From this I am curious if reram from foundry are already formed and given? Like r they by default in set mode of few kohms @Mitch Bailey @Tim Edwards

I don’t see how they could be, but I’m not an expert and don’t have any actual experience.

i also tried to reset but device is still in few KM (3Kohms) .is there other customer from shuttle 2309 who had success?

@samarth jain I’m not sure. @Anton Maurovic (efabless support) does a lot of customer support for chipignite. But you may have to just keep asking on the slack channel.

@Mitch Bailey @Tim Edwards i am able to set the devices and reset. so i tried my other chip c2_022 but on /off ratio is 2 only sadly. Also initially pristine device were set ( low resistance) for this. I did three such cycles. I will see how to improve on off ratio further

@samarth jain I see the same behaviour on my chip, and it is for chipignite 2211Q-R. Post is here, but didn't get any response. The average I got on first read for > 1k devices was 2.2kOhm. By any chance what device sizes are you using? My devices' size is 0.32 x 0.32um. Also what voltage are you using to read them? i.e. what's the voltage across them? I kept mine below 0.2V when reading. Also I did manage to reset some to maybe 4/5kOhm, and set down to 1kOhm. I'm limited by my driving circuits as I expected a larger resistance. They do go lower than 1kOhm if you push them, but it depends on what your driving circuit + selector combo can handle.

My read voltage is 0.7 V . So the devices r different from paper published from skywater on these runs.I don’t want to set them too much else my selector will limit current and I cannot reset. My reset voltage WL 2.5 v BL 3 v. And set is Wl 0.7 v and BL 1.5 V . Should u be interested to discuss details we can schedule meeting. I have chips on run 2309 and 2311 @Andrea Mifsud . Also the on off ratio is 2 as if I try to over reset the device will break and set ( a known mechanism) . I also have a chip from ur run 2211Q and I was also limited by driving circuit as it was designed keeping the data published in mind

I wouldn’t be surprised if this was a problem on Skywater’s side. There are a lot of things that you need to be careful with during fabrication to prevent accidental FORMing. READ voltages should typically be kept below 0.3V. 0.7V will probably get a partial SET behavior. May not be an issue now, but good to know for later

I agree with @Akash Levy regarding forming and read voltage . I kept it high because I have still to test my sensing circuits sense margin . Now my major concern is to get around 60Kohms of HRS as currently on off ratio I could get was only 2 😅

If the device is over-SET/over-FORMed, it becomes difficult to achieve proper HRS. I think @Andrea Mifsud said he was able to achieve 10 kOhms on some devices but not much higher

They might not be different, but they certainly behave differently to what is reported on the pdk pages. Happy to discuss further via meeting @samarth jain as it will be quicker for the specifics. I'd be interested to also know your reram layout size. @Akash Levy Yes I have gotten a few in the 5 - 10k range, but it's not reliable atm as I haven't applied specific algorithms on 1 single device at a time. I might try to close the loop later to make decisions based on instantaneous resistor state, but it's not my priority atm. I did notice that some devices might have higher resistance initially, say 80k, and as soon as you apply a high voltage on them they will go through a forming process which brings them down to 10k - 1k range. As you mentioned, it might be they're being accidentally formed somehow as this is far from the expected 1Mohm or more. Answering this will most likely need at least 1 other tapeout to try various things to see if they solve the issue/s or not ...