FPGA Central

Hi,
Coming from a scientific computing standpoint (with no hardware
experience).
I was wondering if you can improve any dedicated tasks by designing a
special
purpose chips ala FPGA to run your code? Does anyone have any
experience
with this?

[email protected] wrote:
> Hi,
> Coming from a scientific computing standpoint (with no hardware
> experience).
> I was wondering if you can improve any dedicated tasks by designing a
> special
> purpose chips ala FPGA to run your code? Does anyone have any
> experience
> with this?

Yes.

Tommy

[email protected] wrote:
> Hi,
> Coming from a scientific computing standpoint (with no hardware
> experience).
> I was wondering if you can improve any dedicated tasks by designing a
> special
> purpose chips ala FPGA to run your code? Does anyone have any
> experience
> with this?

Bioinformatics, specifically Timelogic used FPGAs in the DeCypher
engine many years ago, Timelogic recently changed hands.

Also recently we saw product introductions from two vendors for Opteron
coprocessors for the 2nd socket using Altera and Xilinx FPGAs, no doubt
these will be used to hardwire some interesting algorithms. Also Cray
computer bought out a small Opteron server company (Octiga Bay) with
FPGAs on board rather than in the socket.

We had Mersenne primes on FPGA v highend PCs, FPGAs don't always win on
cost, usually a matter of managing bandwidth, FPGAs sometimes better,
sometimes not.

For the hardware shy, there are quite a few C based products to map
software to hardware, even Fortran has been used (I'm told).

Google fpga with any scientific term and see what pops up.

John Jakson

"Tommy Thorn" <[email protected]> wrote in message
news:[email protected] oups.com...
> [email protected] wrote:
>> Hi,
>> Coming from a scientific computing standpoint (with no hardware
>> experience).
>> I was wondering if you can improve any dedicated tasks by designing a
>> special
>> purpose chips ala FPGA to run your code? Does anyone have any
>> experience
>> with this?
>
> Yes.
>
> Tommy
>

:-)

Google this.

fpga scientific computing

Guess what? Nearly half a million hits! So, I added 'lumberjack' to the
search and got it down to 44 hits.
HTH, Syms.

Everything what can parallelize is possibly running faster in Hardware
than in Software.


[email protected] wrote:
> Hi,
> Coming from a scientific computing standpoint (with no hardware
> experience).
> I was wondering if you can improve any dedicated tasks by designing a
> special
> purpose chips ala FPGA to run your code? Does anyone have any
> experience
> with this?

"Symon" <[email protected]> wrote in message
news:454aa716$[email protected]..
> "Tommy Thorn" <[email protected]> wrote in message
> news:[email protected] oups.com...
>> [email protected] wrote:
>>> I was wondering if you can improve any dedicated tasks by designing a
>>> special purpose chips ala FPGA to run your code? Does anyone have any
>>> experience with this?
>> Yes.
>
> :-)
>
> Google this.
>
> fpga scientific computing
>
> Guess what? Nearly half a million hits! So, I added 'lumberjack' to the
> search and got it down to 44 hits.

Very good.

Here are some other useful qualifiers and their respective scores:

FPGA scientific computing mongoose 49
FPGA scientific computing pokemon 142
FPGA scientific computing stradivarius 9
FPGA scientific computing zoroastrianism 95,600
FPGA scientific computing paragliding 936
FPGA scientific computing OMG ponies 15
FPGA scientific computing Kellogg's 285
FPGA scientific computing jello 16,200
FPGA scientific computing sefirot 1

FPGA scientific computing boring 103,000
FPGA scientific computing interesting 231,000

Hope that helps,

-Ben-

[email protected] wrote:
>Everything what can parallelize is possibly running faster in Hardware
>than in Software.

And pipelined.

On Fri, 3 Nov 2006 09:35:26 -0000, "Ben Jones" <[email protected]>
wrote:

>FPGA scientific computing sefirot 1

A Googlewhackblat! But, you wrecked it. There'll be 2 hits now.

We have a reason number of customers playing with FPGAs on development
boards as scientific and maths co-processors. It is happening in a
surprising number of applications and markets.

John Adair
Enterpoint Ltd. - Home of Broaddown4. The Ultimate Virtex-4 Development
Board.
http://www.enterpoint.co.uk

<[email protected]> wrote in message
news:[email protected] oups.com...
> Hi,
> Coming from a scientific computing standpoint (with no hardware
> experience).
> I was wondering if you can improve any dedicated tasks by designing a
> special
> purpose chips ala FPGA to run your code? Does anyone have any
> experience
> with this?
>

[email protected] wrote:

> Coming from a scientific computing standpoint (with no hardware
> experience).
> I was wondering if you can improve any dedicated tasks by designing a
> special
> purpose chips ala FPGA to run your code? Does anyone have any
> experience
> with this?

I think an interesting application are neural networks:

http://www.google.com/search?q=neural+network+fpga

The speedup compared to normal CPUs should be very high: If you synthize a
small network with e.g. 1,000 interconnections on the FPGA, a FPGA clocked
with 100 MHz would process the neural network as fast as a normal CPU
clocked with over 100 GHz.

--
Frank Buss, [email protected]
http://www.frank-buss.de, http://www.it4-systems.de

[email protected] wrote:
> Hi,
> Coming from a scientific computing standpoint (with no hardware
> experience).
> I was wondering if you can improve any dedicated tasks by designing a
> special
> purpose chips ala FPGA to run your code? Does anyone have any
> experience
> with this?

National Instruments do plug-in FPGA cards for their LabView product.

Dave

[email protected] wrote:

>Everything what can parallelize is possibly running faster in Hardware
>than in Software.

Still, the amount of processing power a modern PC processor can
deliver is enormous. It is problably more cost effective to optimize
an algorithm to run parallel on 10 PC's than to develop a specific
FPGA solution. If space is a constraint, the answer is in using blade
servers.

--
Reply to nico@nctdevpuntnl (punt=.)
Bedrijven en winkels vindt U op www.adresboekje.nl

Nico Coesel wrote:
> [email protected] wrote:
>
> >Everything what can parallelize is possibly running faster in Hardware
> >than in Software.
>
> Still, the amount of processing power a modern PC processor can
> deliver is enormous. It is problably more cost effective to optimize
> an algorithm to run parallel on 10 PC's than to develop a specific
> FPGA solution. If space is a constraint, the answer is in using blade
> servers.
>

This is often true even for obvious applications of FPGA, PC can still
be better.

The analysis needs to account for factors such as floating point use,
high locality referencing or lots of computing on tiny data sets and
use of standard 32, 64 bit operands favors PCs On the other hand bit
mangling, very high rates of memory shuffling, use of unusual bit wise
or odd size mathematical operators and strategic use of buffer memories
are a win for FPGAs.

I suspect that as Flash drives replace hard drives at the 30GByte level
the case for hardware taking control of data management in Flash only
increases since a hardware FPGA design wouldn't usually include disk
systems and can then push the PC host further away.

Some problems can be factored both ways to get similar results in which
case PCs still have the familiarity edge. Still a PC based solution
that is just recompiled to hardware will perform far less well since
it has already been factored to favor the PC.

For a hardware solution, the software version can be viewed as the
simulation of same hardware and most simulations usually run many
orders slower than real hardware.

This is why the best hardware solution is unlikeley to ever be achieved
by software folks who have little experience in hardware who would
force the FPGA to emulate the PC strengths.

John Jakson
transputer guy

The other area where FPGA's win is in latency. Parallel or networked PC's
cannot meet the latency requirements of many large computationally intensive
real-time systems, even though they can meet the throughput requirements

Marco
________________________
Marc Reinig
UCO/Lick Observatory
Laboratory for Adaptive Optics


"JJ" <[email protected]> wrote in message
news:[email protected] ups.com...
>
> Nico Coesel wrote:
>> [email protected] wrote:
>>
>> >Everything what can parallelize is possibly running faster in Hardware
>> >than in Software.
>>
>> Still, the amount of processing power a modern PC processor can
>> deliver is enormous. It is problably more cost effective to optimize
>> an algorithm to run parallel on 10 PC's than to develop a specific
>> FPGA solution. If space is a constraint, the answer is in using blade
>> servers.
>>
>
> This is often true even for obvious applications of FPGA, PC can still
> be better.
>
> The analysis needs to account for factors such as floating point use,
> high locality referencing or lots of computing on tiny data sets and
> use of standard 32, 64 bit operands favors PCs On the other hand bit
> mangling, very high rates of memory shuffling, use of unusual bit wise
> or odd size mathematical operators and strategic use of buffer memories
> are a win for FPGAs.
>
> I suspect that as Flash drives replace hard drives at the 30GByte level
> the case for hardware taking control of data management in Flash only
> increases since a hardware FPGA design wouldn't usually include disk
> systems and can then push the PC host further away.
>
> Some problems can be factored both ways to get similar results in which
> case PCs still have the familiarity edge. Still a PC based solution
> that is just recompiled to hardware will perform far less well since
> it has already been factored to favor the PC.
>
> For a hardware solution, the software version can be viewed as the
> simulation of same hardware and most simulations usually run many
> orders slower than real hardware.
>
> This is why the best hardware solution is unlikeley to ever be achieved
> by software folks who have little experience in hardware who would
> force the FPGA to emulate the PC strengths.
>
> John Jakson
> transputer guy
>

Thanks for the speedy reply everyone, I appreciate the help.

~Lance

On Nov 3, 10:26 am, "Marc Reinig" <M...@newsgroups.nospam> wrote:
> The other area where FPGA's win is in latency. Parallel or networked PC's
> cannot meet the latency requirements of many large computationally intensive
> real-time systems, even though they can meet the throughput requirements
>
> Marco
> ________________________
> Marc Reinig
> UCO/Lick Observatory
> Laboratory for Adaptive Optics
>
> "JJ" <johnjak...@gmail.com> wrote in messagenews:[email protected] oglegroups.com...
>
>
>
> > Nico Coesel wrote:
> >> helmut.leonha...@gmail.com wrote:
>
> >> >Everything what can parallelize is possibly running faster in Hardware
> >> >than in Software.
>
> >> Still, the amount of processing power a modern PC processor can
> >> deliver is enormous. It is problably more cost effective to optimize
> >> an algorithm to run parallel on 10 PC's than to develop a specific
> >> FPGA solution. If space is a constraint, the answer is in using blade
> >> servers.
>
> > This is often true even for obvious applications of FPGA, PC can still
> > be better.
>
> > The analysis needs to account for factors such as floating point use,
> > high locality referencing or lots of computing on tiny data sets and
> > use of standard 32, 64 bit operands favors PCs On the other hand bit
> > mangling, very high rates of memory shuffling, use of unusual bit wise
> > or odd size mathematical operators and strategic use of buffer memories
> > are a win for FPGAs.
>
> > I suspect that as Flash drives replace hard drives at the 30GByte level
> > the case for hardware taking control of data management in Flash only
> > increases since a hardware FPGA design wouldn't usually include disk
> > systems and can then push the PC host further away.
>
> > Some problems can be factored both ways to get similar results in which
> > case PCs still have the familiarity edge. Still a PC based solution
> > that is just recompiled to hardware will perform far less well since
> > it has already been factored to favor the PC.
>
> > For a hardware solution, the software version can be viewed as the
> > simulation of same hardware and most simulations usually run many
> > orders slower than real hardware.
>
> > This is why the best hardware solution is unlikeley to ever be achieved
> > by software folks who have little experience in hardware who would
> > force the FPGA to emulate the PC strengths.
>
> > John Jakson
> > transputer guy

Marc Reinig wrote:
> The other area where FPGA's win is in latency. Parallel or networked PC's
> cannot meet the latency requirements of many large computationally intensive
> real-time systems, even though they can meet the throughput requirements
>
> Marco

Ofcourse, whipping up a quick list of platform benefits bound to miss
some obvious ones.

Also both platforms continually changing/improving but also getting
harder to understand and less transparent. That alone makes it harder
to compare since one mans optimized code is another mans starting
point.

Still the basic decider is likely to be whether one knows how to even
start a FPGA. project.

> ________________________
> Marc Reinig
> UCO/Lick Observatory
> Laboratory for Adaptive Optics
>
>
> "JJ" <[email protected]> wrote in message
> news:[email protected] ups.com...
> >
> > Nico Coesel wrote:
> >> [email protected] wrote:
> >>
> >> >Everything what can parallelize is possibly running faster in Hardware
> >> >than in Software.
> >>
> >> Still, the amount of processing power a modern PC processor can
> >> deliver is enormous. It is problably more cost effective to optimize
> >> an algorithm to run parallel on 10 PC's than to develop a specific
> >> FPGA solution. If space is a constraint, the answer is in using blade
> >> servers.
> >>
> >
> > This is often true even for obvious applications of FPGA, PC can still
> > be better.
> >
> > The analysis needs to account for factors such as floating point use,
> > high locality referencing or lots of computing on tiny data sets and
> > use of standard 32, 64 bit operands favors PCs On the other hand bit
> > mangling, very high rates of memory shuffling, use of unusual bit wise
> > or odd size mathematical operators and strategic use of buffer memories
> > are a win for FPGAs.
> >
> > I suspect that as Flash drives replace hard drives at the 30GByte level
> > the case for hardware taking control of data management in Flash only
> > increases since a hardware FPGA design wouldn't usually include disk
> > systems and can then push the PC host further away.
> >
> > Some problems can be factored both ways to get similar results in which
> > case PCs still have the familiarity edge. Still a PC based solution
> > that is just recompiled to hardware will perform far less well since
> > it has already been factored to favor the PC.
> >
> > For a hardware solution, the software version can be viewed as the
> > simulation of same hardware and most simulations usually run many
> > orders slower than real hardware.
> >
> > This is why the best hardware solution is unlikeley to ever be achieved
> > by software folks who have little experience in hardware who would
> > force the FPGA to emulate the PC strengths.
> >
> > John Jakson
> > transputer guy
> >

"JJ" <[email protected]> wrote:

>
>Marc Reinig wrote:
>> The other area where FPGA's win is in latency. Parallel or networked PC's
>> cannot meet the latency requirements of many large computationally intensive
>> real-time systems, even though they can meet the throughput requirements
>>
>> Marco
>
>Ofcourse, whipping up a quick list of platform benefits bound to miss
>some obvious ones.
>
>Also both platforms continually changing/improving but also getting
>harder to understand and less transparent. That alone makes it harder
>to compare since one mans optimized code is another mans starting
>point.
>
>Still the basic decider is likely to be whether one knows how to even
>start a FPGA. project.

For any serious project this shouldn't be an issue. Last year I needed
to warp video in real time to project video on an odd shape. It turned
out a PC was more than fast enough for this task.

--
Reply to nico@nctdevpuntnl (punt=.)
Bedrijven en winkels vindt U op www.adresboekje.nl

"Ben Jones" <[email protected]> wrote in message
news:eif2gv$[email protected]..
>
> FPGA scientific computing OMG ponies 15
>
:-)
....and you had the cheek to post that nonsense thread between me and Bob on
your blog! Keep up the good work!
Cheers mate, Syms.

JJ wrote
>
> Still the basic decider is likely to be whether one knows how to even
> start a FPGA. project.

Yes. Start by looking at scientific apps on GPUs: www.gpgpu.org

On Fri, 03 Nov 2006 16:36:12 +0000, Nico Coesel wrote:

> [email protected] wrote:
>
>>Everything what can parallelize is possibly running faster in Hardware
>>than in Software.

On Fri, 03 Nov 2006 16:36:12 +0000, Nico Coesel wrote:

> Still, the amount of processing power a modern PC processor can deliver is
> enormous. It is problably more cost effective to optimize an algorithm to
> run parallel on 10 PC's than to develop a specific FPGA solution. If space
> is a constraint, the answer is in using blade servers.

Plus, the FPGA can be clocked at say 200 MHz, 10 times less than the the
CPU. Of course that doesn't mean that the CPU will be 10x faster, but the
CPU's 'speed of light' is definitely much larger.

<[email protected]> wrote in message
news:[email protected] oups.com...
> Hi,
> Coming from a scientific computing standpoint (with no hardware
> experience).
> I was wondering if you can improve any dedicated tasks
> by designing special purpose chips a la FPGA to run your code?


Video/audio coprocessing?

DNA pattern searching is a good example.

Already been done though.


Some jobs (e.g. running an OS) are best suited to general purpose
microprocessors, and some (like video display controllers) demand high-speed
yet relatively simple workings and are ideal for dedicated hardware.

If you're looking for a big idea, it is best to look for applications that
need a vast amount of small simple operations?

How about doing OCR front end to correlate 2D-images with many font
characters simultaneously?

"Symon" <[email protected]> wrote in message
news:[email protected]..
> "Ben Jones" <[email protected]> wrote in message
> news:eif2gv$d6p1@c[email protected]..
>>
>> FPGA scientific computing OMG ponies 15
>>
> :-)
> ...and you had the cheek to post that nonsense thread between me and Bob
> on your blog! Keep up the good work!

Some things deserve to be preserved for prosperity. :-D

Surprised you found the blog though, must be more careful with my robots.txt
in future...

Cheers,

-Ben-

Another advantage to FPGAs is that the work can be migrated to an ASIC,
yeilding further speed gains. For serious projects, that can be a real
advantage.

JJ wrote:
> Nico Coesel wrote:
>> [email protected] wrote:
>>
>>> Everything what can parallelize is possibly running faster in Hardware
>>> than in Software.
>> Still, the amount of processing power a modern PC processor can
>> deliver is enormous. It is problably more cost effective to optimize
>> an algorithm to run parallel on 10 PC's than to develop a specific
>> FPGA solution. If space is a constraint, the answer is in using blade
>> servers.
>>
>
> This is often true even for obvious applications of FPGA, PC can still
> be better.
>
> The analysis needs to account for factors such as floating point use,
> high locality referencing or lots of computing on tiny data sets and
> use of standard 32, 64 bit operands favors PCs On the other hand bit
> mangling, very high rates of memory shuffling, use of unusual bit wise
> or odd size mathematical operators and strategic use of buffer memories
> are a win for FPGAs.
>
> I suspect that as Flash drives replace hard drives at the 30GByte level
> the case for hardware taking control of data management in Flash only
> increases since a hardware FPGA design wouldn't usually include disk
> systems and can then push the PC host further away.
>
> Some problems can be factored both ways to get similar results in which
> case PCs still have the familiarity edge. Still a PC based solution
> that is just recompiled to hardware will perform far less well since
> it has already been factored to favor the PC.
>
> For a hardware solution, the software version can be viewed as the
> simulation of same hardware and most simulations usually run many
> orders slower than real hardware.
>
> This is why the best hardware solution is unlikeley to ever be achieved
> by software folks who have little experience in hardware who would
> force the FPGA to emulate the PC strengths.
>
> John Jakson
> transputer guy
>

[email protected] wrote:
> Hi,
> Coming from a scientific computing standpoint (with no hardware
> experience).
> I was wondering if you can improve any dedicated tasks by designing a
> special
> purpose chips ala FPGA to run your code? Does anyone have any
> experience
> with this?
>
<http://www.nallatech.com/?node_id=1.2.1&id=8>

<http://www.sgi.com/products/rasc/>

<http://www.cray.com/products/xd1/index.html>

....

There are plenty of scientific apps out there that are being sped up by
FPGAs. I, for instance am just finishing up a 2D 32-2048 point floating
point FFT engine in Virtex4 that can compute a 2D 2Kx2K FFT on complex
data in under 13 msec including raster order input and output for an
imaging application. It resides in a single Virtex4 SX55 with external
QDR memory.