Intel 13-3220, ECC support or not, Confused

[danb35]

Yes its on page 2 post #29 ecc_check.zip How do I get the actual program/file onto my NAS box so as to run it, do I put it on a USB stick or something?

Download the file to somewhere on your NAS, unzip it (which will yield ecc_check.py), and run, as root, "python ecc_check.py"

 

[Ericloewe]

Just for a couple of datapoints--my Linux server has an i3-3240 with a single 8GB stick of Crucial ECC RAM, and my FreeNAS server has an E3-1230v2 with four sticks of the same. Both have X9SCL-F motherboards. Output on the Linux server:

Code:

[root@e-smith files]# ./ecc_check
5004-5007h: 20 0 63 3
5008-500Bh: 0 0 60 0 

Output on the FreeNAS box:

Code:

[root@freenas2] /mnt/tank/media# python ecc_check.py
5004-5007h: 20 20 66 3
5008-500Bh: 20 20 66 3 

As I understand it, the right-most byte being '3' indicates that ECC is present and enabled. I'm guessing the absence of that in the second line on my Linux server is a result of only one stick of RAM being installed.

Right, one line for each DDR3 channel.

 

[avalon60]

Ok, but how do I actually get the file onto my NAS box? I put it on a USB stick, then plug the said USB stick into a port on the NAS box, then what , how do I get it transferred.

 

[Ericloewe]

Ok, but how do I actually get the file onto my NAS box? I put it on a USB stick, then plug the said USB stick into a port on the NAS box, then what , how do I get it transferred.

Don't do that.

Copy the extracted file into a share, browse to the share with the CLI and run the command.

 

[avalon60]

When in a terminal window on my NAS box, and type /mnt/Backup_Data/.
I get: 'Is a Directory', and then goes back to root@freenas again.

Backup_Data is a share on my NAS box

 

[danb35]

Ok, but how do I actually get the file onto my NAS box? I put it on a USB stick, then plug the said USB stick into a port on the NAS box, then what , how do I get it transferred.

How do you get anything else onto the NAS? You don't put your media files on a USB stick to transfer them to your NAS, do you? I'd assume that you copy them into a shared directory on the NAS. There are lots of ways to do this, but the easiest is probably to pull up some share from your NAS on your PC and either download the file directly to the share, or copy it there from wherever you did download it to.

 

[danb35]

When in a terminal window on my NAS box, and type /mnt/Backup_Data/.
I get: 'Is a Directory', and then goes back to root@freenas again.

That's expected. You've given it a directory name, but haven't told it what to do. If you want to change to that directory (which I assume you do), type 'cd /mnt/Backup_Data' (without the quotes).

 

[avalon60]

Ok thanks for that. I had forgotten my DOS days, lol

This is what I get from running the ecc_check program:
[root@freenas /mnt/Backup_Data]# python ecc_check.py
5004-5007h: 20 0 63 3
5008-500Bh: 20 0 63 3

Does the number 3 at the end mean good news or bad.

Edit, I just found this on Hardforum:
According to Intel's datasheet:
0: ECC disabled
1: ECC is active in I/O; ECC logic is not active In this case.
2: ECC is disabled in I/O, but ECC logic is enabled.
3: ECC active in both I/O and ECC logic

So it does look like my i3 does support ECC memory now.

 

[avalon60]

Also just seen this:

It could be that the board will work with ECC installed and an i3 processor although it does not enable ECC. For these boards (supermicro) its very unclear if the ECC is even enabled on a xeon processor.

So not a definitive yea or nay yet.

 

[Ericloewe]

If the information is accurate on both Xeons and lower-end processors, the 3 means ECC is working normally.

 

[avalon60]

I don't know at all now. I could but the xeon cpu just to be in the same boat a now with my i3.3220 cpu, in that it may or may not work with ECC memory.

 

[Ericloewe]

I don't know at all now. I could but the xeon cpu just to be in the same boat a now with my i3.3220 cpu, in that it may or may not work with ECC memory.

The Xeon works for sure,* that's the safer option. Honestly, I can only say I'm glad I went with Haswell.

*If Intel isn't outright lying (they could've found an Oh $#!@ bug in the Ivy Bridge DDR3 controller that made ECC useless and affected all processors, but decided to only update ark for i3s - this is just conspiracy theory-level speculation, not any kind of fact).

 

[cyberjock]

As I understand it, the right-most byte being '3' indicates that ECC is present and enabled. I'm guessing the absence of that in the second line on my Linux server is a result of only one stick of RAM being installed.

Not exactly.

Ok thanks for that. I had forgotten my DOS days, lol

This is what I get from running the ecc_check program:
[root@freenas /mnt/Backup_Data]# python ecc_check.py
5004-5007h: 20 0 63 3
5008-500Bh: 20 0 63 3

Does the number 3 at the end mean good news or bad.

Edit, I just found this on Hardforum:
According to Intel's datasheet:
0: ECC disabled
1: ECC is active in I/O; ECC logic is not active In this case.
2: ECC is disabled in I/O, but ECC logic is enabled.
3: ECC active in both I/O and ECC logic

So it does look like my i3 does support ECC memory now.

It means that, if you have the proper chipset and that memory register is being used for that function. The problem is that all this info about the proper chipset and memory register stuff has been lost and people are now seeing 3, wanting it to be ECC supported and functional, so making the leap that it is the case. Unfortunately, it's all a big fat grey area and the harsh reality is that we will never really know for 100% certainty that it is or isn't working properly.


For all we know, since i3s are beefier CPUs that are "downgraded", some i3s may actually use ECC and others may not, but there may be no discernible way to prove that. (this is all conjecture, no fact to prove either way).

The only thing you should really bank on is the Xeons with server chipets *do* support ECC. The rest is all a question mark.

 

[avalon60]

Oh well, now at least I have not ordered a cpu that would be only as good as what I already have.