing your invention.   ;-)

Good Luck!
Rich

On Mar 5, 7:33 pm, Rich Grise <r...@example.net> wrote:
> On Mon, 05 Mar 2007 10:06:45 -0800, rickman wrote:
> > Perhaps you missed my point.  I am not saying that a Patent is everything.
> >  Read my original post here.  I am saying that a Patent is required to
> > prevent others from stealing your invention, ...
>
> This is not entirely true. A patent _and more money than the other guy_
> are required to prevent others from stealing your invention.   ;-)

I understand your point, but that it not really true.  It is true that
deep pockets can throw up a lot of legal roadblocks that take
financing to overcome.  But if you have a patent that is clearly worth
something, you can get backers.  It does not take a billion dollars to
fight GM... in fact there was a case where a gentleman got a patent on
the automatic intermittent windshield wiper control.  IIRC, Ford would
not pay him royalties and used the invention.  It took the guy
something like 15 or 20 years, but he eventually did win and I believe
they had to pay his legal costs and damages.

We had this conversation about how good/bad patents were a short time
ago in another group.  I really don't want to get into that.  I will
suffice to say that patents are useful and often a necessary way to
protect one's invention.  Then you still need a lot of other stuff to
make money from that invention... including money!

Wilco Dijkstra wrote:
[...]
> Companies have tried making x86 CPUs but nobody had enough resources to
> compete with Intel/AMD, so no patent protection is required for x86.

Au contraire, x86 is probably the architecture with the most patents 
covering it. It's impossible to make an x86-compatible CPU without a 
licence, as many ISA features are patented. Intel and AMD have a 
cross-licencing agreement regarding x86 (any x86-related patents that Intel 
gets AMD can use and vice versa), and various licencing agreements exist 
with other companies regarding x86 as well (IBM, Via, Transmeta, probably a 
few others). The ones that I know of require per-CPU royalty payments to 
Intel.

This cross-licencing with AMD is suspected to be part of the reason Intel 
was/is throwing so much money at Itanium to keep it afloat. If IA64 becomes 
more widespread than x86, they have AMD over a barrel since the 
cross-licencing agreements don't allow AMD to create an IA64 CPU. 
Unsurprisingly, AMD reacted with x86-64, and so far IA64 is showing no signs 
of an imminent takeover ...

[...]

-- 
Michael Brown
Add michael@ to emboss.co.nz - My inbox is always open 

On 2007-03-06, Jim Granville <no.spam@designtools.maps.co.nz> wrote:

> and that 'other stuff' is getting easier to clone these days.
>
> Freescale are about the spin their Coldfire-V1 core,(and did a new 
> Starcore recently), Zilog have Zneo, MIPS refuses to go away, Atmel have 
> AVR32, and DSP devices from Analog Devices, and TI continue to hold the 
> high ground on speed.
>
> Lattice have an open source 32 bit CPU - so the core itself
> is less and less of a hold over customers.
>
> Of course, a core-vendor will claim the core is everything :)

I'm working on a product line that just migrated from ARM7 to
NIOS2.  Nobody cares about the CPU core.  We're using the same
OS and same compiler.  Changing CPU cores was trivial -- and it
even involved going from big-endian to little-endian.

Changing Ethernet controllers and various other peripherals was
the painful part. ;)

It's the peripherals where all the work is -- the CPU just
doesn't matter.

-- 
Grant Edwards                   grante             Yow!  I wonder if I should
                                  at               put myself in ESCROW!!
                               visi.com            

> This is not entirely true. A patent _and more money than the other guy_
> are required to prevent others from stealing your invention.   ;-)

That's a particular deficiency of the US legal system. In some other relevant 
parts of the world, a patent and a small, predictable amount of money are 
quite enough.

	Jan

"rickman" <gnuarm@gmail.com> wrote in message 
news:1173134382.140402.93880@n33g2000cwc.googlegroups.com...
On Mar 5, 3:46 pm, "Wilco Dijkstra" <Wilco_dot_Dijks...@ntlworld.com>
wrote:
> "rickman" <gnu...@gmail.com> wrote in message
>
> news:1173118005.268771.265510@8g2000cwh.googlegroups.com...
> On Mar 5, 4:44 am, torb...@app-0.diku.dk (Torben &#4294967295;gidius Mogensen)
> wrote:
>
> > "rickman" <gnu...@gmail.com> writes:
>> So I think it is pretty unlikely someone just takes the idea and runs 
>> with
>> it -
>> having a patent on it or not is pretty much irrelevant as the patent on 
>> its
>> own is only a tiny part of a complete working system. I think the best 
>> way
>> is to set up your own company and do it yourself. Once you've got it all
>> working you can start thinking about selling the proven technology to 
>> other
>> companies.

> I'm not sure I understand what you are saying.  If you are saying that
> there is much more than just patent to makeing money on an invention,
> then we agree.  If you are saying you don't need a patent, we disagree
> totally.  The patent is the absolute requirement on a technical
> invention if you don't want people to copy your work and make money
> without giving you a cut.

I'm saying that you don't need a patent immediately. Only when you've
got a product that works (and we agree that takes a lot more than just
having a patent) and are ready to sell you need patents.

However a patent means the invention is published, so anyone (say
in a country where you didn't patent) could copy it easily. World-wide
patents are very expensive, so most only patent in the US and UK.
Patents are very useful when you're selling things that can be easily
copied or reverse engineered, but they are a double-edged sword.

>> The difference between ARM and x86 is that ARM doesn't have fabs,
>> let alone ones that are the best in the world, so it can only sell IP. 
>> The
>> patents are needed to protect this IP.
>
> I disagree.  They could protect the IP with any number of mechanisms
> such as encrypted libraries, non-disclosure agreements, etc.  But the
> invention could be duplicated by simply reading the ISA documentation
> and designing your own... if there were no patents involved.

Yes, without patents the IP would be unprotected. I think you agree...

> If ARM did not have a patent, their chips would go the path of the
> 8051 and be even more widely popular than they are now, but like Intel
> gets nothing from the 8051 clones, ARM would get nothing.

Indeed. My guess of what would happen is that everybody creates
incompatible extensions and there is less incentive to invest into
innovative designs as they are easily cloned. We'd still use ARM2's...

>> In addition, an ARM design takes
>> far less effort to copy than an x86 design, so the risk of someone making
>> a competitive clone without paying ARM is much higher. Companies
>> have tried making x86 CPUs but nobody had enough resources to
>> compete with Intel/AMD, so no patent protection is required for x86.
>
> It has nothing to do with fabs or the size of the company.  BTW, did
> you try to slip one past us by saying Intel/AMD as if they were one
> company???  Doesn't AMD *compete* with Intel in the x86 market???

Intel and AMD are a duopoly, and yes it's great there is fierce competition
between them (thanks for reminding me), but that's not the subject of
discussion. My point is that the cost of making an x86 clone is far higher
than making an ARM clone, so protection is more important for ARM.

Fabs are an important difference: if you have a fab you sell chips to
end users, not IP to companies with fabs. Without patents you can still
sell chips (end users can't make chips themselves), but it becomes a
lot harder to sell IP to a fab company (they can make the IP themselves
and avoid royalties). Selling IP is hard with patents - and impossible 
without.

>> Not really. ARM7 is 12 years old and already at the end of its useful
>> life, so it was becoming urgent to design something better. It's always
>> better to obsolete your own products rather than let a competitor do it!
>
> Funny, when I talk to the vendors of ARM MCUs they tell me that they
> have great new ARM7 parts coming out nearly every day.  So far only

They would say that, wouldn't they? If you read ARM's latest financial
statements you can see that licensing of ARM7 is clearly declining fast
(under 5% of total license revenue in 2006, down from 18% in 2002).
That doesn't imply it won't stay popular for the next few years, but ARM9
and Cortex-M3 are appearing fast.

> Luminary has a CM3 chip line out and they can't compete with the ARM7
> except in limited areas.  One way their parts are very uncompetitive
> is in power consumption.  But we'll have to wait for others to produce
> the CM3 in more modern technologies.

Yes, I'd like to see ARM7 and Cortex-M3 on the same process

>> Incidentally, ARM2 is now over 20 years old, so the earliest patents
>> must be close to running out. Will anyone bother cloning an ARM2?
>> I think it's very unlikely - ARM2 was obsoleted by ARM7 in 1995.
>> So few will be interested in Thumb-1 when the Thumb patents run out.
>
> That is the point of continuing to reinvent the archtecture and aquire
> new patents... to continue to lock out competition from using the ISA
> that you made popular.

That's totally wrong. Improving the architecture is essential - without it
nobody would be using ARM today. An ARM2 cannot compete with
MIPS or any other modern CPU. So the goal is to compete, not to
stifle competition. There is a lot of competition in the 32-bit space,
and even within the ARM community there is fierce competition.

The patents are there to protect the investment and clearly don't
lock out competition - they just stop cloners.

Wilco 

"Wilco Dijkstra" <Wilco_dot_Dijkstra@ntlworld.com> writes:

> Incidentally, ARM2 is now over 20 years old, so the earliest patents
> must be close to running out. Will anyone bother cloning an ARM2?
> I think it's very unlikely - ARM2 was obsoleted by ARM7 in 1995.

While ARM2 has been superseded by later processors, I don't think it
is useless and not wortwhile to copy.  Granted, it can not compete
with the later ARM processors, but it is smaller, so it could be
useful for embedded or low-cost handheld designs (or something like
the $100 laptop).  And there is a lot of software (from the Archimedes
homecomputer) that can run on ARM2.

I'm not saying people would rush to do it as soon as the patents run
out, but I wouldn't rule the possibility out.

	Torben