Those can be unlocked with a pencil, usually.

kanecvr wrote:

Yeah, you can't change multipliers on most KT400 / KT600 boards. They were considered by manufacturers as "budget chipsets". Not […]
Show full quote

Yeah, you can't change multipliers on most KT400 / KT600 boards. They were considered by manufacturers as "budget chipsets". Not even my KV7 has a multiplier setting in bios - or my NF7-S witch is a higher end nforce board. The only boards that I know of that support multiplier changes are:

- Abit AN7
- Asus A7N8X Deluxe (regular version and early revisions DO NOT support multiplier changes)
- ASUS A7V880 Deluxe
- MSI K7N2 Delta
- MSI K7N2 Delta2 Platinum
- Shuttle AK35GT2

These are all NF2 boards, except for the A7V880 witch uses the VIA KT880 chipset, and the shuttle AK35GT2 witch is a KT333a. The rest of my boards (see list above) do not support multiplier overclocking.

A lot of VIA based boards do not have these 'very useful' multiplier settings in the BIOS thats true.

But there are VIA based boards with this option... Phil has told last page already that he has a KT600 board with it.

My KT400 board also has quite a lot options in the BIOS, and also multiplier & voltage settings:
The Epox EP-8k9A3+ (recapping in progress... the old known problem). Back in the day, i was running a standard XP2100+ (T-Bred B) with the variable multi i want: from 1500+(1.33GHz) to 2200+ (1.8GHz) at 133FSB and from 2600+(2.08GHz) to 2700+(2.16GHz) at 166FSB.

As backup, i've bought last year the Asrock K7VT4A Pro (with VIA KT400A chipset).
http://www.asrock.com/mb/VIA/K7VT4A%20PRO/

In the manual, i've seen there is also a multiplier change possible, but not like your NF2 boards or my Epox board in the BIOS, .... it's the old traditional way 😊 with jumpers:

melbar wrote:

A lot of VIA based boards do not have these 'very useful' multiplier settings in the BIOS thats true. […]
Show full quote

kanecvr wrote:

Yeah, you can't change multipliers on most KT400 / KT600 boards. They were considered by manufacturers as "budget chipsets". Not […]
Show full quote

Yeah, you can't change multipliers on most KT400 / KT600 boards. They were considered by manufacturers as "budget chipsets". Not even my KV7 has a multiplier setting in bios - or my NF7-S witch is a higher end nforce board. The only boards that I know of that support multiplier changes are:

- Abit AN7
- Asus A7N8X Deluxe (regular version and early revisions DO NOT support multiplier changes)
- ASUS A7V880 Deluxe
- MSI K7N2 Delta
- MSI K7N2 Delta2 Platinum
- Shuttle AK35GT2

These are all NF2 boards, except for the A7V880 witch uses the VIA KT880 chipset, and the shuttle AK35GT2 witch is a KT333a. The rest of my boards (see list above) do not support multiplier overclocking.

A lot of VIA based boards do not have these 'very useful' multiplier settings in the BIOS thats true.

But there are VIA based boards with this option... Phil has told last page already that he has a KT600 board with it.

My KT400 board also has quite a lot options in the BIOS, and also multiplier & voltage settings:
The Epox EP-8k9A3+ (recapping in progress... the old known problem). Back in the day, i was running a standard XP2100+ (T-Bred B) with the variable multi i want: from 1500+(1.33GHz) to 2200+ (1.8GHz) at 133FSB and from 2600+(2.08GHz) to 2700+(2.16GHz) at 166FSB.

As backup, i've bought last year the Asrock K7VT4A Pro (with VIA KT400A chipset).
http://www.asrock.com/mb/VIA/K7VT4A%20PRO/

In the manual, i've seen there is also a multiplier change possible, but not like your NF2 boards or my Epox board in the BIOS, .... it's the old traditional way 😊 with jumpers:

With this in mind it seems a bit odd that ECS did not include multiplier control on their Photon AF1 as their Photon series was their "high end".

It seems 3dmark05 needs SSE so Spitfire core Durons and Thunderbird Athlons can't run it. It's mostly relevant for newer faster Socket A CPUs anyhow so it's not much of an issue.

I'm installing FEAR at the moment, it's a slow process using a 900 MHz Spitfire core Duron. 😁

Edit

FEAR turned out to be a great benchmark. I used the medium presets for both CPU and video card.

/Edit

Edit2

This ECS Photon AF1 KT600 motherboard is really odd! It was advertised as an overclocking motherboard back when it was new and it has voltage control in the BIOS, there are 8 settings and the most voltage you can add is 2.7%! 🙄 🤣

The Spritfire Duron 900 would not run at 9x 120 = 1080MHz but 9x115 = 1035MHz seems fully stable, this is at 1.55 x 1.027 = 1.59V (CPU-Z reports 1.568V). With 2.0V this same CPU could probably run at 1200++ MHz, overclocking motherboard...yea right!

/Edit2

Edit3

As I'm done with testing the Duron 900 I looked for more Spitfire Durons, it seems the 700 MHz version was very popular as the rest of my Spitfire core CPUs are all Duron 700 😜. Well this speeds up the testing and the scaling will look very consistent with a perfectly straight line between two points... Perhaps I should investigate what bridges to cross to change the default multiplier to 800 MHz to get a third point...

/Edit3

The list of boards I posted contains the boards I OWN that support multiplier changes via BIOS.... there are lots of other boards that support this. The thing is, on some boards it simply won't work. For example the Gigabyte KT400 I have has bios options for changing the multiplier, but they seem to do nothing.

I found a Duron 800 so I won't have to investigate how to cut and connect bridges to create one! 😁

Here is a picture showing most of my Thunderbirds and Spitfires, two Morgans has infiltrated the picture. One more Thunderbird and one more Spitfire I know about because they sit on motherboards but I probably have even more as my computer parts are not that well organized... Edit I found some more CPUs! Make that 3 more Thunderbirds and one more Spitfire besides these on the picture... /Edit

do someone know if the MSI K7N2 is any good?

I have one for an uncoming build, it uses the FSB 333 nForce 2 without onboard video

Carlos S. M. wrote:

do someone know if the MSI K7N2 is any good?

I have one for an uncoming build, it uses the FSB 333 nForce 2 without onboard video

From what I have read it should be decent.

After 4 CPUs I took a short break in the Socket A CPU scaling benchmarking to test my new Athlon XP-M 2400+. This was just a quick test to see that everything works as I hoped.

The CPU boots at the 6x multiplier and with 1.525V (1.48V - 1.49V sensor reading) as the XP-Ms 1.35V VID is the same as the desktop CPUs 1.525V VID. With the help of magic* the multiplier can be changed to any multiplier up to the max for that CPU model. Cutting all L6 bridges should get you 24x as max "PowerNow!" multiplier as these bridges control the maximum allowed. For motherboards capable of 200+ MHz FSB the original max multiplier of 13.5x (Barton XP-M 2400+) is enough so cutting bridges is not needed.

*Changing multiplier on the fly does not seem to be supported with nForce2 motherboards but they often have multiplier control in the BIOS. The program that does the magic http://www.cpuheat.wz.cz/html/Download.htm

L11 sets the VID, or max VID in the case of XP-M. By closing the third L11 bridge I can change the default XP-M max VID from 1.35V to 1.55V which in turn translates to a desktop Athlon XP VID of 1.625V. If I need more voltage I instead close the second bidge to get an XP-M VID of 1.75V which translates to 1.725V*. If I close both these briges I get 1.95V = 1.825V and if I also close the fifth and last bridge I get 2.0V = 1.85V. It's possible to get all voltages in the voltage table by cutting bridges closed by default in combination with closing bridges which are cut.

*Desktop Athlon XP CPUs and XP-M CPUs VID tables have different resolution in different voltage ranges, this is why the XP-M VIDs are lower in the low VID range compared to the desktop version while increasingly higher in the high VID range. A desktop motherboard should always interpret the VID as if the CPU was a desktop CPU even if it's correctly identified as an Athlon XP-M, or that is at least what I have read.

All this have to wait though. I need to get back to benching Thunderbirds.

Here is at least a screenshot showing the XP-M 2400+ @3200+ (11x200) with ~1.5V running the 7-Zip benchmark. The reported CPU temp is 40C with the CPU fan turned down to almost totally silent.

Who needs mobile athlons 😜 :

At 500 MHz my 3200+ barton is slower then a k6-3 in integer operations, but much faster in FPU intensive apps and games. This is as low as I can get it. The chip tops out at 2700MHz / 1.785v, but my cooler isn't really up to the task (thermaltake volcano)

kanecvr wrote:

Who needs mobile athlons 😜 :

I heard mobile Athlons might be better binned for overclocking and underclocking/undervolting as well compared to desktop Athlons besides mobile Athlons has already unlocked multi out of the box

Carlos S. M. wrote:

kanecvr wrote:

Who needs mobile athlons 😜 :

I heard mobile Athlons might be better binned for overclocking and underclocking/undervolting as well compared to desktop Athlons besides mobile Athlons has already unlocked multi out of the box

My statement was in jest 😜

My 3200+ is unlocked. Didn't do a thing to it. And yea, mobile chips are better binned so they make less head and use less voltage. They still have trouble clocking over 2.7 GHz....

Binning and vcore aside, quite a few barton CPUs happen to be unlocked. Personally I've never seen a 200MHz FSB 3200+ with a locked multiplier in my area, but I did buy one off ebay that is locked. The locked one has a green package, while my unlocked ones are brown.

I show one of my Athlon XP builds, is based on an Athlon XP 3000+ FSB 200 model and mainly parts from trash picked PCs (around 95% of the PC)

Specs are:
Athlon XP 3000+ FSB 200
1 GB DDR 400 CL 2.5
MSI KT6V-LSR (MS-7021)
ASUS Geforce FX 5500 HP OEM (took from a dead HP, temporay GPU)
Samsung Spinpoint 80 GB IDE
DVD + CD driver
Cooler Master Jet 7 cooler
300 watt PSU
Windows XP SP3

kanecvr wrote:

Yeah, you can't change multipliers on most KT400 / KT600 boards. They were considered by manufacturers as "budget chipsets". Not […]
Show full quote

Yeah, you can't change multipliers on most KT400 / KT600 boards. They were considered by manufacturers as "budget chipsets". Not even my KV7 has a multiplier setting in bios - or my NF7-S witch is a higher end nforce board. The only boards that I know of that support multiplier changes are:

- Abit AN7
- Asus A7N8X Deluxe (regular version and early revisions DO NOT support multiplier changes)
- ASUS A7V880 Deluxe
- MSI K7N2 Delta
- MSI K7N2 Delta2 Platinum
- Shuttle AK35GT2

These are all NF2 boards, except for the A7V880 witch uses the VIA KT880 chipset, and the shuttle AK35GT2 witch is a KT333a. The rest of my boards (see list above) do not support multiplier overclocking.

My KT600 based MSI KT6V-LSR (MS-7021) supports multiplier changing, there are a CPU ratio option on the BIOS which lets me set it

kanecvr wrote:

My statement was in jest :P […]
Show full quote

Carlos S. M. wrote:

kanecvr wrote:

Who needs mobile athlons 😜 :

I heard mobile Athlons might be better binned for overclocking and underclocking/undervolting as well compared to desktop Athlons besides mobile Athlons has already unlocked multi out of the box

My statement was in jest 😜

My 3200+ is unlocked. Didn't do a thing to it. And yea, mobile chips are better binned so they make less head and use less voltage. They still have trouble clocking over 2.7 GHz....

Binning and vcore aside, quite a few barton CPUs happen to be unlocked. Personally I've never seen a 200MHz FSB 3200+ with a locked multiplier in my area, but I did buy one off ebay that is locked. The locked one has a green package, while my unlocked ones are brown.

You mean that there are actually desktop Barton 3200+ chips which are not locked?
Iirc mine had a brown package and it's quite possible it's an earlier sample. Maybe not good for overclocking but if I can change the multi's, it might make the CPU a little bit more fun 😀

Tetrium wrote:

You mean that there are actually desktop Barton 3200+ chips which are not locked?
Iirc mine had a brown package and it's quite possible it's an earlier sample. Maybe not good for overclocking but if I can change the multi's, it might make the CPU a little bit more fun 😀

IIRC: Thoroughbred's and Barton's were unlocked until week 39 or week 40 of 2003.

K7 CPU's were favourite here too! My first 'up to date' system was a Duron 1200 'Morgan' with KT266A chipset. After that I upgraded to an Athlon XP 1700+ 'Thoroughbred' (0308MPMW) on an EPoX 8RDA nForce2 motherboard. Eventually I overclocked this system to 2400MHz (and with a cheap upgrade to an 1800+ (0311VPMW) to 2533MHz). Due the overclock I had a high-end system for years 😀. I loved the nForce2 system with low latencies and actually the upgrade to a Pentium D ES was a step back. It was not until the Core 2 Duo that I rediscovered the snappyness that my Athlon XP had.

For legacy reasons I like K7 too. Slot A and Irongate marked a new era and both Thunderbird and Palomino are indestructible chips in terms of overclocking/overvolting and what heat they can take. Just be careful when swapping heatsinks 😀. And overclocking with pencils... Duron 600 at 950? Great!

I got an Athlon XP-M 2400+ from a dead HP Pavilion ze4500

exact model is AXHM2400FQQ4

is 45 watt TDP, 1.45 v and Barton core

Carlos S. M. wrote:

I got an Athlon XP-M 2400+ from a dead HP Pavilion ze4500

exact model is AXHM2400FQQ4

is 45 watt TDP, 1.45 v and Barton core

Nice! 😀 That CPU should have a very good chance to do 2400 MHz at 1.75V.

I'm still benching Thunderbirds but the progress is slow because of work.

I'm missing multiplier 9.5x and 11x and my Athlon 850 will not do 1133MHz at stock voltage surprise surprise 🙄 so no 8.5x at 133 MHz FSB. I just discoverd that although I just fixed the pins on my Athlon 1000C (133) good enough to at least get it into the socket it still wont work and it's my only 7.5x multi Thunderbird so it makes it impossible to bench the 750 MHz version aswell.

I'm thinking of changing motherboard but my Abit KV7 VIA KT600 (has the needed multiplier control) has a blown cap and I do not think I have a suitable one, I need to order some sooner rather than later.

I have at least benched some CPUs. 😀

Spitfire Duron 700, 800, 900. (The only speed grades I own, it turned out that half of my Spitfire CPUs were the 900 MHz version.)

Athlon Thunderbird 700. 800, 850, 900, 1000(100), 1333(133) (same multiplier as 1000)

I still need to bench my Athlon 1200(100), 1200(133), 1300(100) and 1400(133) CPUs.

I looked on the net to see if I could find any cheap Athlon 950*, Athlon 1000(133) Athlon 1100 and Athlon 1133 CPUs but they were at least 10 euro each with shipping and I have already spent the next months "retro computing money" buying some other crap on Ebay today as it was pay day (one day early for once). It's probably a better idea to order caps or go on yet another Socket A board hunt in my storage units anyhow.

*The Athlon 1266 version seems rare as is the Athlon 1400(100).

I have an Athlon XP-M 2000+ 1.45V Thoroughbred CPU incoming, it was only 7.5 euro with shipping so I could not resist. It will be interesting so see how it stacks up against the Barton XP-M.

Pic from my Athlon XP-M 2400 along with the RAM and HDD from the dead laptop

Anyone else have any weird flukes with win98 and a nforce2 chipset board, driver issues?

Skyscraper wrote:

Nice! :) That CPU should have a very good chance to do 2400 MHz at 1.75V. […]
Show full quote

Carlos S. M. wrote:

I got an Athlon XP-M 2400+ from a dead HP Pavilion ze4500

exact model is AXHM2400FQQ4

is 45 watt TDP, 1.45 v and Barton core

Nice! 😀 That CPU should have a very good chance to do 2400 MHz at 1.75V.

I'm still benching Thunderbirds but the progress is slow because of work.

I'm missing multiplier 9.5x and 11x and my Athlon 850 will not do 1133MHz at stock voltage surprise surprise 🙄 so no 8.5x at 133 MHz FSB. I just discoverd that although I just fixed the pins on my Athlon 1000C (133) good enough to at least get it into the socket it still wont work and it's my only 7.5x multi Thunderbird so it makes it impossible to bench the 750 MHz version aswell.

I'm thinking of changing motherboard but my Abit KV7 VIA KT600 (has the needed multiplier control) has a blown cap and I do not think I have a suitable one, I need to order some sooner rather than later.

I have at least benched some CPUs. 😀

Spitfire Duron 700, 800, 900. (The only speed grades I own, it turned out that half of my Spitfire CPUs were the 900 MHz version.)

Athlon Thunderbird 700. 800, 850, 900, 1000(100), 1333(133) (same multiplier as 1000)

I still need to bench my Athlon 1200(100), 1200(133), 1300(100) and 1400(133) CPUs.

I looked on the net to see if I could find any cheap Athlon 950*, Athlon 1000(133) Athlon 1100 and Athlon 1133 CPUs but they were at least 10 euro each with shipping and I have already spent the next months "retro computing money" buying some other crap on Ebay today as it was pay day (one day early for once). It's probably a better idea to order caps or go on yet another Socket A board hunt in my storage units anyhow.

*The Athlon 1266 version seems rare as is the Athlon 1400(100).

I have an Athlon XP-M 2000+ 1.45V Thoroughbred CPU incoming, it was only 7.5 euro with shipping so I could not resist. It will be interesting so see how it stacks up against the Barton XP-M.

Let me look through my bin of stuff, odds are I have those missing CPUs if you want them. I also have a rare mobile Duron that is aactually a palomino core and you can adjust the l2 cache from 64k>128k>256k using a pencil across the bridges.

gbeirn wrote:

Let me look through my bin of stuff, odds are I have those missing CPUs if you want them. I also have a rare mobile Duron that is aactually a palomino core and you can adjust the l2 cache from 64k>128k>256k using a pencil across the bridges.

Sure! 😀

CPUs are luckily small and easy to ship. Send me a PM f you find any of the missing ones, I will of course cover the postage costs.