Vivado 2014.1工具流程结果不可重复

嗨，预计该工具使用相同的输入和工具选项给出相同的结果。两台机器的操作系统是什么？Windows-32位机器存在已知问题。
有关详细信息，请参阅http://www.xilinx.com/support/answers/61599.html。
谢谢，迪皮卡。
谢谢，迪皮卡.----------------------------------------------
---------------------------------------------- Google之前的问题
张贴。
如果某人的帖子回答了您的问题，请将帖子标记为“接受为解决方案”。
如果你看到一个特别好的和信息丰富的帖子，考虑给它Kudos（左边的明星）



以下为原文

Hi,

It is expected that the tool gives same results with same sets of inputs and tool options.

What is the OS of the two machines?

There is a known issue with windows-32 bit machine.
 
Refer to http://www.xilinx.com/support/answers/61599.html  for details.

Thanks,
Deepika.
Thanks,
Deepika.
--------------------------------------------------------------------------------------------
Google your question before posting. If someone's post answers your question, mark the post as answer with "Accept as solution". If you see a particularly good and informative post, consider giving it Kudos (the star on the left)

你好，谢谢你的回复。
我怀疑这是一个错误的行为，但不能确定导致它的原因。
我尝试了三台不同的64位机器，两台运行的Linux和一台运行的Windows-7。
这三者产生不同的结果。
这些工具完成了这项工作，在日志中不会产生任何错误（仅警告），但.bit文件和时间是不同的。
这三个中使用的.tcl脚本和源代码相同。
我正在编译的项目是用于V7，相当大（在8核3.5GHz CPU上需要2个小时），包括一组vhdl，verilog源和Xilinx IP内核。
任何指针将不胜感激。
谢谢，
亚历克斯



以下为原文

Hi, thanks for the reply. I suspected it's a wrong behavour but can't quite pin point what's causing it. I tried three different 64-bit machines, two runnning Linux and one rinning Windows-7. All three produce different results. The tools finish the job, genenerate no errors (only warnings) in the logs, but .bit files and timings are different. The same .tcl scripts and sources used in all three. The project I'm compiling is for a V7, is quite big (takes 2 hours on 8-core 3.5GHz CPU), comprising a set of vhdl, verilog sources and Xilinx IP cores.
 
Any pointers would be appreciated.
 
Thanks,
Alex
 

嗨，亚历克斯，
您可以比较实现runme.log文件中的校验和，以找出哪个阶段不同。
如果您希望始终获得相同的结果，请参阅http://www.xilinx.com/support/answers/61599.html中的“最大化可重复性”部分
谢谢，
迪皮卡。
谢谢，迪皮卡.----------------------------------------------
---------------------------------------------- Google之前的问题
张贴。
如果某人的帖子回答了您的问题，请将帖子标记为“接受为解决方案”。
如果你看到一个特别好的和信息丰富的帖子，考虑给它Kudos（左边的明星）



以下为原文

Hi Alex,
 
You can compare the checksums in the implementation runme.log files to find out which phase is differing.
 
If you want to achieve same results all the time refer to "maximizing repeatability" section of http://www.xilinx.com/support/answers/61599.html
 
Thanks,
Deepika.
Thanks,
Deepika.
--------------------------------------------------------------------------------------------
Google your question before posting. If someone's post answers your question, mark the post as answer with "Accept as solution". If you see a particularly good and informative post, consider giving it Kudos (the star on the left)

我在运行Centos 6.4的Sun Grid Engine（SGE）上以非项目模式运行Vivado（2014.1,2014.2），我根据从处理器上运行的任务观察不同的时序结果......一般情况下，时序关闭更好
PAR作业本身在处理器上运行，为Vivado提供了完整的8个线程。
如果我在同一台机器上安排2个PAR工作......时间会受到影响。
这只是一个轶事观察..有没有人注意到这一点？
不要忘记通过接受帖子作为解决方案来尽可能地关闭线程。



以下为原文

I run Vivado (2014.1, 2014.2) in non-project mode on Sun Grid Engine (SGE) running Centos 6.4, and I'm observing different timing results depending on the tasks running on the slave processors... in general timing closes better when a PAR job is running on a processor by itself with a full 8 threads available for Vivado.   If I l schedule 2 PAR jobs on the same machine... timing suffers.     This is just an anecdotal observation..   Has anybody else noticed this?
Don't forget to close a thread when possible by accepting a post as a solution.