FPGA Intellectual Property
PCI Express*, Networking and Connectivity, Memory Interfaces, DSP IP, and Video IP
Intel Support hours are Monday-Fridays, 8am-5pm PST, except Holidays. Thanks to our community members who provide support during our down time or before we get to your questions. We appreciate you!

Need Forum Guidance? Click here
Search our FPGA Knowledge Articles here.
5952 Discussions

HARD IP DDR3 with UniPHY + ArriaVGX: too long not local_ready

Honored Contributor II



I am long enough work with DDR/DDR2/DDR3 IPs using Altera Device's. But this see first time. 


I use 2 chips together (32 bit width) Hard IP DDR3 600 MHz. And see that during long read write transactions local_ready signal may not ready up to 1.12 us (1400 tacts of 125MHz clock). I specially at first write RAM then read using bursts and do it repeatedly (in loop). And see these holes in ready very often... I have no any timing slacks after compilation. IP was created with megawizard. 


What is it? Refresh? Why is so long?  

Any suggestion?  

Please Help! )
0 Kudos
1 Reply
Honored Contributor II

Problem has been solved. If change parameters to control refresh then local_ready nearly always ready (but refresh must be done every 4 us). In my case I don't do refresh at all because algorithm which works with RAM write and read from the start address to N address without delays (all needed rows are always charged). 


Goodluck to every one!