Spartan-3e order of availability?

"Eric Smith" schrieb im Newsbeitrag news: snipped-for-privacy@ruckus.brouhaha.com...

Antti's forecast

1) NO S3E before September 2005 2) NO chipscale BGA before Q1 2006

lets hope I am wrong :)

Antti,

I got ours promised (but not 100% confirmed) for end of June (3S500E, CP132). Let's hope YOU are wrong... ;-)

Thomas (sitting here sweating, maybe it's just the heat in Austria, maybe it's also the fear of not getting the parts in time...)

"Antti Lukats" schrieb im Newsbeitrag news:d9g8if$p07$02$ snipped-for-privacy@news.t-online.com...

Wau!

I like to be wrong on some occasions :)

Antti

"Thomas Entner" schrieb im Newsbeitrag news:42bbafca$0$13522$ snipped-for-privacy@newsreader01.highway.telekom.at...

also

Hi Mr. Smith,

The 3S500E in the CP132 package will be sampling to the general customer base by mid- to end-July 2005.

The reason that the higher density CP132 devices are Spartan-3E instead of Spartan-3 is that Spartan-3E delivers more gates at a lower price in that footprint.

--------------------------------- Steven K. Knapp Applications Manager, Xilinx Inc. General Products Division Spartan-3/-3E FPGAs

--------------------------------- The Spartan(tm)-3 Generation: The World's Lowest-Cost FPGAs.

Hi Steven,

What prices can we expect in comparison to Spartan3? Let's say a XC3S250E compared to a XC3S200 in the same package. Will the CP132 package be priced next to the TQ144? When XC3E250E components will be available?

Thanks,

Luiz Carlos

My answers are inserted below.

For low to medium pin count packages, Spartan-3E is priced at or below a comparably sized Spartan-3 FPGA. Spartan-3 is llower cost for high pin count packages.

For 2006 delivery, the XC3S250E will be lower cost.

The CP132 package is generally about the same price as the TQ144.

The XC3S250E FPGA starts sampling in the September 2005 time frame, after the XC3S100E and XC3S500E FPGAs.

--------------------------------- Steven K. Knapp Applications Manager, Xilinx Inc. General Products Division Spartan-3/-3E FPGAs

--------------------------------- The Spartan(tm)-3 Generation: The World's Lowest-Cost FPGAs.

Hi Steven,

Thankyou for your answers.

I have another doubt: At XILINX web site we have: TQ144 PQ208 FT256 XC3S200-4 $15.10 $20.25 $25.55 XC3S400-4 $22.00 $25.50 $27.00

I know these prices are for small quantities but, why the package has so different impact for the two components?

69.2% from TQ144 to FT256 on XC3S200, and 22.7% from TQ144 to FT256 on XC3S400! Note that both components use all available pins/balls in all mentioned packages.

If we need a lot of pins, it looks like using two XC3S200-TQ144 is better than using one XC3S200-FT256. Some more pins, double the logic, more components = better discount! I know some facts are missing, but I really never understood why this huge price difference.

Can we expect the same behavior for SPARTAN3E (XC3S250E to XC3S500E) ?

Luiz Carlos

Don't fade away before an answer!

If the 200 part die is pad limited then the die size(and therfore cost?) is more closely related to the number of pads(IO) rather than the amount of logic.

Maybe the 400 is not pad limited.

Who knows how Xilinx do their pricing, if its based on value to the customer - maybe IO is more important to the customer than logic at those logic densities.

If you are looking at small production volumes I would choose a single part to minimise development cost. If larger volumes then I would have thought this pricing is not so relevant anyway.

It used to be that packages were simple and cheap. The TQ144 is really just a lead-frame for direct bonding to the die pads, and some plastic on either side. FT256 is a much more complicated package, a multi-layer pc-board at the bottom that routes all die pads from the periphery to their individual locations that are evenly spread over the package area. And the routing is done with attention to the characteristic impedance, to avoid reflections. The result is much better signal integrity, and a smaller packege, i.e. potentially a smaller pc board. You pay for that with a few dollars at these single-quantity prices. Of course far less in production quantities...

When it comes to really large, 1500-ball packages, there are:

10 metal layers on the die, 10 metal layers in the package, and perhaps 10 to 20 layers in your pc board. Maybe the electrons go dizzy with all this complicated routing... Peter Alfke

Peter,

I understand that FT256 is a more complicated package than TQ144. But why does it (the FT256 package) have a very heavier price impact on XC3S400 than on XC3S200? Can I expect the same for Spartan3E?

Andrew,

"Who knows how Xilinx do their pricing," Yes, but I don't feel good when something looks illogical to me!

Luiz Carlos