Hello, please help find an updated firmware version for this (beta/early/china) card version - it currently has 15.00.00.00 fw version.
I’ve purchased it from taobao and it looks exactly as original LSI 9305-16i card (long), but sas3flash -list gives controller SAS3216 instead of SAS3224.
It also says product ID Avago SAS3216 instead of SAS9305-16i
When I try to flash it with Broadcom 16.* firmware for 9305-16i-IT , there is ERROR: NVDATA Image does not match Controller Device ID! Device ID - NVDATA:0xc4 Controller:0xc9
(as it has another controller, not SAS3224) . Looking to another topic, “Is this HBA card legit?” , I’ve tried to flash fw from supermicro site - it flashes without errors, but don’t work at all. Red led on the board is on, and the card becomes very slow to respond, sas3flash -list says it is not operational, flashing is needed. Good that I’ve made a backup of original fw before flashing (15.0.0) - after reflashing it/rebooting, card works again.
So, is there a chance updated (16.*) firmware exist for this card? Is 15.0.0 firmware ok for Truenas?
It sounds like you have a variant of the LSI 9305-16i that’s not a standard SAS3224 controller but instead a SAS3216 (often found on early or China-market versions). That explains why flashing the official 16.x firmware fails—the controller IDs don’t match, and NVDATA is tied to the hardware.
For these cards, 15.00.00.00 firmware is the latest officially compatible version. Trying to force 16.x firmware isn’t recommended; it can brick the card or make it unstable, as you’ve seen.
For TrueNAS, the 15.0.0 firmware is stable and fully usable. Just make sure:
You keep the card in IT mode (for passthrough).
You don’t attempt to flash firmware from a SAS3224 card.
You have a backup of the original firmware before any experiment.
In short, stick with 15.0.0 for your card—it’s safe and supported for TrueNAS. The 16.x branch likely doesn’t exist for the SAS3216 variant.
Absolute legend! I yolo’d the update on my card and discovered it was one of these non-genuine cards. I’ve tried tons of firmware that took (other 3216), but always failed to work, even version 15 of official 9305-16e firmware, which used the 3216, would not work. Thanks for uploading this, it restored functionality to my card; created an account just to say thank you!
With the backup that you uploaded, I was able to compare it to the official firmware - standard documented MPI2 firmware structure. The official P15 release for a 9305-16i and your backup are nearly identical and only differed in NVDATA information. NVDATA drives the Chip ID string, Board name string, PCI subsystem IDs, and the physical port (PHY) mapping tables. The reason the official P15 9305-16e firmware would flash was because its NVDATA chip id matched the board, but then the PHY mapping was wrong (external vs. internal) for how this clone is actually physically wired - so then Linux couldn’t port enable anything.
Your backup firmware confirms that your card and my card are true hybrids - the PHY mapping is identical to the P15 release for the 9305-16i, but the Chip ID and Board name were changed for the SAS3216.
With this knowledge, creating a patch process for P16 releases was not hard for Claude to figure out. P16 changed where PHY mapping was stored, but otherwise, it was only a matter of repeating the same changes to a P16 release for the 9305-16i.
I’ve flashed the new firmware, P16.12, which was the latest that I had previously downloaded and my card is finally working with recent firmware in my test bed. I’m currently exercising some drives attached to it, watching for errors. If things are clean, I’ll post the firmware and the python code for the patch.
I was able to find the firmware, thanks for the work on this. Card is back and except for a couple of cosmetic quirks in the interface it’s running Scale 25.10.3 with no issues so far.
These cards are being sold by many sellers on eBay. Make sure if you purchased a 9305-16i card on eBay you check the chip version. SAS3216 chip is not the same as the SAS3224 chip and requires a different upgrade process. User Nialpo nailed this one with a copy of the original 15.00.00.00 firmware he downloaded from his card. The file is in his Mar 15 post. Using his process I was able to restore my card to working order. (see his original posts from Jan 13).
Everything checked out. I hammered 4 drives for 12-hours (each one connected to an individual SFF) and there were zero complaints registered by mpt3sas. Thanks again to Nialpo for making a backup
Full write-up, python code to reproduce the patched P16.12 firmware, and the firmware that I flashed are on Github. My account is too new / hasn’t posted enough to share links.
Thanks for the heads up!
If I remember correctly the last updates beyond .10 were for SSD-specific issues. Supermicro P16.00.11.00 is probably recent enough for HDD pools.
You have saved my sanity. Thankfully I had made a backup of my weird P15 firmware which I was able to restore (after a few hours failing other attempts) by EFI booting a flash tool and wiping the entire flash, because the Supermicro P16 firmware was accepted to be flashed as an upgrade but totally failed to load with many BIOS boot errors, and locked the kernel of every OS I attempted in endless adapter reset loops. I have the “long card” with notch in the heat sync for the extra ports of the 24i, and the 3216 PHY map has 8-15 disabled. Your AI patched firmware has been working perfectly. I don’t even use TrueNAS but signed up just to say thank you!
