XC2000

Do not start a new design with XC2000 products. They are of pre-1990 vintage. My rule is: One year in FPGA evolution equals 15 years in human aging. So these parts correspond to a 15 x 15 = 225 year old great-grandfather. Do not disturb him in his grave. Newer chips are so much easier to design with, and Spartan 3 devices are ridiculously cheap. Don't waste any new-design effort on FPGAs that were introduced at any time in the previous century. Some people may have no choice when they need to fix old designs, but nobody should do this voluntarily. Just aim for the nearest trash can... Peter Alfke, Xilinx Applications

Actually, he might be in the same boat as me. Some of us need to maintain hardware and software developed 15-20 years ago. And this means FPGA development as well.

For this, I need to keep an old.... real old... PC around just to maintain my XC3090 stuff.

-- Ed

There is a dichotomy between the extreme reliability and longevity of the electronic hardware, and its rapid obsolescence. The equipment does not die, but it gets harder and harder to maintain, let alone upgrade it. The military suffers most from this, but industrial and medical may not be so far behind. Long ago, even telecom moved slowly, but that has changed dramatically during the past decades.

Certain chips seem to stay around forever, like the 8051 and the Z80 (I was Zilog Applications in 1978/9), but I don't think you can find a '286 or even a '386 today. Xilinx still offers some XC3000A chips (I think), but the software and the computer operating system are a different matter.

My warning was just that nobody should venture into this archaeological trap without a very compelling reason. Peter Alfke, from home.

There is a dichotomy between the extreme reliability and longevity of the electronic hardware, and its rapid obsolescence. The equipment does not die, but it gets harder and harder to maintain, let alone upgrade it. The military suffers most from this, but industrial and medical may not be so far behind. Long ago, even telecom moved slowly, but that has changed dramatically during the past decades.

Certain chips seem to stay around forever, like the 8051 and the Z80 (I was Zilog Applications in 1978/9), but I don't think you can find a '286 or even a '386 today. Xilinx still offers some XC3000A chips (I think), but the software and the computer operating system are a different matter.

My warning was just that nobody should venture into this archaeological trap without a very compelling reason. Peter Alfke, from home.

Yes, I had a reason for using XC2000 device. I'm from Poland and here there is no way to buy any of Xilinx FPGA device, even if I found in some Internet shop in Poland FPGA, it is too expensive. Buying FPGA in shop in foreign countries where FPGA are chip, price of shipping to Poland is also too high. I have two XC2064 and I want use it to build FIFO dedicated driver for SRAMs memory. I know that it could be builded in this device because I have old PC/104 single board computer which has FIFO driver in XC2064.

If there is no way to get old XACT software, I probably buy CPLD from Atmel because its price is very less than Xilinx CPLD (in Poland). But on the other hand the software from Atmel for his CPLD is realy poor (in front of Xilinx Webpack). Meybe I design my device in webpack and then convert it to the atmel device. This will be probably the most economical solution for me.

Please explain what makes Poland so expensive. (I have spent some time there). The Iron Curtain is long gone. PL is part of the EU, postal rates from the US are reasonable. It is not a rich country, but people are not poor. Why would you spend days (weeks?) of chasing obsolete software and an old computer to program your tiny and slow part, when you can avoid all that by using the smallest and cheapst modern FPGA, like a Spartan3 device. You seem to believe that you have a reason... Peter Alfke

"Peter Alfke" schrieb im Newsbeitrag news: snipped-for-privacy@z14g2000cwz.googlegroups.com...

Hi Peter,

you did beat me in response :) I was actually offering to send some Spartan-2 chips that are somewhere on my desk to the original poster.

There are different sales channels, and the overhead they may charge may differ greatly, so I can easily belive that buying qty 1 Xilinx FPGA from Poland is not possible without paying for the shipment more than the silicon does cost.

to the OP, the offer is still valid, please ask in private email if you are interested. I defenetly have some spartan-2 in TQ144 overleft

Antti

S3E development board, how low can it go?

Interesting, - but isn't 28 pins a little light - just 16 FPGA I/O ? perhaps double rows, would keep a similar form factor, backward compatible to DIP28, but allow 0.1" expansion to 56 pins ?

No mention of the size/speed of the dataflash ?

re Lattice "newsflash"on link above - hmmm, reads like classic competitor FUD.

OK, their balance sheets are not stellar, but are fairly typical for an average semi company. Their sales trends indicate to me that their older parts are tailing off faster than the new parts ramp up, but that is temporary. Assets look strong, and they are pumping money into Fujitsu for long term FAB edge.

All up, I'd take such FUD as signs the competitor is seriously worried by what they can see comming ! :)

Altera must be miffed by Lattice, who have trumped their MAX II, with the better MachXO, AND also have Zero Power CPLDs, so have Altera's old technology CPLD market in a pincer movement.

What are Xilinx planning in CPLDs ?

-jg

Hi Jim,

there is actually a pin desnsity upgrade available in the same form factor, but reatining compatibility with single density, so board standard goes to some 96 ios

BTW the competitor gossip was not from Altera ;)

and I personally really think that the MachXO is way better than MAX2 and tha new Lattice low cost/flash FPGAs are nice too. But Lattice has also tons of old too expensive chips that nobody really wants I guess, and at the time they are working towards 90nm and 65nm FPGAs the profit lass could be temporary - interesting times in any case

Antti

Jim, I do not know what Xilinx CPLD are (American:" is") planning, and if I knew I would not tout it in this public forum. But think about it: going up against a small but established, hard-working but money-losing, floundering and increasingly desperate competitor is not the most exciting proposition. Might get anybody a bloody nose...

Peter Alfke

"Peter Alfke" schrieb im Newsbeitrag news: snipped-for-privacy@f14g2000cwb.googlegroups.com...

Xilinx is the only FPGA vendor that does not offer non volatile fpga's. (Altera MAX2 is an small FPGA no matter what their marketing says).

Xilinx has not made any (significant) new PLD announcements for a while -so either Xilinx will eventually completly get out from flash-pld nonvolatile things, or

... something would come ?

If it comes, I hope that Xilinx Press Release will appear AFTER Xilinx webshop has been setup to take orders for the new NV device family (for immediate shipping).

That would be something!

When the V4 spins off nicely, then it could be good idea to make a quick take and grab off the cream of NV-pld business as well, there is exactly enough time to monitor how the altera-lattice small pld-nvfpga things (max2-machXO) are doing. Could be there is not much money to grab on that market of course.

I wonder why ST cancelled their project to introduce ARM with on chip PLD? Possible their stacked die technology would have been to expensive, or there wasnt enough interest (my interest doesnt seem to be enough...)

Antti

Hello Antti,

your board is very similar to our GOP_XC3S200 FPGA board, which has a DIL24 PAL/GAL compatible footprint. And your company is in our neighborhood. Let's have a beer together ?

MIKE

Hey, mstrug, you must be kidding!

For years I buy in Poland various Xilinx chips with prices almost the same as in US. And there is no problem to buy even a single piece. This was the reason why I selected (many years ago) Xilinx: in times where it was hard to buy most semiconductors, Xilinx rep offered good availability and support.

Dz.

there is no way to buy any of Xilinx FPGA device, even if I found in some Internet shop in Poland FPGA, it is too expensive. Buying FPGA in shop in foreign countries where FPGA are chip, price of shipping to Poland is also too high. I have two XC2064 and I want use it to build FIFO dedicated driver for SRAMs memory. I know that it could be builded in this device because I have old PC/104 single board computer which has FIFO driver in XC2064.

Atmel because its price is very less than Xilinx CPLD (in Poland). But on the other hand the software from Atmel for his CPLD is realy poor (in front of Xilinx Webpack). Meybe I design my device in webpack and then convert it to the atmel device. This will be probably the most economical solution for me.

Can you provide me more information on this matter

snipped-for-privacy@operamail.co

snipped-for-privacy@hotmail.co