My Thinkpad T460s uses a Intel SSD 600P Series NVMe M.2 stick.
Since yesterday the stick causes the initialization of the laptop to be
Very slow. To reach the first BIOS string output it takes around 5 minutes. When removing the NVMe stick I can enter BIOS setup as normal.
It feels like the NVMe stick causes the system to jam. Maybe interrupting the CPU or causing DMA on the PCIe bus.
I also tried to boot the NVMe stick inside a Gigabyte BRIX, however while the BIOS screen appears faster entering BIOS menue is not possible because the system hangs again (maybe it continues but very slow)
Is there a possibility to get the NVMe stick working again? What can be the reason for the failure? I cannot update the firmware or examine the NVMe device because a inserted stick jams the system.
Link Copied
Hi KEise,
Greetings from Intel® SSD Support Group.
Please accept our apologies for our belated reply, we needed to investigate about your inquiry; today, we can share the following details with you
In order to provide information not available to the general public, such as datasheets or other proprietary information, it’s mandatory to be under the privileges of a NDA (Non Disclosure Agreement) signed by your company with Intel®, meaning this information can’t be shared by you or your employees outside your organization. Might you have already signed a NDA with Intel®, please let us know.
In regard the possibility of wiring the NAND chips removed from a SSD, to a FPGA; this is not validated, nor supported or encouraged by Intel®.
On a previous post, we have requested from you, the markings of the failed SSD; if you have them, please send them to us.
Looking forward to hearing from you.
Have a nice day.
Santiago A.
Intel® Customer Support Technician
Under Contract to Intel Corporation
Hello @KEise
Leon
Problem is that the system doesnt boot at all when the NVMe is inserted in the M.2 socket. So I cannot boot into Windows with the
M.2 stick inserted. There is a USB3 to NVMe M.2 disk cabinet that I could try, but I guess the SSD Toolbox needs raw PCIe access. I can maybe
buy a PCIe to M.2 adaptercard and try boot on a desktop pc, but unclear weather the system will lock-up also.
There was no special workload I ran, suddenly it failed to boot. I'm running Linux. Can the NVMe's firmware maybe get corrupted? I tested
with another NVMe M.2 stick (Samsumg) on the T460s and it boots normally. The Intel 600p NVMe on the other hand fails on
both T460s and Gigabyte Brix. System gets locked up.
Leon
As explained above (please read) the BIOS not even starts up in the T460s case (or at least vivibly). There is a "_" shown in the screen only. Sometimes, after around 5 minutes a "To interrupt startup press space" or similar message is shown.
I cannot boot into i.e. a Ubuntu USB stick or a Windows 10 installation stick because the bootprocess doesnt come that long.
In the Gigabyte Brix case at least the BIOS logo is shown but then it is the same, system hangs.
Short answer propably is that the NVMe stich is simply broken.
O.K. I didn't understood correctly your post. Yes, probably your SSD failed.
Leon
Question to Intel would then be weather there is a possibility to recover the data from the NAND chips. Maybe there is some recovery mode that can be activated by a resistor or similar... ??
No. Your only hope for recovering your data (which should have been backed up; just saying) is to try using a NVMe-to-USB adapter and see if the SSD can be accessed this way.
Sorry, reality bites.
...S
I do have backup of most but not all. The controller chip on the stick labeled "SMI" gets quite hot when inserted and the laptop is trying to boot. Not shure weather this is normal.
I'm almost tempted to connect an oszilloscope to the PCIe serdes pins to see weather the chip is alive or see weather I can get another 600P stick and exchange the SMI controller with a heatgun :-) .
Have ordered a "ICY BOX IB-1816M-C31" and a "PCIe adapter card". Lets see how far I can get.
I'm tempted to buy a replacement stick where I would try
to swap the Flash controller using a heatgun to restore my
failing stick.
The controller on failing SSD stick is labeled:
"SMI N07T79.00 S LLUT TW 1636 4D."
There are quite some sticks available with model number
SSDPEKKW512G7 and the same silver shiny SMI controller,
however the labling is different most of the time: i.e.
"SMI N03X56.00 S LLUT TW 1627 4D" I guess the labling
is production batch specific (?)
On top of that the the 512 GB model SSDPEKKW512G7 is
quite expensive and my guess is that the controller on a
SSDPEKKW128G7 with 128 GB is the same?
So question to the SSD part of Intel:
Is the Silicon Motion Flash controller of the SSDPEKKW128G7
SSD stick the same as the one used on the SSDPEKKW512G7 ?
I want to use a heatgun to heat the SSD sticks from the bottom,
then use a Kapton take to pull the Flash controller off. Then hope that
the BGA balls supply enough lead still so that I can swap the
controller chips. Cannot reball.
Intel: The NAND chips used on the 600p stick's are labled 29f01t2ancmg2 . Is it possible to get the datasheet of this chips? I can wire them to an FPGA and try readout that way.
Hi KEise,
Thank you for contacting Intel® SSD support group with your request for assistance with your Intel® SSD 600P Series.
We have reviewed all interactions on your community post and read the amazing contributions from the experts and it indeed looks the drive its bricked.
There are no possibilities from Intel®, to recover the data stored on the failed SSD; there might be some services offered on the web; we haven’t tested or confirmed the effectiveness of such services.
Might you want our cooperation to verify if it’s possible to start the process for a warranty claim, we can request our warranty team to confirm unit’s warranty status; therefore, before making any further diagnostics, we will require your cooperation, so please send us the markings of your SSD as detailed in the following link: https://www.intel.com/content/www/us/en/support/articles/000025532/memory-and-storage.html
Please let us know if you might need further assistance.
We will be looking forward to your reply.
Best regards,
Santiago A.
Intel® Customer Support Technician
Under Contract to Intel Corporation
Hi KEise,
Greetings from Intel® SSD Support Group.
Please accept our apologies for our belated reply, we needed to investigate about your inquiry; today, we can share the following details with you
In order to provide information not available to the general public, such as datasheets or other proprietary information, it’s mandatory to be under the privileges of a NDA (Non Disclosure Agreement) signed by your company with Intel®, meaning this information can’t be shared by you or your employees outside your organization. Might you have already signed a NDA with Intel®, please let us know.
In regard the possibility of wiring the NAND chips removed from a SSD, to a FPGA; this is not validated, nor supported or encouraged by Intel®.
On a previous post, we have requested from you, the markings of the failed SSD; if you have them, please send them to us.
Looking forward to hearing from you.
Have a nice day.
Santiago A.
Intel® Customer Support Technician
Under Contract to Intel Corporation
Marking is Intel SSD 600P Series:
SN: BTPY65050Q6A512F
WWN : 55CD2E414D93DD4D
Model: SSDPEKKW512G7
SA: J25627-101
PBA: H87797-101
My company has a NDA with Intel, maybe you can send an offline email to me to discuss.
Having the datasheet will let me decide weather I have a chance to recover. If the protocol
is ONFI I can read it.
Hi KEise,
Greetings from Intel® SSD Support Group.
Thank you for your reply.
You will receive an email from our advanced technical support.
Have a nice day.
Santiago A.
Intel® Customer Support Technician
Under Contract to Intel Corporation
For more complete information about compiler optimizations, see our Optimization Notice.