Lattice Semi

I think there's often a problem of poor communication.  Here are some
examples from my own experience (working alone, not as part of a large
company):

-----------------------------------------
Altera:  I'm involved in developing with this vendor's products
(mainly because I came across Altera first), so I have most experience
with them.  I can find all the information I need and the basic
documentation is very good.  The dev kits are excellent, but the
accompanying example code makes you wonder if they know what they're
doing.  I've found the field engineers to be very helpful, but...I
feel repelled from this vendor because of the hassle in installing
tools (registering against a fixed MAC address and so on) and, if I
want to register on their site, I must promise to keep Altera informed
of my career moves for ever and ever (which I refuse to do).  This has
pushed me to look at other suppliers.

-----------------------------------------
Lattice:  Despite the abysmal experience of a visit by the UK/Ireland
Sales manager, I was still keen to learn of competing products, so I
attended a seminar arranged by a distributor.  This was excellent, so
I made enquiries about buying a development kit.  It seemed that the
XP10 demo board would be best and the distributor was offering this at
the "special" price of $555.

Just to make sure I knew what I was getting (since there were two dev
boards with the XP10), I enquired to confirm that this was the
"LatticeXP Advanced Evaluation Board".  The answer came back "The
special offer relates to the "Standard" XP board".  In other words,
the distributor wanted to charge $555 (their "special" price") for the
board which Lattice advertised on their web site for $229.

Bye-bye, Lattice.

A narrow escape there, but at least we didn't waste $555 on a dev kit
which is (as I understand it), an oscillator, an FPGA and a few LEDs.

-----------------------------------------
Xilinx:  Another excellent seminar organised by a distributor and a
great dev board advertised there for an great price ($199).  So I
enquired at the seminar...

I was referred to someone else who was there and given that person's
business card.  The seminar was nearly done and he was busy, so I
decided to e-mail him when I got home.  He telephoned and asked me to
contact their sales office.  Nothing like giving the customer a chance
to get away, eh?

So, I e-mailed the sales office the same day (4th June), quoting the
exact product reference.  They replied by sending me a quotation which
estimated delivery on 4th August and telling us not to pay but to
apply for a credit account, giving our company registration number,
VAT registration number, postal code, telephone number, fax number,
trading names, a trade referee (with telephone number and fax number),
a second trade referee (with telephone number and fax number) - all
this to buy a development kit for about $199.

Am I in a parallel universe?  I reach into my pocket for $199.  I
don't waste my time making credit account applications.

The same product was available in stock at Digi-Key for a few dollars
more, so we ordered it and it came the same week (from a different
continent).

Which would you choose?  Digi-key (21st century) or the distributor
(19th century)?

On 17th July, the distributor e-mailed us (as an attendee at the
seminar) to ask if we wanted to buy any of the Xilinx development
kits.  I think you can guess our reply.

So, maybe the kit is good (I haven't yet had time to do more than
check that it seems to work), but I feel that we don't have a
distributor of any serious value to us, so would it actually be much
use switching it on?

-----------------------------------------
As ever, the chief problem with all the vendors is that, if you want
to know about prices, the response is typically the robotic,
unblinking face, saying "We can talk about prices...tell us about your
application...".

>MK,
>
>Sorry to hear you feel this way.
>
>I would hope that contacting your Xilinx distributor would result in 
>useful and helpful information.
>
>If not, we would certainly like to know.
>
>"Being small" has many disadvantages, but I have to say that many FPGA 
>sockets are created by 'small' businesses.  To that end I would hope 
>that our distributors are doing the job we expect them to do by 
>providing you with the services you require.
>
>Sounds like that did not happen.  Perhaps this is something we should be 
>aware of?
>
>As always, Peter an I stand ready to listen (and act).
>
>Austin

Mike,

I will look into this experience of the demo board.

In general, the field is asked to find out what the customer needs (it 
helps us design better products), so it is unlikely that the field will 
stop asking for information.

If sales or marketing asks me to design in a new feature, they have to 
have conducted a study to convince us that it is worth the millions of 
dollars it takes to do anything nowadays.

However, I agree, that if all you want to do is buy something already 
researched, identified, etc. life is often not easy anymore (from tennis 
shoes to FPGAs).

Thanks for taking the time to let us know how all of the FPGA vendors 
fare through their distributors.  We don't often get that sort of feedback.

Austin

Joseph H Allen wrote:
> In article <A35%i.17632$pr6.16558@newsfe06.phx>,
> Colin Hankins <Colin.Hankins@touit.com> wrote:
> 
>> My particular interest developed in the 
>> Lattice ECP2M because I need a FPGA/SERDES combo and the price for the ECP2M 
>> seems unbelievable compared to their competitor's equivalent FPGAs.
> 
> How about Altera's Aria GX?  The 60K LE device is $262.50 from Arrow, about
> the same as the 70K device from Lattice.  The SERDES in the Lattice device
> looks better (and there are more of them), but Altera's synthesis tools are 
> better.  The free web edition software supports this device.
> 
> I know they support x4 PCIe, if that's why you need the SERDES.

The Altera Arria GX may be a technically good device but the prices from 
the factory were - at least at the time - double what we were quoted for 
the Lattice and Xilinx alternatives.  These were quantity quotes with 
production over a year away.  Double?  Maybe the quote was a fluke since 
prototype pricing for a larger part came in much better than I expected 
(given the recent experience) when our coworkers on another coast got 
quotes from their local sales folks.

- John_H

"austin" <austin@xilinx.com> wrote in message 
news:fhq5o1$1721@cnn.xilinx.com...
>
> Thanks for taking the time to let us know how all of the FPGA vendors fare 
> through their distributors.  We don't often get that sort of feedback.

Austin,

My experience (also as a small company in the UK, like some of the other 
respondents) is that the franchised distributors are all entirely useless 
for every aspect of everything to do with electronic design, whether it's 
technical info, prices or delivery of parts.

They're an infuriating intermediary who do nothing except turn us against 
manufacturers who have been ill-advised enough to abandon all their 
customer-facing activities to distributers.

To the utmost degree possible, we only select components where we can find a 
*guide* volume price using Findchips or on the manufacturer's website, buy 
ex-stock parts from RS, Farnell or Digikey and obtain the technical 
information we need directly from the manufacturer (without filling in 
forms).

Over the lifetime of any product I've worked on, obtaining the actual parts 
to actually build things has been just as big a deal as doing the original 
design.   As a customer, putting oneself at the mercy of the Arrows of the 
world is crazy - and I've never been able to work out why the manufacturers 
do it.

I can't believe that X or A have more than a few hundred shipping products 
at any one time (even with all the speed/package variants).   Why on earth 
can't you sell the wretched things directly from your website - your average 
corner-store handles a large range of products with more sophisticated 
storage and handling requirements.

As far as I can tell, the conventional distributors:

1. Don't hold stock
2. Don't break volume in any useful way, so you can't buy small quantities
3. Can't give proper prices without going back to the mfg
4. Can't offer hard technical advice (i.e. not in the published litterature) 
without going back to the mfg

What are they for?  Why are you still using them?

Will

In article <4741af3c$0$507$bed64819@news.gradwell.net>,
Will Dean <will@nospam.demon.co.uk> wrote:
>"austin" <austin@xilinx.com> wrote in message 
>news:fhq5o1$1721@cnn.xilinx.com...

>> Thanks for taking the time to let us know how all of the FPGA vendors fare 
>> through their distributors.  We don't often get that sort of feedback.

>As far as I can tell, the conventional distributors:

>1. Don't hold stock
>2. Don't break volume in any useful way, so you can't buy small quantities
>3. Can't give proper prices without going back to the mfg
>4. Can't offer hard technical advice (i.e. not in the published litterature) 
>without going back to the mfg

>What are they for?  Why are you still using them?

Here's one thing they do: they sell parts to large companys based on a PO.
In other words they loan money between the time of part delivery and the PO
getting paid.  If you pay them late, I bet the interest rate they charge is
much better than credit-card interest rates, and I bet you can frequently
talk them out of charging interest.

So you can directly buy parts from many semiconductor manufactures using
your credit card, even from A and X (although X redirects you to a
distributer).  This is my complaint: when you pay them by credit card, you
should get the best possible price plus the 2.5% credit card fee.  I mean,
they get paid immediately, so they should be very happy with credit card
orders.

What the manufactures want even more: accurate projections of your future
purchases.  Then they don't have to guess how many chips to make.  Your
contract manufacture should be able to do a much better job of this than the
distributer.
-- 
/*  jhallen@world.std.com AB1GO */                        /* Joseph H. Allen */
int a[1817];main(z,p,q,r){for(p=80;q+p-80;p-=2*a[p])for(z=9;z--;)q=3&(r=time(0)
+r*57)/7,q=q?q-1?q-2?1-p%79?-1:0:p%79-77?1:0:p<1659?79:0:p>158?-79:0,q?!a[p+q*2
]?a[p+=a[p+=q]=q]=q:0:0;for(;q++-1817;)printf(q%79?"%c":"%c\n"," #"[!a[q-1]]);}

Will,

I have forwarded your post onto the people who should care.

I will keep everyone up to speed on what I find out.

Austin

On Mon, 19 Nov 2007 08:51:12 -0800, 
austin <austin@xilinx.com> wrote:

>I have forwarded your post onto the people who should care.

It's encouraging, and perhaps not too surprising, to see
you responding thus.  However, for small customers in 
Europe, the poor performance of distis in supporting 
smaller customers, both technically and in making 
strategic product choices, is very old news.  And it's
not just FPGAs; my bleat applies to the whole component
distribution business.

I don't buy parts from distis any more; I'm not in
manufacturing, and on the rare occasions when I need
to buy silicon I just take the one-off price hit of
going to Digikey or whatever.  But in earlier lives
I was on both sides of the disti - as a small 
development-oriented customer, and as a technical 
FAE in a device manufacturer.  In neither of these
roles did I feel well-served by the distis.  They
often made noises about doing a technical sell and
needing in-house technical expertise, but their
FAEs were always paid peanuts and had to make their
living through sales-target bonuses; so, surprise
surprise, their focus was always on design-in 
numbers.  Anything with a timeline longer than
three months was beyond the next bonus payment 
and therefore irrelevant.  Anything technically
difficult could either be punted upwards to the
manufacturers' FAEs, or (more likely in the case
of small customers) simply ignored because the
customer would not cost them very much if they
merely went away.  There was simply no motivation
for the disti FAEs to get real technical expertise.
They were usually spread way too thinly across
a raft of disparate products from several vendors,
so had no chance to gain real depth of knowledge
in anything.

I had hoped that with the recent consolidation of
distribution in Europe, things might have improved.
Maybe they have, but the feeling I get from this
group is not encouraging.

Maybe things are better in the US.  Perhaps there
are so many design starts there that even a 
modest distribution organisation can retain
useful technical expertise.  I simply don't know.

It doesn't have to be like this.  For example,
I know a few of the distributors for various
EDA tools here in the UK, and I perceive their
technical expertise to be pretty impressive and
their customer support responsive and helpful.
Component distribution, by contrast, seems to 
be stuck in a 1970s road-warrior timewarp
despite the obvious importance to future
sales of effective support of new designs.
They're not even very good at being a stockholding
supermarket, as others have pointed out.  (Sheesh,
they are there to *sell* stuff.  But the typical
experience is to be told that you can't have one
unless you want a truckload and you don't mind
waiting until three weeks after the next festival
of St.Polycarp, or something; and you aren't allowed 
to know what price it will be until you say how
many you want, so you are barred from making your
own intelligent decisions about what parts to use.)

Thanks for letting me get that off my chest :-)
-- 
Jonathan Bromley, Consultant

DOULOS - Developing Design Know-how
VHDL * Verilog * SystemC * e * Perl * Tcl/Tk * Project Services

Doulos Ltd., 22 Market Place, Ringwood, BH24 1AW, UK
jonathan.bromley@MYCOMPANY.com
http://www.MYCOMPANY.com

The contents of this message may contain personal views which 
are not the views of Doulos Ltd., unless specifically stated.

Jonathan,

If one argues that FPGA technology needs to become more user friendly,
and easier to use, in order to make inroads into all possible
applications, your experience is merely one example of an opportunity
(for Xilinx).

Sure, we could say "well, that is life" but who would we be serving?

Or, we could strive to improve the situation and potentially walk away
with stellar growth, and excellent sales figures?

Personally, I will work towards the latter choice.

Austin

On Mon, 19 Nov 2007 10:00:22 -0800, austin <austin@xilinx.com> wrote:

>If one argues that FPGA technology needs to become more user friendly,
>and easier to use, in order to make inroads into all possible
>applications, your experience is merely one example of an opportunity
>(for Xilinx).

Please don't forget that the experiences I was reporting 
(ranting about) are a decade old.  To a large extent I think
it's an opportunity you've already grasped - webshop, 
excellent online documentation, free software downloads 
for smaller customers, good appnotes.... and, dare I say,
informal access to top-notch factory expertise via public
media such as this group.

>Sure, we could say "well, that is life" but who would we be serving?

Nah.  You wouldn't do that.  You need to eat too :-)

Like I said: it seems to me that major FPGA vendors have somewhat
taken that opportunity already.  Almost without exception, they
(or at least their FPGA business) started small, with reasonably
close relations between factory and even the smaller customers.
As they've grown they seem to have remembered that heritage,
and the quality of technical information available through 
FPGA vendors' websites and other resources is pretty good.
(Yes, I know people complain all the time.  And you put a 
brave face on it, because you know that there's at least a
grain of truth in most complaints, and you want happy customers
and happy potential customers.  We appreciate it.  Really.)

I see the result of all this being a sidelining of the traditional
distributor role as intermediary between factory and customer.
Instead, the successful distributors either act as supermarkets
(in which case let us treat them as such, and punish with our
lack of custom those who can't keep their shelves full of 
attractive products at attractive prices) or act as agencies
helping large customers collaborate effectively with the
factory to ensure a good deal for both parties.  The 
traditional component disti, who tries to erect walls between
customer and factory to protect his own revenue monopoly
at the customer's expense, is surely a dinosaur whose 
meteorite is already bright in the sky.  Their passing 
will be mourned by but few.
-- 
Jonathan Bromley, Consultant

DOULOS - Developing Design Know-how
VHDL * Verilog * SystemC * e * Perl * Tcl/Tk * Project Services

Doulos Ltd., 22 Market Place, Ringwood, BH24 1AW, UK
jonathan.bromley@MYCOMPANY.com
http://www.MYCOMPANY.com

The contents of this message may contain personal views which 
are not the views of Doulos Ltd., unless specifically stated.

Wow, how a discussion can evolve from a single reply from Austin to
contact my disti to a discussion about the distribution landscape in
Europe.

At Austin I would like to say: Austin, I've contacted my disti, but
they don't carry Xilinx. I'm sorry, I'm a Lattice user, so if I would
concider Xilinx (or Altera) the disti should contact me, shouldn't
they?

At all they other posters in this discussion: yes, you are right,
distribution has changed, but I wouldn't generalize this statement.
There are still good FAE's in this world. But the supplier should
train them well, and let them do their job, get experienced, find the
best solution (even if this is not in line with the directives from
the supplier to sell the newest part), ...

Concider this statement: distribution FAE's are the sausage between
the hot dog. They are sitting between the supplier and the customer
and are also trying to do what they are payed for.

At the end, I haven't found out the latest and greatest from Xilinx -
and as long as the parts aren't available in production, I will not
concider them for a new design. The ECP2M that I'm using is in
production, I get my parts at a decent price (better than the "one
off" price at Digikey), and I get from time to time annoying questions
from the disti FAE. But that's life I guess. (refering to Austin)

At least, I'm happy that my design is working. Now I can look for
someone who can build it in smaller quantities. But that's another
discussion, isn't it?

Luc

On Mon, 19 Nov 2007 18:24:03 +0000, Jonathan Bromley
<jonathan.bromley@MYCOMPANY.com> wrote:

>On Mon, 19 Nov 2007 10:00:22 -0800, austin <austin@xilinx.com> wrote:
>
>>If one argues that FPGA technology needs to become more user friendly,
>>and easier to use, in order to make inroads into all possible
>>applications, your experience is merely one example of an opportunity
>>(for Xilinx).
>
>Please don't forget that the experiences I was reporting 
>(ranting about) are a decade old.  To a large extent I think
>it's an opportunity you've already grasped - webshop, 
>excellent online documentation, free software downloads 
>for smaller customers, good appnotes.... and, dare I say,
>informal access to top-notch factory expertise via public
>media such as this group.
>
>>Sure, we could say "well, that is life" but who would we be serving?
>
>Nah.  You wouldn't do that.  You need to eat too :-)
>
>Like I said: it seems to me that major FPGA vendors have somewhat
>taken that opportunity already.  Almost without exception, they
>(or at least their FPGA business) started small, with reasonably
>close relations between factory and even the smaller customers.
>As they've grown they seem to have remembered that heritage,
>and the quality of technical information available through 
>FPGA vendors' websites and other resources is pretty good.
>(Yes, I know people complain all the time.  And you put a 
>brave face on it, because you know that there's at least a
>grain of truth in most complaints, and you want happy customers
>and happy potential customers.  We appreciate it.  Really.)
>
>I see the result of all this being a sidelining of the traditional
>distributor role as intermediary between factory and customer.
>Instead, the successful distributors either act as supermarkets
>(in which case let us treat them as such, and punish with our
>lack of custom those who can't keep their shelves full of 
>attractive products at attractive prices) or act as agencies
>helping large customers collaborate effectively with the
>factory to ensure a good deal for both parties.  The 
>traditional component disti, who tries to erect walls between
>customer and factory to protect his own revenue monopoly
>at the customer's expense, is surely a dinosaur whose 
>meteorite is already bright in the sky.  Their passing 
>will be mourned by but few.