Tualatin PIII-S running at 1.62 V OK? \ VOGONS

Tualatin PIII-S running at 1.62 V OK?

Topic actions

First post, by feipoa

Posted on 2018-10-01, 12:18

feipoa Offline

Rank l33t++

Rank: l33t++
Posts: 10274
Joined: 2011-03-07, 13:54
Location: Canada

I am trying to get a Tyan Thunder HEsl S2567 motherboard working with dual PIII-S 1.4 GHz CPUs. I am using two Powerleap PL-Neo/T FCPGA-to-FCPGA2 adapters to facilitate Tualatin use on the S2567. According to the product sheet provided by Powerleap, these adapters have a fixed VID request of 1.55 V. http://www.powerleap.com/pdf_files/PL-NeoT.pdf I'm not sure why Powerleap wanted to set it for 1.55 V. From what I can discern, the PIII-S runs at 1.45 Vcore, while the Celeron Tualatin runs at 1.5 V.

When I view the PC status page in Tyan's BIOS, it says that the CPU's Vcore on one CPU is at 1.60 V, while the other is at 1.62 V. Has anyone run these chips long term at 1.62 V? What was the outcome? I've only ever run these at 1.45 V.

Do the Korean Tualatin mods have a non-fixed VID setting to allow the CPU to request the voltage from the motherboard's VRM?

As far as I can tell, the Thunder HEsl does not have built-in Tualatin support and I do not plan to test my luck putting the Tualatin's in the MB without an adapter. Tyan later released the HEsl-T for Tualatin chips.

The attachment Tualatin_Powerleap_PL-NeoT.jpg is no longer available

The attachment Tyan_S2567.jpg is no longer available

Plan your life wisely, you'll be dead before you know it.

Reply 1 of 4, by The Serpent Rider

Posted on 2018-10-01, 13:13

The Serpent Rider Offline

Rank l33t++

Rank: l33t++
Posts: 6160
Joined: 2017-03-25, 19:07
Location: The Hole

That's merely 10% voltage increase not including Vdrop. Should be absolutely fine.

I must be some kind of standard: the anonymous gangbanger of the 21st century.

Reply 2 of 4, by PARKE

Posted on 2018-10-01, 15:22

PARKE Offline

Rank Oldbie

Rank: Oldbie
Posts: 1181
Joined: 2015-07-31, 12:46

The maximum rating is 1.75v according to the datasheet - see attached

The attachment 24965705.pdf is no longer available

Reply 3 of 4, by feipoa

Posted on 2018-10-02, 01:25

feipoa Offline

Rank l33t++

Rank: l33t++
Posts: 10274
Joined: 2011-03-07, 13:54
Location: Canada

I did find the spec sheet previously and saw the "Absolute Maximum" Vcore setting of 1.75 V, but it seems like it is awefully high. It wasn't clear if 1.75 V was OK for sustained use. Normally, it seems CPUs have about a 0.1 V, above and below the target voltage for which they can run at. In this case, it is 1.45 V. So I was wondering if anyone has run these at at least 1.6 V for a sustained period of time, meaning years, as a daily driver? I suspect those who ran these in slockets or s370 adapters may have run into this, especially if using the PL-Neo/T.

I'm running Memtest 4.20 on the system now to find the optimal chipset settings. Using 4 GB reg/ecc. There seems to be something wrong with CPU2's temperature sensor because it always shows 92 C, while CPU1 is at 50 C. Perhaps something related to the Neo/T? Do these chips have
built-in thermistors or could the issue be on the MB?

Plan your life wisely, you'll be dead before you know it.

Reply 4 of 4, by The Serpent Rider

Posted on 2018-10-02, 07:04

The Serpent Rider Offline

Rank l33t++

Rank: l33t++
Posts: 6160
Joined: 2017-03-25, 19:07
Location: The Hole

It wasn't clear if 1.75 V was OK for sustained use.

It is normal for 130nm CPU. Barton works at default 1.65v without any problem.

I must be some kind of standard: the anonymous gangbanger of the 21st century.

Go to top of page Go to top of page

Back to General Old Hardware