I'm starting a new project at the moment, and I'm looking at upgrading to ISE 7.1, since I prefer not to change synth/par tool versions mid-project. I noted that a number of people complained about 7.1 when it first came out, but also noted that Service Pack 2 is out now. Can anybody comment on the state of ISE 7.1 at the moment?
Thanks, Jeremy
Howdy Jeremy,
I don't use Linux, but Windoze based 7.1i has stablized enough that chances are slim you'd run into any problems with it - and even if you do, they are likely fixed in SP3 (due out in the next week or so).
Have fun,
Marc
Thanks Marc,
Looks like I'll upgrade after all :)
Jeremy
Unless you are putting RLOCs on the DSP48's. That is still broken in SP2. Last version it worked correctly in is ISE6.3 SP3. ise7.1 SP3 fixes that, but has a problem with the C ports on the DSP48 (the C Port is physically shared by two DSP48 slices, but shows up individually for each slice in the library. If both DSP48s do not have the same value tied to the C Port, its CE input and its reset input, the mapper crashes. If you are careful, that isn't a problem. The problem comes if one DSP48 uses the Cport and one doesn't, you still have to specify the same inputs on both. That creates a packing problem unless you've pre-packed the DSP48s making sure both Cports are wired identically.
-- --Ray Andraka, P.E. President, the Andraka Consulting Group, Inc.
I was just told that they managed to get the CR for the Cport error into SP3, so it sounds like that might work correctly. Time will tell:-)
-- --Ray Andraka, P.E. President, the Andraka Consulting Group, Inc.
For what it's worth..we were told by visiting Xilinx engineer to NOT upgrade our tools to 7.x but instead wait for 8.1 release in August...
Paul C
I'm still using 6.3 SP3 because there are a few show-stoppers in 7.1. Remains to be seen whether they are all fixed in SP3 or not.
-- --Ray Andraka, P.E. President, the Andraka Consulting Group, Inc.
I was told the same thing. We ran into a problem with timing driven map in 7.1 which made our design impossible to implement. 6.3 works well. They have it fixed in 8.1 (it implemented nicely with 8.1) but it won't be fixed in 7.1 sp3. We were told to skip 7.1 unless we need it for S3E or possibly V4.
Jason Daughenbaugh
Looks like I spoke too soon. We also have a design that won't implement in 7.1i (it's very full), but does fine with 6.3i.
So maybe my original response to the OP should have been something closer to "7.1i works fine for all but a few things. If you run into those, you'll need to drop back to 6.3i."
Marc
--
/* jhallen@world.std.com (192.74.137.5) */ /* 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)
Wouldn't fit/route (depending on timing based MAP mode) with 7.1i.
6.3i shows 91% LUTs used in an LX25, and runs through the tools in under an hour while meeting timing (50/50 split between 77 and 155 MHz). I've not noticed any difference between using the V4 global optimization step and not (except that it increase run times by a huge amount) - has anyone else?I have not heard anything, one way or the other, about V2 designs with
7.1i. I will say that if your design is really very full and you are happy with your results in 6.3i, I wouldn't really expect better results with 7.1i. It may (or may not) run through MAP and PAR a bit faster, but I've not see it mapping any more efficiently on two V4 designs.Marc
P.S. It's exceedingly rare to have incorrect logic created - I can't remember the last time I've heard of that.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.