VOGONS


Reply 420 of 550, by Skip94

User metadata
Rank Newbie
Rank
Newbie

Finally got around to doing my first one the other night.
Found moving the resistor much easier than anticipated, just my fat fingers made lining it up afterwards a bit difficult!
Tested my Gigabyte GA 5AA (With dead on motherboard cache) and its stable at 130Mhz, but not at 135.
This particular CPU seems stable at 625 Mhz, 5x125 @ 2.0 volts and 650Mhz 5x130 @2.1v.
650 @2.0V wasn't stable.
Got another 6 to mod at some point, then I shall go through and find the best couple, and really push them. Have got a nice socket A copper heatsink for cooling.
Andrew

Reply 421 of 550, by Sphere478

User metadata
Rank l33t
Rank
l33t
Skip94 wrote on 2022-04-05, 03:13:
Finally got around to doing my first one the other night. Found moving the resistor much easier than anticipated, just my fat fi […]
Show full quote

Finally got around to doing my first one the other night.
Found moving the resistor much easier than anticipated, just my fat fingers made lining it up afterwards a bit difficult!
Tested my Gigabyte GA 5AA (With dead on motherboard cache) and its stable at 130Mhz, but not at 135.
This particular CPU seems stable at 625 Mhz, 5x125 @ 2.0 volts and 650Mhz 5x130 @2.1v.
650 @2.0V wasn't stable.
Got another 6 to mod at some point, then I shall go through and find the best couple, and really push them. Have got a nice socket A copper heatsink for cooling.
Andrew

These are 570s I assume?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 422 of 550, by Skip94

User metadata
Rank Newbie
Rank
Newbie
Sphere478 wrote on 2022-04-05, 03:35:
Skip94 wrote on 2022-04-05, 03:13:
Finally got around to doing my first one the other night. Found moving the resistor much easier than anticipated, just my fat fi […]
Show full quote

Finally got around to doing my first one the other night.
Found moving the resistor much easier than anticipated, just my fat fingers made lining it up afterwards a bit difficult!
Tested my Gigabyte GA 5AA (With dead on motherboard cache) and its stable at 130Mhz, but not at 135.
This particular CPU seems stable at 625 Mhz, 5x125 @ 2.0 volts and 650Mhz 5x130 @2.1v.
650 @2.0V wasn't stable.
Got another 6 to mod at some point, then I shall go through and find the best couple, and really push them. Have got a nice socket A copper heatsink for cooling.
Andrew

These are 570s I assume?

Yup, had a batch of 3 before this mod was discovered, sold 2 and kept one for myself. Once this was a thing, I ordered another batch of 6.
Andrew

Reply 423 of 550, by Skalabala

User metadata
Rank Member
Rank
Member
Skip94 wrote on 2022-04-05, 03:13:
Finally got around to doing my first one the other night. Found moving the resistor much easier than anticipated, just my fat fi […]
Show full quote

Finally got around to doing my first one the other night.
Found moving the resistor much easier than anticipated, just my fat fingers made lining it up afterwards a bit difficult!
Tested my Gigabyte GA 5AA (With dead on motherboard cache) and its stable at 130Mhz, but not at 135.
This particular CPU seems stable at 625 Mhz, 5x125 @ 2.0 volts and 650Mhz 5x130 @2.1v.
650 @2.0V wasn't stable.
Got another 6 to mod at some point, then I shall go through and find the best couple, and really push them. Have got a nice socket A copper heatsink for cooling.
Andrew

Stable 650MHz @ 2.1V. Now that sounds like business. How about 120x5.5 for 660MHz? For how long can it handle 720MHz? Mine can do about 10 seconds 🤣.

Reply 424 of 550, by Fritzchens Fritz

User metadata
Rank Newbie
Rank
Newbie

Got a batch of these 570ACZ CPUs . Tested all with 128KB L2$, Water and 2,4V. The best one running @ 682MHz primestable.

I moved the resistor of the best one, printed a frame for the heatspreader and using the clamp of der bauers delid die mate 2 to build up sufficient pressure. Thermal grease is Aeronaut and as glue I used black "UHU high temperature silicone". Will now wait 24 hours and then see what the CPU says. Hope the voltage can be lowered.

DSC05075.JPG
Filename
DSC05075.JPG
File size
382.83 KiB
Views
889 views
File license
Public domain
DSC05077.JPG
Filename
DSC05077.JPG
File size
233.01 KiB
Views
889 views
File license
Public domain

Reply 425 of 550, by Sphere478

User metadata
Rank l33t
Rank
l33t
Fritzchens Fritz wrote on 2022-04-05, 22:15:
Got a batch of these 570ACZ CPUs . Tested all with 128KB L2$, Water and 2,4V. The best one running @ 682MHz primestable. […]
Show full quote

Got a batch of these 570ACZ CPUs . Tested all with 128KB L2$, Water and 2,4V. The best one running @ 682MHz primestable.

I moved the resistor of the best one, printed a frame for the heatspreader and using the clamp of der bauers delid die mate 2 to build up sufficient pressure. Thermal grease is Aeronaut and as glue I used black "UHU high temperature silicone". Will now wait 24 hours and then see what the CPU says. Hope the voltage can be lowered.

DSC05075.JPG

DSC05077.JPG

Beautiful work!

Can you post 3d files? And where you got that clamp and its part number?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 426 of 550, by Repo Man11

User metadata
Rank Oldbie
Rank
Oldbie

Now that this impossible thing has been discovered, could we discover a way to have a 7x multiplier so I can run my TXP4 at 583 MHz?

"A lot of times when you first start out on a project you think, This is never going to be finished. But then it is, and you think, Wow, it wasn't even worth it." - Jack Handey

Reply 427 of 550, by Sphere478

User metadata
Rank l33t
Rank
l33t
Repo Man11 wrote on 2022-04-06, 03:22:

Now that this impossible thing has been discovered, could we discover a way to have a 7x multiplier so I can run my TXP4 at 583 MHz?

it's too bad that the multiplier is in the cpu its self..

I still hold out hope for New processors for old sockets.

interestingly a intel Atom processor was actually simulated and run on a socket 7 mobo.

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 428 of 550, by Repo Man11

User metadata
Rank Oldbie
Rank
Oldbie
Sphere478 wrote on 2022-04-06, 04:39:
it's too bad that the multiplier is in the cpu its self.. […]
Show full quote
Repo Man11 wrote on 2022-04-06, 03:22:

Now that this impossible thing has been discovered, could we discover a way to have a 7x multiplier so I can run my TXP4 at 583 MHz?

it's too bad that the multiplier is in the cpu its self..

I still hold out hope for New processors for old sockets.

interestingly a intel Atom processor was actually simulated and run on a socket 7 mobo.

Yeah, it's just a shame. The performance of the TX chipset shows that if Intel had made a Super 7 chipset, it would have really been something.

Compare the memory bandwidth of the TX Vs. the MVP3. The MVP3 is running 112 MHz FSB against 83 for the TX, has 1024k L3 (the L3 is disabled on the TX because with that much memory and the K6-3+ the score is lower with it enabled) and it has four way interleave enabled, while the TX doesn't have that feature.

Attachments

"A lot of times when you first start out on a project you think, This is never going to be finished. But then it is, and you think, Wow, it wasn't even worth it." - Jack Handey

Reply 429 of 550, by Fritzchens Fritz

User metadata
Rank Newbie
Rank
Newbie

So, the 570ACZ running very well with the full amount of 256KB L2 cache.

672MHz (112x6) @ 2.1V with Water and activated L3 cache
682Mhz (124x5.5) @ 2.2V with Water and no L3 cache

700MHZ+ are not possible (even with 2.4V)

2022-04-07_111637_306.png
Filename
2022-04-07_111637_306.png
File size
1.25 MiB
Views
782 views
File license
Public domain
Sphere478 wrote on 2022-04-06, 00:41:

Beautiful work!

Can you post 3d files? And where you got that clamp and its part number?

I think there is no part number for the clamp. The clmap is only available together with the "delid die mate 2" kit. But you could also print such a clamp or get a normal screw clamp at the local hardware store.

Filename
AMD K6 Sticky Frame.zip
File size
838 Bytes
Downloads
29 downloads
File license
Public domain

Reply 431 of 550, by Sphere478

User metadata
Rank l33t
Rank
l33t
Fritzchens Fritz wrote on 2022-04-07, 10:47:
So, the 570ACZ running very well with the full amount of 256KB L2 cache. […]
Show full quote

So, the 570ACZ running very well with the full amount of 256KB L2 cache.

672MHz (112x6) @ 2.1V with Water and activated L3 cache
682Mhz (124x5.5) @ 2.2V with Water and no L3 cache

700MHZ+ are not possible (even with 2.4V)

2022-04-07_111637_306.png

Sphere478 wrote on 2022-04-06, 00:41:

Beautiful work!

Can you post 3d files? And where you got that clamp and its part number?

I think there is no part number for the clamp. The clmap is only available together with the "delid die mate 2" kit. But you could also print such a clamp or get a normal screw clamp at the local hardware store.

AMD K6 Sticky Frame.zip

That’s an amazing result. You should take that chip to the 3d mark 2000 aiming for the atars thread.

What will it do on air cooling?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 432 of 550, by bloodem

User metadata
Rank Oldbie
Rank
Oldbie
Sphere478 wrote on 2022-04-07, 15:51:

That’s an amazing result. You should take that chip to the 3d mark 2000 aiming for the atars thread.

It's very unlikely that an Epox MVP3 based board will break any records, even with such a high overclock. 😀

2 x PGA132 / 5 x Socket 3 / 9 x Socket 7 / 12 x SS7 / 1 x Socket 8 / 14 x Slot 1 / 5 x Slot A
5 x Socket 370 / 8 x Socket A / 2 x Socket 478 / 2 x Socket 754 / 3 x Socket 939 / 7 x LGA775 / 1 x LGA1155
Current PC: Ryzen 7 5800X3D
Backup PC: Core i7 7700k

Reply 433 of 550, by Repo Man11

User metadata
Rank Oldbie
Rank
Oldbie
Repo Man11 wrote on 2022-03-29, 23:34:

As far as overclocking goes, I can hit 616 on my Epox board with no issue; that's the speed I've left it at. In my Soyo board I could hit 620 (5x124) with this cpu, as well as with my K6-2+ 550, but when I try it in this board it won't always POST, etc. So I'm going to recap my Epox board to make sure that isn't the issue. I ordered the caps today.

After going through the trouble and expense of recapping my Epox MVP3G-M I found that nothing changed (though it does look nicer); when I selected 124 FSB (5x124) it would get stuck in a POST loop that you can only get out of by clearing the CMOS or by putting the FSB jumper on 66 MHz. Knowing that this CPU would do that speed on my Soyo board, I was stumped. I finally tried loading the BIOS defaults, and only then did I discover that this board won't work with the FSB set higher than 112 MHz if ACPI is disabled! Once ACPI was enabled, it ran fine with the higher bus speed. But once again the motherboard's cache won't work at that speed and has to be disabled. The net result is a 3D 2000 score at least two hundred points lower than the highest score @ 5.5x112 with the motherboard's cache enabled.

It will POST at 4.5x133, but the memory settings have to be set to their slowest values, and it's very unstable. it will POST at 6x112, but it would always blue screen when trying to load Windows no matter what I tried. At 5.5x124, I couldn't get it to be stable.

I guess the key is to buy them by the lot and look for the best out of them. It reminds me of Charlie and the Chocolate Factory where some of the characters were buying huge lots of Wonka Bars to find the ones that had a Golden Ticket inside.

"A lot of times when you first start out on a project you think, This is never going to be finished. But then it is, and you think, Wow, it wasn't even worth it." - Jack Handey

Reply 435 of 550, by Repo Man11

User metadata
Rank Oldbie
Rank
Oldbie
Nemo1985 wrote on 2022-04-11, 17:54:

Are you doing all of these testing with the l2 cache disable I hope.
The mvp3 use very tight timings on the cache and even with -4ns chips it's hard to overclock the bus over 100mhz.

On both of my MVP3 boards the motherboard's cache works fine at 112 MHz FSB, but no higher.

"A lot of times when you first start out on a project you think, This is never going to be finished. But then it is, and you think, Wow, it wasn't even worth it." - Jack Handey

Reply 437 of 550, by debs3759

User metadata
Rank Oldbie
Rank
Oldbie
Nemo1985 wrote on 2022-04-11, 18:20:

You should try with l2 cache disabled, it may allow you to reach higher fsb frequencies.

If you read his post, he tried with cache disabled, but the performance was dismal (as can be expected)

See my graphics card database at www.gpuzoo.com
Constantly being worked on. Feel free to message me with any corrections or details of cards you would like me to research and add.

Reply 438 of 550, by Sphere478

User metadata
Rank l33t
Rank
l33t

I’ve stubbled across a 570 that isn’t stable unlocked at 550. Or 500 it seems.

Might have to pull it apart and re lock it.

edit: not even stable at 400mhz 2.1v (better though)

Edit2
Re: SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod

Last edited by Sphere478 on 2022-04-24, 10:45. Edited 2 times in total.

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 439 of 550, by bloodem

User metadata
Rank Oldbie
Rank
Oldbie
Sphere478 wrote on 2022-04-12, 05:53:

I’ve stubbled across a 570 that isn’t stable unlocked at 550. Or 500 it seems.

Might have to pull it apart and re lock it.

Too bad, but I guess it's normal (some might have actually had issues with that portion of the cache).

Why didn't you test it before reassembling? If you're afraid of damaging the die, don't be, you only need to put a bit of thermal paste and a cooler on top (without using the retention clips).
Just make sure that the cooler doesn't slide. 😀 I've already done this four times and it worked great for 2+ hours during the Prime 95 + 3DMark2000 stability testing.

2 x PGA132 / 5 x Socket 3 / 9 x Socket 7 / 12 x SS7 / 1 x Socket 8 / 14 x Slot 1 / 5 x Slot A
5 x Socket 370 / 8 x Socket A / 2 x Socket 478 / 2 x Socket 754 / 3 x Socket 939 / 7 x LGA775 / 1 x LGA1155
Current PC: Ryzen 7 5800X3D
Backup PC: Core i7 7700k