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 142400

Article: 142400
Subject: Spartan-6 Boards - Your Wish List
From: John Adair <g1@enterpoint.co.uk>
Date: Sun, 9 Aug 2009 05:44:15 -0700 (PDT)
Links: << >>  << T >>  << A >>
Guys and Girls

Your chance to have input to what we bring to market as Spartan-6
products.

We have several release candidates nearly ready now but we can still
add or modify features and I am interested in features you would all
like to have in these boards either based on things we hve done before
in our product range or even new features you would like to see.

So go ahead and let us know what you want.

John Adair
Enterpoint Ltd. - Home of Merrick1. The HPC solution.



Article: 142401
Subject: Re: Spartan-6 Boards - Your Wish List
From: "Antti.Lukats@googlemail.com" <Antti.Lukats@googlemail.com>
Date: Sun, 9 Aug 2009 07:45:20 -0700 (PDT)
Links: << >>  << T >>  << A >>
On Aug 9, 3:44=A0pm, John Adair <g...@enterpoint.co.uk> wrote:
> Guys and Girls
>
> Your chance to have input to what we bring to market as Spartan-6
> products.
>
> We have several release candidates nearly ready now but we can still
> add or modify features and I am interested in features you would all
> like to have in these boards either based on things we hve done before
> in our product range or even new features you would like to see.
>
> So go ahead and let us know what you want.
>
> John Adair
> Enterpoint Ltd. - Home of Merrick1. The HPC solution.

if you could ship BEFORE Xilinx SP-601 availability,
there would be interest. if not.. sure there is still
place for more boards, but at the moment its the
time that counts, when would you ship?
guaranteed time of delivery?

can you answer that today?
[No]

I have ordered from those who have announced first,
only to find out that orders are accepted but fulfilled
way later (after competitors have better products)

Antti












Article: 142402
Subject: Re: Spartan-6 Boards - Your Wish List
From: nico@puntnl.niks (Nico Coesel)
Date: Sun, 09 Aug 2009 15:07:39 GMT
Links: << >>  << T >>  << A >>
John Adair <g1@enterpoint.co.uk> wrote:

>Guys and Girls
>
>Your chance to have input to what we bring to market as Spartan-6
>products.
>
>We have several release candidates nearly ready now but we can still
>add or modify features and I am interested in features you would all
>like to have in these boards either based on things we hve done before
>in our product range or even new features you would like to see.
>
>So go ahead and let us know what you want.

I'd go for as few options as possible and loads of room for additional
circuitry and small pads to attach (enameled) wires to the FPGA. 

Whenever I start using a new microcontroller I make a prototype board
with the least possible features. The less features, the greater the
possibilities. If people want to do high-speed stuff they'll have to
make their own boards anyway.

-- 
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
                     "If it doesn't fit, use a bigger hammer!"
--------------------------------------------------------------

Article: 142403
Subject: Re: Peter Alfke
From: nico@puntnl.niks (Nico Coesel)
Date: Sun, 09 Aug 2009 15:28:28 GMT
Links: << >>  << T >>  << A >>
Peter Alfke <alfke@sbcglobal.net> wrote:

>On Aug 8, 2:26=A0pm, n...@puntnl.niks (Nico Coesel) wrote:
>
>>
>> Well, if Peter is no longer employed by Xilinx he is free to post
>> whatever he wants in this newsgroup :-)
>>
>Nico, I was never ever held back, censored or reprimanded by Xilinx
>management.

Previous posts from Austin indicated Xilinx has (strict) rules for
Xilinx employees regarding posting messages on the internet. Which is
understandable for a company like Xilinx.

>After the end of this month I will no longer be an employee, but still
>a shareholder of Xilinx.
>And I owe this company a lot of gratitude, financial, intellectual,
>and emotional.
>It was "the best years of my life".
>So don't expect me to divulge any "dirty secrets", whatever you might
>imagine them to be.

I wasn't refering to exposing dirty secrets (goes against my
professional ethics too). I just guessed you have more time on your
hands and no management to approve messages. In other words: less
hassle. My posting was intended as 'Take it easy and enlighten us if
you feel like it'.

-- 
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
                     "If it doesn't fit, use a bigger hammer!"
--------------------------------------------------------------

Article: 142404
Subject: Re: Spartan-6 Boards - Your Wish List
From: John Adair <g1@enterpoint.co.uk>
Date: Sun, 9 Aug 2009 08:28:54 -0700 (PDT)
Links: << >>  << T >>  << A >>
Antti

At the moment all I can say is we don't have production amounts of
silicon available. The dates for production silicon arrival are still
flexible and confidential so we can't give any guarantees on dates for
boards as yet. xilinx probably will feed their own asembly first but
I'm sure we won't be far behind in our shipping.

To a dregree our product announcement will be linked to a reason
expectation of when we have silicon. There also questions of whether
customers want silicon without the fully fixed memory controller in
there.

John Adair
Enterpoint Ltd.

On 9 Aug, 15:45, "Antti.Luk...@googlemail.com"
<Antti.Luk...@googlemail.com> wrote:
> On Aug 9, 3:44=A0pm, John Adair <g...@enterpoint.co.uk> wrote:
>
> > Guys and Girls
>
> > Your chance to have input to what we bring to market as Spartan-6
> > products.
>
> > We have several release candidates nearly ready now but we can still
> > add or modify features and I am interested in features you would all
> > like to have in these boards either based on things we hve done before
> > in our product range or even new features you would like to see.
>
> > So go ahead and let us know what you want.
>
> > John Adair
> > Enterpoint Ltd. - Home of Merrick1. The HPC solution.
>
> if you could ship BEFORE Xilinx SP-601 availability,
> there would be interest. if not.. sure there is still
> place for more boards, but at the moment its the
> time that counts, when would you ship?
> guaranteed time of delivery?
>
> can you answer that today?
> [No]
>
> I have ordered from those who have announced first,
> only to find out that orders are accepted but fulfilled
> way later (after competitors have better products)
>
> Antti


Article: 142405
Subject: Re: Spartan-6 Boards - Your Wish List
From: John Adair <g1@enterpoint.co.uk>
Date: Sun, 9 Aug 2009 08:54:21 -0700 (PDT)
Links: << >>  << T >>  << A >>
Nico

It's umlikely we would encourage direct soldering to development
boards principally because we might end up with warranty issues.
However it is very much in our tradition that we make a higher number
of I/Os "free of function" than our compeditors and that is very
unlikely to change. We are adding a little more to boards now than the
early days but that is driven by customer demand and not what product
lines we cover like disti board designs. When we can we do compensate
for the extra bits by design or bigger I/O packages so lot's of I/O
are left free.

If had not seen our boards up close our standard customer headers for
expansion is based on 0.1 inch (2,54mm) pitch headers. These headers
are also spaced on multiples of 0.1 inch.That layout allows stripboard
or simple add modules to be attached to the development board. The
headers work well at low speed and not so bad high speed as well. If
you also want a maximum I/O for customer use do have a look at our
Darnaw1 http://www.enterpoint.co.uk/moelbryn/darnaw1.html or
Craignell1/2 http://www.enterpoint.co.uk/component_replacements/craignell2.=
html,
http://www.enterpoint.co.uk/component_replacements/craignell.html.
Polmaddie1 also gives some of these features as do some other boards.
We do have long term plans for all of these products to extend into
Spartan-6 replacements as and when demand dictates and our engineering
team has time. It is our general policy to be "compatible" as we can
with older products so that any customer designed add-ons can be used
in later designs wherever that is possible. Our headers have evolved
in the last few designs and those seen on Mulldonnoch2
http://www.enterpoint.co.uk/oem_industrial/mulldonnoch2.html are going
to be the norm whenever we do that although some boards may not have
features like the user set regulator and all the ancillary fixed
voltages. The complete length strips of 0V/GND and 3.3V seem on
Mulldonnoch2 are now the minimum to be expected on boards supporting
headers. In general that would be Broaddown, Raggedstone, Drigmorn and
Merrick ranges.

John Adair
Enterpoint Ltd.

On 9 Aug, 16:07, n...@puntnl.niks (Nico Coesel) wrote:
> John Adair <g...@enterpoint.co.uk> wrote:
> >Guys and Girls
>
> >Your chance to have input to what we bring to market as Spartan-6
> >products.
>
> >We have several release candidates nearly ready now but we can still
> >add or modify features and I am interested in features you would all
> >like to have in these boards either based on things we hve done before
> >in our product range or even new features you would like to see.
>
> >So go ahead and let us know what you want.
>
> I'd go for as few options as possible and loads of room for additional
> circuitry and small pads to attach (enameled) wires to the FPGA.
>
> Whenever I start using a new microcontroller I make a prototype board
> with the least possible features. The less features, the greater the
> possibilities. If people want to do high-speed stuff they'll have to
> make their own boards anyway.
>
> --
> Failure does not prove something is impossible, failure simply
> indicates you are not using the right tools...
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0"If it doesn't fit, use a bigg=
er hammer!"
> --------------------------------------------------------------


Article: 142406
Subject: Re: Spartan-6 Boards - Your Wish List
From: Muzaffer Kal <kal@dspia.com>
Date: Sun, 09 Aug 2009 13:11:03 -0700
Links: << >>  << T >>  << A >>
On Sun, 9 Aug 2009 05:44:15 -0700 (PDT), John Adair
<g1@enterpoint.co.uk> wrote:

>Guys and Girls
>
>Your chance to have input to what we bring to market as Spartan-6
>products.
>
>We have several release candidates nearly ready now but we can still
>add or modify features and I am interested in features you would all
>like to have in these boards either based on things we hve done before
>in our product range or even new features you would like to see.
>
>So go ahead and let us know what you want.

I'm looking for a standard high speed extension bus. I used the Avnet
exp to good results but these days FMC seems to be the new standard. A
LPC FMC connector header would be nice.
-
Muzaffer Kal

DSPIA INC.
ASIC/FPGA Design Services

http://www.dspia.com

Article: 142407
Subject: Re: Spartan-6 Boards - Your Wish List
From: glen herrmannsfeldt <gah@ugcs.caltech.edu>
Date: Sun, 9 Aug 2009 23:42:16 +0000 (UTC)
Links: << >>  << T >>  << A >>
John Adair <g1@enterpoint.co.uk> wrote:
 
< Your chance to have input to what we bring to market as 
< Spartan-6 products.
 
< We have several release candidates nearly ready now but we can still
< add or modify features and I am interested in features you would all
< like to have in these boards either based on things we hve done before
< in our product range or even new features you would like to see.

I was thinking not so long ago that it would be nice to
have a board ready to build a clock radio, including the
appropriate external parts for an SDR (software defined
radio, I believe).  The idea was a board for beginning and
intermediate EE/CS projects that could also be mounted in
a nice box for a completed project.  One could use or not use
the LED display and/or radio parts, as usual for such boards.

-- glen

Article: 142408
Subject: Re: AES encryption of bitstream - is my design secure?
From: Theo Markettos <theom+news@chiark.greenend.org.uk>
Date: 10 Aug 2009 02:14:25 +0100 (BST)
Links: << >>  << T >>  << A >>
glen herrmannsfeldt <gah@ugcs.caltech.edu> wrote:
> In most cases, the mass market software sells in large enough
> quantities to overcome a small side usage.  If not, then a new
> business model is needed.  Note that no grocery items are protected
> by AES, and yet they seem to be able to seel them and make a profit.

Coming in a little late at this thread, but anyway...

One way to tackle this issue is general economics.  Don't try to drive your
product into the ground by piling it high with layers of costly DRM (which
may well backfire on you, and may distract from your actual market
intentions).  So build a better product than the competition, or a cheaper
one, or build up enough of a market share by being first that imitations
don't matter (like I suspect Apple doesn't care about $10 Chinese knockoff
MP3 players because they're hardly in the same marketplace).

The next way is to think about security economics.  If it costs someone
$1million to clone your product, is it worth it?  What about $100K, $10K,
$1K?  So make it economically unaffordable to do so.  Now Russian or Chinese
labour is cheap, so this has become more difficult of late.  But, for
example, Datel put in $17million (IIRC) into reverse engineering Sony's
MagicGate chip (the DRM controller Sony's MemoryStick flash format) and
didn't succeed.  The chip was a mass of random gates, with no structure. 
With semi-automated netlist generation tools they got a netlist, but it
didn't work.

That investment, if successful, would probably have been worthwhile given
the worldwide sales of MemorySticks.  But if your field is smaller, your
attacker has to spend less money before it becomes economically infeasible. 
So perhaps simply potting your board in epoxy is enough (but a bit of a pain
from a service point of view).

If you're worried about your board being repurposed for something else, make
their life difficult.  Use wierd connectors, and route all the traces on an
internal PCB layer so they can't easily be tapped and patched (a mistake the
Xbox people made).  If the attacker has to spend $100 on parts and some
hours customising your board almost everyone probably won't bother.  And if
your board costs much more than $100 less than another major application
area, just increase the faff-cost to the customiser until they're level.

If someone is really keen, they'll be desoldering FPGAs or depackaging your
chip, applying lasers or Focused Ion Beams to it, looking at its EM
emissions or many other evil things.  The way to protect against that is
either to make that economically infeasible (so reduce the gain from doing
this) or else apply a higher level control.  I may have an attack where I'm
able to deduce a PIN in 50 guesses.  But if the bank only allows me three,
I'll only be successful about 6% of the time (and to do this frequently
enough to get a decent return I run the risk that I'm intercepted and locked
up).

Or a combination may apply.  So, despite all the security in smartcards,
let's assume I can copy your bank card. I can go to ATMs and withdraw lots
of money.  This is only worthwhile if the cost for me to clone the card was
less than you have in your account.  But the bank will notice this unusual
pattern after a while and block the card.  So even if I clone Bill Gates'
card I can't go on raiding his account forever.  If the bank blocks the card
before I've received back the investment I put in in cracking the card, the
attack isn't worth it.

Theo

-- 
Security research, University of Cambridge Computer Laboratory
http://www.cl.cam.ac.uk/~atm26/

Article: 142409
Subject: Re: iCore7 vs Core2 simulation & FPGA tool performance?
From: Allan Herriman <allanherriman@hotmail.com>
Date: 10 Aug 2009 03:30:57 GMT
Links: << >>  << T >>  << A >>
On Thu, 06 Aug 2009 14:21:34 +0000, Allan Herriman wrote:

> On Tue, 28 Jul 2009 15:22:25 +0000, Allan Herriman wrote:
> 
>> On Sun, 26 Jul 2009 10:28:03 -0700, Muzaffer Kal wrote:
>> 
>>> On 26 Jul 2009 16:23:16 GMT, General Schvantzkoph
>>> <schvantzkoph@yahoo.com> wrote:
>>> 
>>>>Has anyone benchmarked Core7 vs Core2 on NCverilog, Questa, Xilinx and
>>>>Altera FPGA tools?
>>> 
>>> I'd also be very interested in Core7 vs Phenom II performance too
>>> (45nm AMD ie Phenom II 955 etc.)
>> 
>> 
>> We just got a new i7 machine for FPGA builds.  It tested at just over
>> twice as fast as our high-end AMD box that was 2-3 years old.
>> 
>> I will publish the results in this ng within a week or two, assuming I
>> can ever get the licensing for ISE 11.2 running on it.  (Thanks Xilinx,
>> flexlm was a really good move.)
> 
> 
> It's been a week and we still can't get the Xilinx licensing working on
> the new machine, so I can't post any ISE11.2 results.  I assume we will
> be able to get the licensing running soon, because I have told the local
> Xilinx reps that I won't be designing their parts into new products if
> we can't run the software.  My last big design used an Altera FPGA, so
> they know I'm not joking.
> 
> Here are the two machines:
> 
> Old machine:
> AMD Phenom 9750 CPU 2.40GHz
> MSI K9A2-CD-F Motherboard  (AMD 790X Chipset) 8GB Corsair
> TWIN2X2048-8500C5DF Xtreme CL5 DDR2 RAM  (1066MHz) Windows XP
> Professional x64 Edition Version 2003 Service Pack 2
>  
> New machine:
> Intel Core i7 Extreme Edition 975 CPU 3.33GHz Asus Rampage-II-Extreme
> Motherboard  (Intel X58 Chipset) 12GB OCZ Platinum Low-Voltage Triple
> Channel CL7 DDR3 RAM (1600MHz), Windows Server 2003 R2 Standard x64
> Edition Service Pack 2.
> 
> 
> Using ISE8.2 SP3 (yes, 8.2) on an old test design that almost filled a
> medium sized fpga, we recorded a run time of 6 hours and 12 minutes for
> the older AMD machine, and 3 hours and 3 minutes for the new i7 machine.
> That's 2.03 times as fast.
> 
> The runs contained XST, ngdbuild, map, par, etc. In both cases, the 32
> bit version of software was used.  Peak memory usage (during map) was
> just under 3GBytes. Map was run with the -speed option.
> 
> Regards,
> Allan

I should point out that hyperthreading was turned ON on the i7 machine.  
The results might be a little quicker with hyperthreading turned off.  
I'm not worried about the loss of parallelism here; the thing has four 
cores, each capable of hyperthreading, and most of the cores will be idle 
most of the time.

The FPGA build machines here are dedicated to doing just that and are 
headless.  We use different machines for desktop work and simulations, 
etc.

Regards,
Allan

Article: 142410
Subject: Re: Spartan-6 Boards - Your Wish List
From: John Adair <g1@enterpoint.co.uk>
Date: Mon, 10 Aug 2009 00:36:11 -0700 (PDT)
Links: << >>  << T >>  << A >>
We have been looking at the FMC standard but it does have a pile of
problems for us and our potential customers. This specification was
written by a group of people all from companies, in the same market
sectors, and with a very limited vison as a result. Problems as I see
it:

(1) The specification costs a lot to buy and will limit widespread
adoption and potential customer base.
(2) The format is very small even on the bigger version.
(3) A lot of on-module regulation will be needed further limiting what
else can be done on the module.
(4) On small and medium size devices, e.g. most Spartans, it uses up
most I/O that we would normally use for our own headers and we can't
maintain our support for existing customers.
(5) Connectors are very expensive and single source. I will say Samtec
is a very good supplier of connectors but even so if they go out of
business where are the connectors coming from.
(6) The currently available FMC modules are many times the price of a
potential Spartan-6 board especially a starter kit Spartan-6.
(7) There may be IP or licensing issues as a result of the way the
spec is controlled and owned.

That all said the attraction of common standard is a nice idea. I'm
just not sure of the practical and cost aspects of it as FMC currently
appears. There are so many connector standards that could have been
either reused or just used as is without a totally new standard. One
of my personal favorites is the PC104 connectors that we allready use
both in normal mode but also in custom fashions for some of our
products. We also do that just by FPGA configuration and that
selection doesn't affect the hardware aspects of our boards.

I think we will adopt FMC on our high end Spartan-6 boards and maybe
have an adaptor available for our low end. We are looking at the
practicality of that. Our Virtex-6 boards may have this as standard
but we debating that. We will also

John Adair
Enterpoint Ltd.

On 9 Aug, 21:11, Muzaffer Kal <k...@dspia.com> wrote:
> On Sun, 9 Aug 2009 05:44:15 -0700 (PDT), John Adair
>
> <g...@enterpoint.co.uk> wrote:
> >Guys and Girls
>
> >Your chance to have input to what we bring to market as Spartan-6
> >products.
>
> >We have several release candidates nearly ready now but we can still
> >add or modify features and I am interested in features you would all
> >like to have in these boards either based on things we hve done before
> >in our product range or even new features you would like to see.
>
> >So go ahead and let us know what you want.
>
> I'm looking for a standard high speed extension bus. I used the Avnet
> exp to good results but these days FMC seems to be the new standard. A
> LPC FMC connector header would be nice.
> -
> Muzaffer Kal
>
> DSPIA INC.
> ASIC/FPGA Design Services
>
> http://www.dspia.com


Article: 142411
Subject: Re: Spartan-6 Boards - Your Wish List
From: John Adair <g1@enterpoint.co.uk>
Date: Mon, 10 Aug 2009 00:39:41 -0700 (PDT)
Links: << >>  << T >>  << A >>
As it happens one of our first release candidates may be quite good
for this. It will need an add-on module for the front end RF and maybe
our simple R/2R DAC module to drive a speaker but probably very close
to what you want.

John Adair
Enterpoint Ltd.

On 10 Aug, 00:42, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote:
> John Adair <g...@enterpoint.co.uk> wrote:
>
> < Your chance to have input to what we bring to market as
> < Spartan-6 products.
>
> < We have several release candidates nearly ready now but we can still
> < add or modify features and I am interested in features you would all
> < like to have in these boards either based on things we hve done before
> < in our product range or even new features you would like to see.
>
> I was thinking not so long ago that it would be nice to
> have a board ready to build a clock radio, including the
> appropriate external parts for an SDR (software defined
> radio, I believe). =A0The idea was a board for beginning and
> intermediate EE/CS projects that could also be mounted in
> a nice box for a completed project. =A0One could use or not use
> the LED display and/or radio parts, as usual for such boards.
>
> -- glen


Article: 142412
Subject: Re: Spartan-6 Boards - Your Wish List
From: kclo4 <alexis.gabin@gmail.com>
Date: Mon, 10 Aug 2009 00:55:25 -0700 (PDT)
Links: << >>  << T >>  << A >>
On Aug 10, 1:42=A0am, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote:
> John Adair <g...@enterpoint.co.uk> wrote:
>
> < Your chance to have input to what we bring to market as
> < Spartan-6 products.
>
> < We have several release candidates nearly ready now but we can still
> < add or modify features and I am interested in features you would all
> < like to have in these boards either based on things we hve done before
> < in our product range or even new features you would like to see.
>
> I was thinking not so long ago that it would be nice to
> have a board ready to build a clock radio, including the
> appropriate external parts for an SDR (software defined
> radio, I believe). =A0The idea was a board for beginning and
> intermediate EE/CS projects that could also be mounted in
> a nice box for a completed project. =A0One could use or not use
> the LED display and/or radio parts, as usual for such boards.
>
> -- glen

Hi john,

 I was thinking about the same: a nice board that will be able to do
SDR application so mean include at least one fast ADC and DAC
(>100MSPS, 200 would be better in order to play in FM bandwidth
without any analogic stage)
I am currently playing with a Terasic board : DE2-70 and using the
video ADV7123 as a DAC: I have been able to generate FM broadcast with
it,but I am now missing an ADC in order to do a loopback and analyze
what i generate. Like this a board including those component would be
able to do SDR applications, video applications, or things like a
scope, spectrum analyzer...
I have been thinking and a video ADC like AD9483 might be a good
choice as it offer 3 fast channel , the resolution is low but i
believe is enought for lab testing.
But if you add ADC/DAC on your board you should have as much output
channel than input , they should be same resolution and same clock
speed...

Can you tell us about the targeted price for you board is it like
around a few hundred euros/pound, or more a thousand?

Article: 142413
Subject: Re: Spartan-6 Boards - Your Wish List
From: "Nial Stewart" <nial*REMOVE_THIS*@nialstewartdevelopments.co.uk>
Date: Mon, 10 Aug 2009 10:05:28 +0100
Links: << >>  << T >>  << A >>
John,

A header option you might not have considered is Altera's HSMC
interface.

This is becoming a 'standard' on all their new dev boards with an
expanding range of add onn cards available off the shelf.

http://www.altera.com/products/devkits/kit-daughter_boards.jsp

It does take quite a few device pins though, this might rule it
out.


(I have to admit to producing one of the add on boards listed).

Nial



----------------------------------------------------------
Nial Stewart Developments Ltd        Tel: +44 131 516 8883
32/12 Hardengreen Business Park      Fax: +44 131 663 8771
Dalkeith, Midlothian
EH22 3NX
www.nialstewartdevelopments.co.uk






Article: 142414
Subject: Re: Quartus fitter put a user pin on an already assigned pin
From: "Nial Stewart" <nial*REMOVE_THIS*@nialstewartdevelopments.co.uk>
Date: Mon, 10 Aug 2009 10:48:14 +0100
Links: << >>  << T >>  << A >>
> So I don't already understand where the pin LVDS195p/nCEO come from
> and even more why the fitter put it on an assigned pin

What configuration type have you selected, it might be that one of the
required pins is one you've selected as user io?

Also check Assigmnents->Device -> Device and Pin Options, there are options
to allow some config pins to be used as user IO when configuration's
finished. (I've only got an old version of Quartus on this PC so can't
check).


Nial 



Article: 142415
Subject: delta-signa DAC with FPGA
From: Antti <Antti.Lukats@googlemail.com>
Date: Mon, 10 Aug 2009 05:38:26 -0700 (PDT)
Links: << >>  << T >>  << A >>
Hi

it seems that i may have lost all my skills of using google :(
as I havent found any ready to use HDL for for DS DAC
with higher order than 1

only research papers, or then commercial cores, but absolutly
no HDL sources. this cant be true?

or is there some magic reason why all online DS DAC code
is first order only? so far i have failed to figure out what the
reason could be.

Antti

Article: 142416
Subject: Re: Using OPEN in port map
From: KJ <kkjennings@sbcglobal.net>
Date: Mon, 10 Aug 2009 05:48:05 -0700 (PDT)
Links: << >>  << T >>  << A >>
On Aug 6, 1:37=A0pm, JimLewis <J...@SynthWorks.com> wrote:
> Kevin,
> It is here:https://bugzilla.mentor.com/show_bug.cgi?id=3D240
>
> Note it is IR 2132, however it is bugzilla issue 240.
>
> Best,
> Jim

Bug #275 has been submitted as an enhancement to the language to allow
for opens on vector subelements.

Kevin Jennings

Article: 142417
Subject: Re: What would be the best method to terminate GTX_CLK signal in Gigabit Ethernet PHY
From: "MM" <mbmsv@yahoo.com>
Date: Mon, 10 Aug 2009 10:07:14 -0400
Links: << >>  << T >>  << A >>
"Sudhir Singh" <Sudhir.Singh@email.com> wrote
>
> Are you saying
> that regardless of what ever speed the line is operating at
> (10/100/1000 Mbps), the PHY will always have the interface to the MAC
> operating at 1000Mbps, i.e in GMII?

It won't be operating at 1000 Mbps regardless, but the clock direction will 
not change either. The clock frequency however should change to 25 MHz in 
100Mb mode, and to 2.5MHz in 10Mb mode.

>
> The Micrel PHY I am looking at has a boot strap option of starting in
> GMII/MII mode, I couldn't see anything to force a GMII only mode. I am
> wondering whether PHY switches to MII when it finds a 10/100 Mbps
> device at the other end, and switches to GMII if it finds a 1000Mbps
> device.

It really shouldn't. It should stay in the mode found at the boot time. 
Think of the MII mode as just a legacy thing required to interface to an 
obsolete MAC.


/Mikhail




Article: 142418
Subject: Re: Spartan-6 Boards - Your Wish List
From: John Adair <g1@enterpoint.co.uk>
Date: Mon, 10 Aug 2009 08:51:38 -0700 (PDT)
Links: << >>  << T >>  << A >>
The biggest problem with fitting fast ADC/DAC on the board is the
cost. It means that other people that might buy the board don't and
hence you end up with a board with small sales and hence large costs.
To some extend FMC or another module system is an answer allowing a
more generic product for price. The other problem is the choice of ADC/
DAC is very variable andhard fitting a particular ADC/DAC combo can
reduce the sales. For the meantime the modular approach is best
although we do often do customised variants of standard boards for
specif customers in numbers as low as 25 off.

The range of products that we are about to release range for $200 to
$200K so I'm sure there something of a match in there. I think a lower
module will probably do it but the ADC/DAC combo maybe more
significant.

John Adair
Enterpoint Ltd.


On 10 Aug, 08:55, kclo4 <alexis.ga...@gmail.com> wrote:
> On Aug 10, 1:42=A0am, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote:
>
>
>
> > John Adair <g...@enterpoint.co.uk> wrote:
>
> > < Your chance to have input to what we bring to market as
> > < Spartan-6 products.
>
> > < We have several release candidates nearly ready now but we can still
> > < add or modify features and I am interested in features you would all
> > < like to have in these boards either based on things we hve done befor=
e
> > < in our product range or even new features you would like to see.
>
> > I was thinking not so long ago that it would be nice to
> > have a board ready to build a clock radio, including the
> > appropriate external parts for an SDR (software defined
> > radio, I believe). =A0The idea was a board for beginning and
> > intermediate EE/CS projects that could also be mounted in
> > a nice box for a completed project. =A0One could use or not use
> > the LED display and/or radio parts, as usual for such boards.
>
> > -- glen
>
> Hi john,
>
> =A0I was thinking about the same: a nice board that will be able to do
> SDR application so mean include at least one fast ADC and DAC
> (>100MSPS, 200 would be better in order to play in FM bandwidth
> without any analogic stage)
> I am currently playing with a Terasic board : DE2-70 and using the
> video ADV7123 as a DAC: I have been able to generate FM broadcast with
> it,but I am now missing an ADC in order to do a loopback and analyze
> what i generate. Like this a board including those component would be
> able to do SDR applications, video applications, or things like a
> scope, spectrum analyzer...
> I have been thinking and a video ADC like AD9483 might be a good
> choice as it offer 3 fast channel , the resolution is low but i
> believe is enought for lab testing.
> But if you add ADC/DAC on your board you should have as much output
> channel than input , they should be same resolution and same clock
> speed...
>
> Can you tell us about the targeted price for you board is it like
> around a few hundred euros/pound, or more a thousand?


Article: 142419
Subject: Re: Spartan-6 Boards - Your Wish List
From: John Adair <g1@enterpoint.co.uk>
Date: Mon, 10 Aug 2009 08:57:31 -0700 (PDT)
Links: << >>  << T >>  << A >>
On 10 Aug, 10:05, "Nial Stewart"
<nial*REMOVE_TH...@nialstewartdevelopments.co.uk> wrote:
> John,
>
> A header option you might not have considered is Altera's HSMC
> interface.
>
> This is becoming a 'standard' on all their new dev boards with an
> expanding range of add onn cards available off the shelf.
>
Using an Altera stabdard for a Xilinx board might ruffle a few
feathers back in San Jose but again havinf a common standard would be
good. The Altera way of add-on is more sensible than FMC allowing
stanardish connectors to be used and a less rigid physical form too.
The connectors still are not cheap and I don't see a good way to have
several modules covering a requirement. If you see they dip headers we
use now we can in some cases have 5 different, or even the same,
modules fitted and that is selectable totally on a user need.

The comment on pins is the same unfortunately although the Altera
header may make onto our Altera product range at some point.

John Adair
Enterpoint Ltd.

> http://www.altera.com/products/devkits/kit-daughter_boards.jsp
>
> It does take quite a few device pins though, this might rule it
> out.
>
> (I have to admit to producing one of the add on boards listed).
>
> Nial
>
> ----------------------------------------------------------
> Nial Stewart Developments Ltd =A0 =A0 =A0 =A0Tel: +44 131 516 8883
> 32/12 Hardengreen Business Park =A0 =A0 =A0Fax: +44 131 663 8771
> Dalkeith, Midlothian
> EH22 3NXwww.nialstewartdevelopments.co.uk


Article: 142420
Subject: Re: Spartan-6 Boards - Your Wish List
From: glen herrmannsfeldt <gah@ugcs.caltech.edu>
Date: Mon, 10 Aug 2009 18:13:53 +0000 (UTC)
Links: << >>  << T >>  << A >>
John Adair <g1@enterpoint.co.uk> wrote:

< The biggest problem with fitting fast ADC/DAC on the board is the
< cost. It means that other people that might buy the board don't and
< hence you end up with a board with small sales and hence large costs.

The idea was for a board that could be used for undergraduate
CS/EE courses.  For a freshman course, they could just do the
clock part.  A nice six digit LED display would do for that.
(Why are six digit digital clocks so rare now?)  For that,
it would help to keep the price down, but you might be able
to sell large numbers of them.  I would prefer one that didn't
need separate modules, but I do like the LED display on
a separate board so that it can face out when the board is
lying flat.  

As for ADC/DAC, for SDR as I understand it you would need a
fast but not so wide (in bits) ADC, and a much slower DAC.

You could put just the pads on, such that one could solder
the parts in later.  Or maybe a breadboard area, so that
the students had to do a little work to get it running.

< To some extend FMC or another module system is an answer allowing a
< more generic product for price. The other problem is the choice of ADC/
< DAC is very variable andhard fitting a particular ADC/DAC combo can
< reduce the sales. For the meantime the modular approach is best
< although we do often do customised variants of standard boards for
< specif customers in numbers as low as 25 off.

If you design a board that one medium sized school uses for 
an introductory undergrad CS/EE course, that could already
be a significant number of boards.  Even more, if many schools
would use it.

-- glen

Article: 142421
Subject: FPGA-Camp - A mini conference on FPGAs, (Aug'26, Silicon Valley)
From: Vikram <vkr101@gmail.com>
Date: Mon, 10 Aug 2009 11:28:46 -0700 (PDT)
Links: << >>  << T >>  << A >>
Announcing the 1st ever FPGACamp event at Silicon Valley on Aug'26th.
Attendance is FREE.
More details & registration at http://www.fpgacentral.com/fpgacamp

TOPIC: "High Speed Serial Interface: Protocols, IPs & Devices"

The idea behind FPGA Camp is to bring semiconductor engineers together
and discuss FPGA, mainly NextGen FPGA technology, application,
methodology, best practices and challenges. Also provide a location to
meet other local FPGA designers to share their stories.

We are hoping that this would act as a platform to bring all the FPGA
users together more often.

Agenda:

5:30 PM- 6:00 PM: Registration and demo

6:00 PM-  7:00 PM: Tech Talk- "High Speed Serial Interface: Protocols,
IPs & Devices"

7:00 PM- 7:30 PM: Vendor talk- a brief talks from the vendors offering
High Speed  Serial Interface devices or IPs. (5 to 10 Mins each).

7:30 PM- 8:00 PM: Networking and exhibits

Recent expansion in the video usage and growth in the Internet use
have created a demand to move more data faster than ever. To meet this
demand, system & chip designers are moving towards high speed serial
interfaces such as PCIe, XAUI, Interlaken, XFI, 10GbE etc.

With FPGA devices currently supporting speeds upto ~12.5Gb per IO
pair, which makes FPGAs a unique choice for the next design.

This talk will focus on familiarizing people with various protocols
(both currently used & emerging), IPs & Devices which can be used to
solve the next system problem.

The talk will be followed by quick presentations from some of the
vendors offering these solutions (no marketing talk, only technical).

Registration: FREE! Feel free to bring a friend. For Registration/
RSVP visit http://www.fpgacentral.com/fpgacamp or
http://events.linkedin.com/FPGA-Camp-High-Speed-Serial-Interfaces/pub/104860

Thanks,
Vikram

Article: 142422
Subject: Re: iCore7 vs Core2 simulation & FPGA tool performance?
From: General Schvantzkoph <schvantzkoph@yahoo.com>
Date: 10 Aug 2009 19:42:53 GMT
Links: << >>  << T >>  << A >>
On Sun, 26 Jul 2009 16:23:16 +0000, General Schvantzkoph wrote:

> Has anyone benchmarked Core7 vs Core2 on NCverilog, Questa, Xilinx and
> Altera FPGA tools?

I have an iCore7 system up and running. It's a 920 overclocked to 3.6GHz 
using a Thermalright Ultra 120 Extreme. 3.6GHz seems to be the limit for 
stable operation. I also have a Core2 P8400 overclocked to 4GHz, it's 
been running at that speed for a year without problems.

The iCore7 seems to be better for Xilinx tools, worse for NCSim. Here are 
the times (in seconds) for a Xilinx build (11.1 tools, Virtex6) and an 
NCVerilog regression. Both systems are running on 64 bit Fedora 11 with 
8G of RAM.

		iCore7		Core2		iCore7/Core2	Normalized
GHz		3.6		4			
Xilinx		2040		2259		110.74%		123.04%
NCverilog	1660		1358		81.81%		90.90%

NC is incredibly cache sensitive, my guess is that the iCore7's three 
level cache doesn't work very well for NC.

Article: 142423
Subject: Re: Spartan-6 Boards - Your Wish List
From: "Pete Fraser" <pfraser@covad.net>
Date: Mon, 10 Aug 2009 13:03:26 -0700
Links: << >>  << T >>  << A >>
"John Adair" <g1@enterpoint.co.uk> wrote in message 
news:d4baa37f-f871-4143-9711-23ce3d444009@g1g2000vbr.googlegroups.com...

> Your chance to have input to what we bring to market as Spartan-6
> products.

I do a lot of video algorithm development.
DVI (dual-link) I/O would be great.
HDMI and DisplayPort would be a nice addition.

Pete 



Article: 142424
Subject: Re: Spartan-6 Boards - Your Wish List
From: John Adair <g1@enterpoint.co.uk>
Date: Mon, 10 Aug 2009 14:04:43 -0700 (PDT)
Links: << >>  << T >>  << A >>
Yes if you can hit a combination that suits a particular school you
could make it worth doing. That's only every really happened though
with a couple of boards in the market as far as I know. If you look at
the sales of Spartan-3 boards apart from those boards the sales are
relatively low and not in the critical mass area to make that all
work. However if you have particular suggestions I am open to looking
at them and pricing the cost of adding them to a product.

John Adair
Enterpoint Ltd.

On 10 Aug, 19:13, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote:
> John Adair <g...@enterpoint.co.uk> wrote:
>
> < The biggest problem with fitting fast ADC/DAC on the board is the
> < cost. It means that other people that might buy the board don't and
> < hence you end up with a board with small sales and hence large costs.
>
> The idea was for a board that could be used for undergraduate
> CS/EE courses. =A0For a freshman course, they could just do the
> clock part. =A0A nice six digit LED display would do for that.
> (Why are six digit digital clocks so rare now?) =A0For that,
> it would help to keep the price down, but you might be able
> to sell large numbers of them. =A0I would prefer one that didn't
> need separate modules, but I do like the LED display on
> a separate board so that it can face out when the board is
> lying flat. =A0
>
> As for ADC/DAC, for SDR as I understand it you would need a
> fast but not so wide (in bits) ADC, and a much slower DAC.
>
> You could put just the pads on, such that one could solder
> the parts in later. =A0Or maybe a breadboard area, so that
> the students had to do a little work to get it running.
>
> < To some extend FMC or another module system is an answer allowing a
> < more generic product for price. The other problem is the choice of ADC/
> < DAC is very variable andhard fitting a particular ADC/DAC combo can
> < reduce the sales. For the meantime the modular approach is best
> < although we do often do customised variants of standard boards for
> < specif customers in numbers as low as 25 off.
>
> If you design a board that one medium sized school uses for
> an introductory undergrad CS/EE course, that could already
> be a significant number of boards. =A0Even more, if many schools
> would use it.
>
> -- glen




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