HP Z620: installing a second CPU riser

[xberg]

Hi,

Question: Can I use my CPU riser from my V1 HP Z620 in a V2 HP Z620?

Context:

I currently have a V1 HP Z620 with 2x Sandy Bridge CPUs. 

I want to upgrade to 2x Ivy Bridge CPUs but for this I need a V2 motherboard (part 708614-001).

I found a second hand V2 HP Z620 with a single CPU. I would like to buy that and use my current riser from my V1 HP Z620.

Is it possible?

thanks!

[MtothaJ · Accepted solution]

---

@xberg wrote:

Hi,

Question: Can I use my CPU riser from my V1 HP Z620 in a V2 HP Z620?

Context:

I currently have a V1 HP Z620 with 2x Sandy Bridge CPUs. 

I want to upgrade to 2x Ivy Bridge CPUs but for this I need a V2 motherboard (part 708614-001).

I found a second hand V2 HP Z620 with a single CPU. I would like to buy that and use my current riser from my V1 HP Z620.

Is it possible?

thanks!

---

Yes, you can use your current riser.

There are no hardware differences between the Zx20 'v1' and 'v2' machines whatsoever (unlike in the case of the Zx00) - the only difference is that the 'v1' is missing the microcode for Ivybridge CPU's in the write protected bootblock segment of the bios.

[xberg]

Thanks!

[xberg]

Thanks!

That's partially good news. Even better news would be that HP provides a BIOS update 🙂

[MtothaJ]

---

@xberg wrote:

Thanks!

That's partially good news. Even better news would be that HP provides a BIOS update 🙂

---

The updated bios with the '2013' v2 boot block is not the issue - it is readily available for download on the HP site. If you open up the latest image with hex editor, nearer the end you will see that it has the 2013 bootblock date.

The problem is that perhaps as a design choice (i.e. corupted bootblock would not allow for a boot block recovery procedure to be run) or just simple oversight, the boot block area does not get updated when flashing the bios using the normal HP flashing procedure hence the need for alternative methods.

[xberg]

I see.

And I read your messages: the only method that has proven to work involves soldering.

You do mention an option "switching the debug mode and  Intel FPT" but I have yet to understand what this means or hear that this method was indeed successful for my needs 🙂

[MtothaJ]

---

@xberg wrote:

I see.

And I read your messages: the only method that has proven to work involves soldering.

You do mention an option "switching the debug mode and  Intel FPT" but I have yet to understand what this means or hear that this method was indeed successful for my needs 🙂

---

Debug mode is acheived by shorting pins 1 and 5 of the sound chip (the one with the crab / lobbster logo) with a 1k ohm resistor during boot (just for a moment, no need to hold it in place). This method is contained in official Intel documentation and is sometimes used as a workaround to enable bios flashing on boards where this is not possible otherwise (e.g. locked flash descriptor). It persists until the next reboot. It may take a few times to get it right since the pins are pretty small - it took me AFAIR 4 attempts. All this while booting from a USB drive with DOS and Intel FPT.

When you are DOS you check whether the debug state is in effect by running fptw -d BIOS.bin

This part is perfectly safe - the command will dump your bios contents into the file bios.bin if debug state has been successful. If not you will get Error 26, which means that you need to polish up on the resistor holding part during boot.

If you managed to dump your bios then at this point you in theory have full read / write access to the bios chip.

This is as far as I have got with my experimentation with this method - next step would obviously be attempting to flash the new bios to the chip using FPT's write commands.

As mentioned before, only go down this path if you are comfortable with the thought that you might end up with a brick (then its either a new board or the soldering / desoldering route).