VOGONS


First post, by supergeek

User metadata
Rank Newbie
Rank
Newbie

There's a total of three sSpecs I'm seeing:

SL293
SL2BM
SL27S

Which one is the newest and presumably the"best" one?

MY SYSTEMS...
main: i7-12700KF + 64 GB DDR5 @ 6200 + RTX 4070 FE
server: i5-3570 + 16 GB
NAS: E3-1230 v2 + 32 GB ECC + 2 x 18 TB
mobile: X9000 + 4 GB
90s rig: P166 + 96 MB EDO + TBD

Reply 1 of 14, by PARKE

User metadata
Rank Oldbie
Rank
Oldbie
supergeek wrote on 2025-08-01, 21:23:
There's a total of three sSpecs I'm seeing: SL293 SL2BM SL27S Which one is the newest and presumably the"best" one? […]
Show full quote

There's a total of three sSpecs I'm seeing:
SL293
SL2BM
SL27S
Which one is the newest and presumably the"best" one?

Steppings are the same - see page 23

The attachment 24248041.pdf is no longer available

Reply 3 of 14, by supergeek

User metadata
Rank Newbie
Rank
Newbie
PARKE wrote on 2025-08-02, 14:58:

Steppings are the same - see page 23

The attachment 24248041.pdf is no longer available

OK but why have different sSpecs is what I'm asking I guess.

MY SYSTEMS...
main: i7-12700KF + 64 GB DDR5 @ 6200 + RTX 4070 FE
server: i5-3570 + 16 GB
NAS: E3-1230 v2 + 32 GB ECC + 2 x 18 TB
mobile: X9000 + 4 GB
90s rig: P166 + 96 MB EDO + TBD

Reply 4 of 14, by gerwin

User metadata
Rank l33t
Rank
l33t
supergeek wrote on 2025-08-02, 18:19:
PARKE wrote on 2025-08-02, 14:58:

Steppings are the same - see page 23

The attachment 24248041.pdf is no longer available

OK but why have different sSpecs is what I'm asking I guess.

Only difference I see is in the notes:
https://www.cpu-world.com/sspec/SL/SL293.html Notes on sSpec SL293 - Boxed Pentium processor with an attached fan heatsink.
https://www.cpu-world.com/sspec/SL/SL2BM.html Notes on sSpec SL2BM - Boxed Pentium processor without an attached fan heatsink.
https://www.cpu-world.com/sspec/SL/SL27S.html Notes on sSpec SL27S - This processor was also offered as an embedded microprocessor.

--> ISA Soundcard Overview // Doom MBF 2.04 // SetMul

Reply 5 of 14, by supergeek

User metadata
Rank Newbie
Rank
Newbie
gerwin wrote on 2025-08-02, 19:23:

https://www.cpu-world.com/sspec/SL/SL27S.html Notes on sSpec SL27S - This processor was also offered as an embedded microprocessor.

What is an "embedded" CPU offering in this case? I'm getting mixed results looking it up myself, so thought I'd ask for clarification.

MY SYSTEMS...
main: i7-12700KF + 64 GB DDR5 @ 6200 + RTX 4070 FE
server: i5-3570 + 16 GB
NAS: E3-1230 v2 + 32 GB ECC + 2 x 18 TB
mobile: X9000 + 4 GB
90s rig: P166 + 96 MB EDO + TBD

Reply 7 of 14, by supergeek

User metadata
Rank Newbie
Rank
Newbie
PARKE wrote on 2025-08-02, 20:20:
supergeek wrote on 2025-08-02, 19:38:

What is an "embedded" CPU offering in this case? I'm getting mixed results looking it up myself, so thought I'd ask for clarification.

Embedded in this case means it is soldered on a pcb:
https://www.ebay.com/itm/116042690010

So SL27S is sometimes soldered, sometimes not? Now I'm curious as to why it's "sometimes" as opposed to "always". Like, why not have a given sSpec be only soldered/embedded?

MY SYSTEMS...
main: i7-12700KF + 64 GB DDR5 @ 6200 + RTX 4070 FE
server: i5-3570 + 16 GB
NAS: E3-1230 v2 + 32 GB ECC + 2 x 18 TB
mobile: X9000 + 4 GB
90s rig: P166 + 96 MB EDO + TBD

Reply 8 of 14, by bertrammatrix

User metadata
Rank Member
Rank
Member
supergeek wrote on 2025-08-02, 22:54:
PARKE wrote on 2025-08-02, 20:20:
supergeek wrote on 2025-08-02, 19:38:

What is an "embedded" CPU offering in this case? I'm getting mixed results looking it up myself, so thought I'd ask for clarification.

Embedded in this case means it is soldered on a pcb:
https://www.ebay.com/itm/116042690010

So SL27S is sometimes soldered, sometimes not? Now I'm curious as to why it's "sometimes" as opposed to "always". Like, why not have a given sSpec be only soldered/embedded?

Because the stepping refers to the cpu die revision, NOT the PACKAGE the cpu die is placed in (for instance PGA for a socket or QFP to be soldered).

Reply 9 of 14, by supergeek

User metadata
Rank Newbie
Rank
Newbie
bertrammatrix wrote on Yesterday, 02:49:

Because the stepping refers to the cpu die revision, NOT the PACKAGE the cpu die is placed in (for instance PGA for a socket or QFP to be soldered).

The only real relevant thing is the stepping, correct (assuming you're getting the right package or whatever)? With newer steppings almost always being preferred?

MY SYSTEMS...
main: i7-12700KF + 64 GB DDR5 @ 6200 + RTX 4070 FE
server: i5-3570 + 16 GB
NAS: E3-1230 v2 + 32 GB ECC + 2 x 18 TB
mobile: X9000 + 4 GB
90s rig: P166 + 96 MB EDO + TBD

Reply 10 of 14, by bertrammatrix

User metadata
Rank Member
Rank
Member
supergeek wrote on Yesterday, 02:57:
bertrammatrix wrote on Yesterday, 02:49:

Because the stepping refers to the cpu die revision, NOT the PACKAGE the cpu die is placed in (for instance PGA for a socket or QFP to be soldered).

The only real relevant thing is the stepping, correct (assuming you're getting the right package or whatever)? With newer steppings almost always being preferred?

Correct

Reply 11 of 14, by PARKE

User metadata
Rank Oldbie
Rank
Oldbie

My opinion is that we should not overthink this stepping thing. Steppings were done for various reasons, fixing bugs was a major reason, and steppings that were substantially 'better' to such an extent that we mortals would actually notice a difference are not that common. Even when a certain stepping was for example a bit faster the difference may have been subtle enough to level out differences between individual chips i.e. the 'best' chip of the earlier stepping may be faster than the worst of the later one.

Reply 12 of 14, by bertrammatrix

User metadata
Rank Member
Rank
Member
PARKE wrote on Yesterday, 21:35:

My opinion is that we should not overthink this stepping thing. Steppings were done for various reasons, fixing bugs was a major reason, and steppings that were substantially 'better' to such an extent that we mortals would actually notice a difference are not that common. Even when a certain stepping was for example a bit faster the difference may have been subtle enough to level out differences between individual chips i.e. the 'best' chip of the earlier stepping may be faster than the worst of the later one.

Pretty much. There is always a variance between the quality of the individual chips themselves within a batch, which will often play a much bigger role in the particular chips overclockability then which exact stepping it is. This is true more often then not.

The question shouldn't really be "which is the best/latest stepping for a mmx cpu ", but, "which cpu originating from ANY of the problem-free steppings just happened to originate somewhere near the middle of the wafer and is likely to have the best quality "

Reply 13 of 14, by bertrammatrix

User metadata
Rank Member
Rank
Member

Were you possibly wondering what the -fastest- socket 7 Pentium mmx was?

Because afaik that would be Tillamook core Pentium mmx (whatever stepping). These were only ever officially produced in some solder on form of packaging HOWEVER some made it into some china (?) produced socket7 PGA interposer packaging, known for their gold color as "Golden Tiger" Pentium's. There is a few threads about them on here. They do pop up on ebay every so often. They have a stepping written on them, however usually it is fake. They are supposedly great overclockers, however you probably need a supporting board/ modified bios and SHOULD have proper lower voltage support (though they are known to run even highly overvolted)

Reply 14 of 14, by dionb

User metadata
Rank l33t++
Rank
l33t++

Er no, there were always OPGA Tillamooks without interposers. Not common though and not well supported as everything else you say applies to them too. In any event the differences between P55C and Tillamook are bigger than just a stepping, with significant new functionality (BF2 pin for starters), different voltage spec and changes to cache system (causing the compatibility issues), so it's definitely not just a 'stepping' of Pentium MMX, let alone of the P233MMX specifically.