Site Home   Archive Home   FAQ Home   How to search the Archive   How to Navigate the Archive   
Compare FPGA features and resources   

Threads starting:
1994JulAugSepOctNovDec1994
1995JanFebMarAprMayJunJulAugSepOctNovDec1995
1996JanFebMarAprMayJunJulAugSepOctNovDec1996
1997JanFebMarAprMayJunJulAugSepOctNovDec1997
1998JanFebMarAprMayJunJulAugSepOctNovDec1998
1999JanFebMarAprMayJunJulAugSepOctNovDec1999
2000JanFebMarAprMayJunJulAugSepOctNovDec2000
2001JanFebMarAprMayJunJulAugSepOctNovDec2001
2002JanFebMarAprMayJunJulAugSepOctNovDec2002
2003JanFebMarAprMayJunJulAugSepOctNovDec2003
2004JanFebMarAprMayJunJulAugSepOctNovDec2004
2005JanFebMarAprMayJunJulAugSepOctNovDec2005
2006JanFebMarAprMayJunJulAugSepOctNovDec2006
2007JanFebMarAprMayJunJulAugSepOctNovDec2007
2008JanFebMarAprMayJunJulAugSepOctNovDec2008
2009JanFebMarAprMayJunJulAugSepOctNovDec2009
2010JanFebMarAprMayJunJulAugSepOctNovDec2010
2011JanFebMarAprMayJunJulAugSepOctNovDec2011
2012JanFebMarAprMayJunJulAugSepOctNovDec2012
2013JanFebMarAprMayJunJulAugSepOctNovDec2013
2014JanFebMarAprMayJunJulAugSepOctNovDec2014
2015JanFebMarAprMayJunJulAugSepOctNovDec2015
2016JanFebMarAprMayJunJulAugSepOctNovDec2016
2017JanFebMarAprMayJunJulAugSepOctNovDec2017
2018JanFebMarAprMayJunJulAugSepOctNovDec2018
2019JanFebMarAprMayJunJulAugSepOctNovDec2019
2020JanFebMarAprMay2020

Authors:A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Custom Search

Messages from 102550

Article: 102550
Subject: Re: "disappointing" performance
From: "Peter Alfke" <peter@xilinx.com>
Date: 17 May 2006 09:54:52 -0700
Links: << >>  << T >>  << A >>
Look at other maturing areas:

Commercial aviation has hardly gotten faster since the 747 arrived more
than 30 years ago.
Automobiles are hardly getting faster, except for at the lunatic
fringe.
100 m dash has improved a few measly percent since Jesse Owens in 1936,
70 years ago !
Baseball records improve mainly through chemistry...

But, as Austin wrote, we are still trying.
It is now easier to make circuits smaller, make bigger chips, improve
yield, and thus lower cost, than it is to make circuits faster.

Peter Alfke


Article: 102551
Subject: Re: getting good deals on small qty?
From: "Peter Alfke" <peter@xilinx.com>
Date: 17 May 2006 09:58:39 -0700
Links: << >>  << T >>  << A >>
I am still raising hell inside Xilinx to make it easier for our
customers to buy parts in small volume.
I have not given up, never will...
Peter Alfke


Article: 102552
Subject: Re: Xilinx or Altera...
From: "Slurp" <slip@slop.slap>
Date: Wed, 17 May 2006 18:24:13 +0100
Links: << >>  << T >>  << A >>

"Paul Leventis" <paul.leventis@gmail.com> wrote in message 
news:1147833178.442391.186290@i40g2000cwc.googlegroups.com...
> To clarify, Altera's posted DDR2 SDRAM controller supports operation up
> to 267 Mhz.  There is another core that supports 333 Mhz operation
> available by contacting your local sales rep.
>
> Paul Leventis
> Altera Corp.
>

Thanks for the confirmation Paul, am already using it!

Slurp




Article: 102553
Subject: Re: "disappointing" performance
From: pbdelete@spamnuke.ludd.luthdelete.se.invalid
Date: 17 May 2006 17:30:49 GMT
Links: << >>  << T >>  << A >>
Austin Lesea <austin@xilinx.com> wrote:
>All,

>A recent Intel presentation at an IEEE Workshop admitted that clock 
>frequency has max'd out, and now has to go down (not up) in order to not 
>create heat.

>We have known that for years now.  So has AMD.

>The only choice is "multi."

>Intel proposes a future with more than 200 x86 cores on one die, with a 
>"communications fabric" and many memories.  All on one die.  Small 
>software problem to be solved by the need to have it solved....

>One attendee of the conference (not me!) quipped, "sounds like you are 
>describing a FPGA..."

>Boy did the presenter get mad!  To be ccompared to a lowly FPGA!  He was 
>spitting venom back at the attendee.  "There is no comparison!  FPGAs 
>are fine grained, and this is not!"

>Sounds like if that is the only difference, the FPGA wins.  Again.

Maybe we'll see "Xilinx inside" within 20 years ;)

Maybe machines with fpgas interconnected in a giant "web of interconnects"
will be the feature. And parallell computeing as the only way to harness that
capability.

One could even take processed silicon plates and have them unmap faulty
chips and interconnect the rest to have that functionality in one go.


Article: 102554
Subject: Re: "disappointing" 550Mhz performance of V5 DSP slices
From: Falk Brunner <Falk.Brunner@gmx.de>
Date: Wed, 17 May 2006 19:50:16 +0200
Links: << >>  << T >>  << A >>
MikeShepherd564@btinternet.com schrieb:

> Maybe the speed freaks can form their own newsgroup ("FPGA
> overclockers"?)  Stand by for photos of water-cooled chips, lit with
> blue LEDs.

STRIKE!

You made my day!

(Ok, its evening now, but nevertheless)

Regards
Falk


Article: 102555
Subject: Re: "disappointing" performance
From: Falk Brunner <Falk.Brunner@gmx.de>
Date: Wed, 17 May 2006 19:52:23 +0200
Links: << >>  << T >>  << A >>
Peter Alfke schrieb:

> Commercial aviation has hardly gotten faster since the 747 arrived more
> than 30 years ago.

Same marketing problem in car industry. So the change to slogans like

" . . you wont arrive faster, but much more relaxed."

Not too bad. Maybe its time for the software to catch up.

Regards
Falk

Article: 102556
Subject: Re: SystemACE bootloader for PowerPC on Virtex4 FX
From: "Jon Beniston" <jon@beniston.com>
Date: 17 May 2006 11:07:53 -0700
Links: << >>  << T >>  << A >>
Peter,

> assuming that you are using System ACE CF to configure your Virtex-4 FX

Yep.

> can load the boot code directly into the processor caches

Thanks sounds very interesting. Any pointers or app notes on how to do
this? (Sorry if I'm overlooking the obvious)

Cheers,
Jon


Article: 102557
Subject: Re: IEEE-1394 (aka FireWire) Core
From: =?ISO-8859-1?Q?St=E9phane_Goujet?= <stephane.news@pp.fi.invalid>
Date: Wed, 17 May 2006 18:13:54 GMT
Links: << >>  << T >>  << A >>
Michael Schöberl wrote:

> AFAIR there is only one (very old) chip (TI TSB something) that has
> firewire and can connect to an FPGA ... all the others have PCI or PCIe

  Yes, the TSB12LV01B.

Goodbye,
         Stéphane.

Article: 102558
Subject: Re: disappointing 550Mhz performance of V5 DSP slices
From: "JJ" <johnjakson@gmail.com>
Date: 17 May 2006 11:17:54 -0700
Links: << >>  << T >>  << A >>

airtom@gmail.com wrote:
> Hello,
> Can anyone give i explanation for the disappointing 550Mhz performance
> of V5 DSP slices? Couldn't we hope 1GHz multipliers with 65nm
> technology?
>
> By the way, why are not the multipliers  pipelined to increase the
> performance?. Is there any chance to see pipelined multipliers in
> virtex-6?

Over on toms hardware they have been have a good time overclocking a
very affordable $130 Pentium D 805 (dual core) from 2.66 GHz to
something like 4.1GHz.bypassing the $1K cpus from Intel & AMD.

But the speed gain varied from soso to 2x was achieved at very great
cost in extra heat output, the nos are quite large. Typically the power
starts at 95W but goes out to 200W or so on the cpu for the extra
performance and needed water cooling and they also cranked the voltage
way up for the last 10% stretch.

The sweet spot I think would be to stay near the 3.6GHz limit of air
cooling with the giant Zalman coolers and near the nominal voltage, the
extra prize points just not worth the hassle.

Now I said all that in 3 paras, they took 45 pages.

Still my money is on parallel slower cpus too, but what else would a
Transputer person say.
All this stuff about 200 x86s on a chip doesn't seem so difficult
considering Moores law applied to density over 25yrs. Since almost all
software is sequential today and most of the software to exploit 200
cores will have to be rewritten anyway, the holdover of x86 ISA is
really looking plain silly, just what will compatitibility mean to run
old code on only 1 of 200 cores.

If you design a cpu for massive scaling with support for fine grain
concurrency you are better off with a design that does it well and you
get many more than 200 on the same chip. Now if the array is programmed
in a language that is 50/50 HDL and traditional Cxx, the difference
between Transputer arrays and FPGA arrays is only a matter of
granularity, they both execute concurrent processes.

Personaly I think it will be along time though before Intel rediscovers
how to make lots of cpus cooperate the way it was done 20 yrs ago
already.

John Jakson
transputer guy
FPGAs & Transputers 2 sides of the same coin


Article: 102559
Subject: Re: SystemACE bootloader for PowerPC on Virtex4 FX
From: "Jon Beniston" <jon@beniston.com>
Date: 17 May 2006 11:20:13 -0700
Links: << >>  << T >>  << A >>
Hi Antti,

Do you have a link to this particular implementation? Is it open
source? (I have googled, but it seems there are lots of different
dosfs')

Cheers,
Jon


Article: 102560
Subject: DCM
From: "Fizzy" <fpgalearner@gmail.com>
Date: 17 May 2006 11:33:40 -0700
Links: << >>  << T >>  << A >>
Does any buddy know How DCM work and does it require an input clock
sigal or a output signal from external oscillator can work also. Plus
how do i connect my code to the output of DCM

Thanks


Article: 102561
Subject: Re: IEEE-1394 (aka FireWire) Core
From: "MM" <mbmsv@yahoo.com>
Date: Wed, 17 May 2006 14:43:28 -0400
Links: << >>  << T >>  << A >>

"Michael Schöberl" <MSchoeberl@mailtonne.de> wrote in message
news:446b5445$1@news.fhg.de...
>
> AFAIR there is only one (very old) chip (TI TSB something) that has
> firewire and can connect to an FPGA ... all the others have PCI or PCIe

Not true. There a few 1394a LLCs including the Philips chip. There are
however no 1394b non-OHCI (read PCI) chips out there to the best of my
knowledge.

/Mikhail



Article: 102562
Subject: Re: Hold Time Violations in Virtex4
From: "Jim Wu" <jimwu88NOOOSPAM@yahoo.com>
Date: 17 May 2006 11:49:10 -0700
Links: << >>  << T >>  << A >>
This particular path doesn't seem to be routed (note there is an "e" on
the net delay line). You don't need to be concerned about hold time
errors due to estimated delays.

HTH,
Jim
http://home.comcast.net/~jimwu88/tools/


Article: 102563
Subject: Re: DCM
From: Falk Brunner <Falk.Brunner@gmx.de>
Date: Wed, 17 May 2006 20:51:00 +0200
Links: << >>  << T >>  << A >>
Fizzy schrieb:
> Does any buddy know How DCM work and does it require an input clock
> sigal or a output signal from external oscillator can work also. Plus

RTFM.

> how do i connect my code to the output of DCM

Instanciate a component from Coregenerator?

Regards
Falk

Article: 102564
Subject: Re: disappointing 550Mhz performance of V5 DSP slices
From: Jan Panteltje <pNaonStpealmtje@yahoo.com>
Date: Wed, 17 May 2006 18:52:43 GMT
Links: << >>  << T >>  << A >>
On a sunny day (17 May 2006 11:17:54 -0700) it happened "JJ"
<johnjakson@gmail.com> wrote in
<1147889874.095362.247580@y43g2000cwc.googlegroups.com>:

>Still my money is on parallel slower cpus too, but what else would a
>Transputer person say.

AMD came out with four core today.

Article: 102565
Subject: Re: "disappointing" performance
From: reeuwijk@few.vu.nl (Kees van Reeuwijk)
Date: Wed, 17 May 2006 21:02:07 +0200
Links: << >>  << T >>  << A >>
Austin Lesea <austin@xilinx.com> wrote:

> Intel proposes a future with more than 200 x86 cores on one die, with a
> "communications fabric" and many memories.  All on one die.  Small 
> software problem to be solved by the need to have it solved....
> 
> One attendee of the conference (not me!) quipped, "sounds like you are
> describing a FPGA..."
> 
> Boy did the presenter get mad!  To be ccompared to a lowly FPGA!  He was
> spitting venom back at the attendee.  "There is no comparison!  FPGAs
> are fine grained, and this is not!"
> 
> Sounds like if that is the only difference, the FPGA wins.  Again.

Except that there is no way to compile standard software to an FPGA, or
even to compile freshly created software in anything near a normal
programming langage. I hope you don't expect that the bulk of
C/C++/Java/C# programmers will learn VHDL or Verilog.

Of course programming a 200-core x86 processor in C/C++/Java/C# is a
software engineering nighmare too, but with enough coding discipline
there is at least a slim chance that you can get a team of ordinary
programmers to produce working software for it.

It is very hard to predict what a viable mainstream architecture will
look like in ten years, but unless a lot of work is done to create
better compilers for them, it surely isn't going to be an FPGA. I
wouldn't bet on the 200-core x86 either, but that's because I'm an
optimist.

In an emergency, I would prefer a 10000-core Transputer to either of
them, even if it would mean resurrecting Occam, but I hope someone can
come up with something more imaginative.

Article: 102566
Subject: Re: DCM
From: "John_H" <johnhandwork@mail.com>
Date: Wed, 17 May 2006 19:05:39 GMT
Links: << >>  << T >>  << A >>
"Fizzy" <fpgalearner@gmail.com> wrote in message 
news:1147890820.359958.300680@38g2000cwa.googlegroups.com...
> Does any buddy know How DCM work and does it require an input clock
> sigal or a output signal from external oscillator can work also. Plus
> how do i connect my code to the output of DCM
>
> Thanks

In fewer words than the manual which you should read:
o The Delay Locked Loop is at the center.
o A series of delay elements can provide a delay of an input signal such as 
that from an oscillator up to 20 ns, perhaps more.
o This multi-element delay chain is tapped at the right point so the delayed 
input lines up with where you expect the next input to be at the reference 
point you select.
o The DFS mode (frequency synthesis) taps and muxes this same delay line at 
multiple points to give you the M/D frequency ratio.

The output from the DCM is a signal that often goes through a global clock 
buffer.

Fizzy,
  Are you expecting comp.arch.fpga to be an instructional forum for you?  It 
seems many of your posts are very green to the point that it doesn't look 
like you do basic research on the items you have question about.  If you're 
new to programmable logic, welcome to the world ahead of you.  Please read 
the data sheets, read the user guides, read the textbooks on how to code 
HDL.  It's much more effective when this forum is used to help solve 
problems rather than provide basic instruction which otherwise has multiple 
sources.

Happy coding,
- John_H 



Article: 102567
Subject: Re: getting good deals on small qty?
From: Uwe Bonnes <bon@hertz.ikp.physik.tu-darmstadt.de>
Date: Wed, 17 May 2006 19:08:47 +0000 (UTC)
Links: << >>  << T >>  << A >>
Peter Alfke <peter@xilinx.com> wrote:
> I am still raising hell inside Xilinx to make it easier for our
> customers to buy parts in small volume.
> I have not given up, never will...

Much luck for this ...

-- 
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

Article: 102568
Subject: Re: SystemACE bootloader for PowerPC on Virtex4 FX
From: "Antti Lukats" <antti@openchip.org>
Date: Wed, 17 May 2006 21:10:12 +0200
Links: << >>  << T >>  << A >>

"Jon Beniston" <jon@beniston.com> schrieb im Newsbeitrag 
news:1147890013.855465.165380@j33g2000cwa.googlegroups.com...
> Hi Antti,
>
> Do you have a link to this particular implementation? Is it open
> source? (I have googled, but it seems there are lots of different
> dosfs')
>
> Cheers,
> Jon
>
learn too google :)

its simple

google larwe dosfs
3 link

http://www.zws.com/products/dosfs/index.html

Antti 



Article: 102569
Subject: Re: SystemACE bootloader for PowerPC on Virtex4 FX
From: "Antti Lukats" <antti@openchip.org>
Date: Wed, 17 May 2006 21:12:34 +0200
Links: << >>  << T >>  << A >>

"Jon Beniston" <jon@beniston.com> schrieb im Newsbeitrag 
news:1147889273.909649.117060@i40g2000cwc.googlegroups.com...
> Peter,
>
>> assuming that you are using System ACE CF to configure your Virtex-4 FX
>
> Yep.
>
>> can load the boot code directly into the processor caches
>
> Thanks sounds very interesting. Any pointers or app notes on how to do
> this? (Sorry if I'm overlooking the obvious)
>
> Cheers,
> Jon
>
xil appnotes and ref designs

pretty cool they use the USR ACCESS JTAG command to create virtual JTAG TAP 
master that then connects to the PPC JTAG tap and uses PPC ICE registers to 
load the caches. was pretty cool to see that implementation

look at the ultracontroller II, the thing is all there

antti 



Article: 102570
Subject: Re: Make a signal free for glitches?
From: Eric Smith <eric@brouhaha.com>
Date: 17 May 2006 12:21:49 -0700
Links: << >>  << T >>  << A >>
Peter Alfke wrote:
> In general: Whenever you change one input of a LUT, there is no glitch
> at the output if the two input addresses generate the same output
> level.
> I documented this 15 years ago...

This information doesn't seem to have made it into the datasheets, which is
why I asked that question a few months ago.

Eric

Article: 102571
Subject: Cyclone II PCI & Pin Swapping
From: joey@joescan.com
Date: 17 May 2006 12:25:23 -0700
Links: << >>  << T >>  << A >>
I'm implementing a PCI104+ board which will have the opencores PCI
bridge core.  I've started with the suggested layout from Alteras PCI
Megacore.  But the choice of pins does not lend to a really clean
layout.  If I swap some of the pins around it will clean up quite a
bit.  I'm wondering if I'm asking for timing problems if I do this.
I'm only planning to run the PCI at a leasurely 33Mhz, so I'm guessing
it won't be a problem, as long as I keep it in the same pair of banks
(5 & 6).  Any thoughts or suggestions?

Many thanks

Joey


Article: 102572
Subject: Re: Xilinx Platform Cable USB protocol specifications and/or open-source firmware replacement
From: Eric Smith <eric@brouhaha.com>
Date: 17 May 2006 12:35:23 -0700
Links: << >>  << T >>  << A >>
ghelbig@lycos.com writes:
> And let's not forget that Xilinx owns the USB Vendor ID for the device,
> so one can't re-use it without their permission.

Why?  Xilinx doesn't have a copyright, trademark, patent, or trade
secret on their USB vendor ID.  I don't recall that I've ever signed a
contract with Xilinx (or anyone else) stating that I would not use the
Xilinx USB vendor ID for something else (e.g., a Xilinx-compatible
cable).

Anyhow, you could always ship a product with some other USB vendor ID,
and supply a tool that allowed the user to change the vendor ID to
any numeric value of his or her choice.


Article: 102573
Subject: Re: Xilinx Platform Cable USB protocol specifications and/or open-source firmware replacement
From: Eric Smith <eric@brouhaha.com>
Date: 17 May 2006 12:38:14 -0700
Links: << >>  << T >>  << A >>
Ed McGettigan <ed.mcgettigan@xilinx.com> writes:
> I reread the thread and didn't see this asked.  Why aren't you just
> using our iMPACT software.  Linux is one of the supported OSes after all.

Doesn't work on 64-bit Linux.  Jungo supports 64-bit, but Xilinx only
supplies 32-bit versions of the proprietary binaries that get linked to
the Jungo code.

Please, please, please support 64-bit Linux in 8.2i, or at least in
8.2i SP1.

Thanks!
Eric

Article: 102574
Subject: Re: Virtex4 FX12 dynamic clock divider
From: "Erik Widding" <widding@birger.com>
Date: 17 May 2006 12:54:30 -0700
Links: << >>  << T >>  << A >>
Guru wrote:
> Here are some more details:
> Desired clock increment is about 1 MHz.
> Maximum jitter not specified.
> 2 DCMs free for now.
>
> I think that DCM with dynamic FX ratios cannot produce such increments,

Guru,

A five minute exercise with an excel spreadsheet is all it takes to
come up with the answer to the DCM resolution question.  If starting
with a 66MHz input clock (which from the exercise became clear to me
that one would want to use the top frequency of the range to get best
resolution), you can get very close to your desired resolution.  There
will be steps of 1.06MHz going from M/D of 15/31 to 1/2 and from 1/2 to
16/31, so the worst two steps happen to be adjacent, and at the very
low end of your frequency range.  The next worst step pairs are
0.76MHz, 0.71MHz, and 0.53MHz and 0.57MHz.

The interesting thing to note is that the steps immediately prior and
immediately after each of the worst case pairs are all on the order of
0.10MHz.  So as a practical matter the DCM gets you extremely close.
If your range of frequencies were just slightly less than 2:1 (i.e.
34MHz to 66MHz) you would be able to get a worst case 0.76MHz step
size.

The DCM solution will drastically reduce the jitter.  This may or may
not be significant in your application.  The M/D values will want to be
stored in a look up table (i.e. a blockram).

Just more food for thought.


Regards,
Erik.

---
Erik Widding
President
Birger Engineering, Inc.

 (mail) 100 Boylston St #1070; Boston, MA 02116
(voice) 617.695.9233
  (fax) 617.695.9234
  (web) http://www.birger.com




Site Home   Archive Home   FAQ Home   How to search the Archive   How to Navigate the Archive   
Compare FPGA features and resources   

Threads starting:
1994JulAugSepOctNovDec1994
1995JanFebMarAprMayJunJulAugSepOctNovDec1995
1996JanFebMarAprMayJunJulAugSepOctNovDec1996
1997JanFebMarAprMayJunJulAugSepOctNovDec1997
1998JanFebMarAprMayJunJulAugSepOctNovDec1998
1999JanFebMarAprMayJunJulAugSepOctNovDec1999
2000JanFebMarAprMayJunJulAugSepOctNovDec2000
2001JanFebMarAprMayJunJulAugSepOctNovDec2001
2002JanFebMarAprMayJunJulAugSepOctNovDec2002
2003JanFebMarAprMayJunJulAugSepOctNovDec2003
2004JanFebMarAprMayJunJulAugSepOctNovDec2004
2005JanFebMarAprMayJunJulAugSepOctNovDec2005
2006JanFebMarAprMayJunJulAugSepOctNovDec2006
2007JanFebMarAprMayJunJulAugSepOctNovDec2007
2008JanFebMarAprMayJunJulAugSepOctNovDec2008
2009JanFebMarAprMayJunJulAugSepOctNovDec2009
2010JanFebMarAprMayJunJulAugSepOctNovDec2010
2011JanFebMarAprMayJunJulAugSepOctNovDec2011
2012JanFebMarAprMayJunJulAugSepOctNovDec2012
2013JanFebMarAprMayJunJulAugSepOctNovDec2013
2014JanFebMarAprMayJunJulAugSepOctNovDec2014
2015JanFebMarAprMayJunJulAugSepOctNovDec2015
2016JanFebMarAprMayJunJulAugSepOctNovDec2016
2017JanFebMarAprMayJunJulAugSepOctNovDec2017
2018JanFebMarAprMayJunJulAugSepOctNovDec2018
2019JanFebMarAprMayJunJulAugSepOctNovDec2019
2020JanFebMarAprMay2020

Authors:A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Custom Search