R5000PC to R5000SC?
Forum rules
Any posts concerning pirated software or offering to buy/sell/trade commercial software are subject to removal.
Any posts concerning pirated software or offering to buy/sell/trade commercial software are subject to removal.
- Raion-Fox
- Donor
- Posts: 1598
- Joined: Thu Jan 30, 2014 5:01 pm
- Location: near King George, Virginia
- Contact:
R5000PC to R5000SC?
So I have a 150MHz R5000PC board here. I noticed there's little difference between it and a SC other than the RAM chips installed. Assuming one could find the RAM chips sold separately by someone and had them soldered on, would the CPU module recognize it or require a logic upgrade too?






I am probably posting from yangxiaolong, HP Z230 with Xeon E3-1230v3, 16GB RAM, GeForce 750ti, and running NetBSD and Windows 8.1 Embedded.
Owner and operator of http://irix.cc
-
- Posts: 1667
- Joined: Mon Sep 12, 2011 2:28 pm
- Location: Boston
Re: R5000PC to R5000SC?
I assume you mean for an Indy IP24; I don't think the O2 processors went as low as 150 MHz.
As far as I know, all R5000 chips contain secondary cache controllers. The issues you may face are:
1) The processor cards sold as R5000PC may not have been fully tested and qualified for use with external cache chips. I would say this problem is unlikely. It is possible that SGI designed two different E-Test profiles for the two products, but that would double the expensive engineering work for dubious gains.
2) The processor chip itself may be binned based on secondary cache functionality. Again this is possible, but I'm not aware of any evidence that they did this. The datasheet (which I can't find ATM) would show two different product codes.
3) The processor needs to be configured to enable the secondary cache. This means writing certain values to the SPRs at boot time. I believe there is a serial PROM on the module that is used for this? There was discussion some years back about how the SPROM is used to set up R10000 chips and I would guess it would be similar.
Another hack that would be possible is to change XL8 to XL24 or vice versa. The only difference is the presence of some additional VRAM and maybe some different resistors.
As far as I know, all R5000 chips contain secondary cache controllers. The issues you may face are:
1) The processor cards sold as R5000PC may not have been fully tested and qualified for use with external cache chips. I would say this problem is unlikely. It is possible that SGI designed two different E-Test profiles for the two products, but that would double the expensive engineering work for dubious gains.
2) The processor chip itself may be binned based on secondary cache functionality. Again this is possible, but I'm not aware of any evidence that they did this. The datasheet (which I can't find ATM) would show two different product codes.
3) The processor needs to be configured to enable the secondary cache. This means writing certain values to the SPRs at boot time. I believe there is a serial PROM on the module that is used for this? There was discussion some years back about how the SPROM is used to set up R10000 chips and I would guess it would be similar.
Another hack that would be possible is to change XL8 to XL24 or vice versa. The only difference is the presence of some additional VRAM and maybe some different resistors.




Re: R5000PC to R5000SC?
On the Indy, the external cache controller of the R5000 is never used - the secondary cache in R5000SC (and R4600SC) processor boards is a write-through cache specific to Indy, with specific handling in IRIX.
I don't know how the PROM detects that the secondary cache is present on such modules. It might simply be an SPD line being pulled down by the simple presence of the memory chips in the module, in which case nothing else will be required; but there might be small wiring differences between the R5000PC and R5000SC modules.
If you can have the appropriate memory chips soldered correctly, your module *might* be recognized as an SC. The worst outcome would be the module still recognized as PC and consuming more electricity due to the need to get the SC memory chips powered on.
I don't know how the PROM detects that the secondary cache is present on such modules. It might simply be an SPD line being pulled down by the simple presence of the memory chips in the module, in which case nothing else will be required; but there might be small wiring differences between the R5000PC and R5000SC modules.
If you can have the appropriate memory chips soldered correctly, your module *might* be recognized as an SC. The worst outcome would be the module still recognized as PC and consuming more electricity due to the need to get the SC memory chips powered on.





















among more than 150 machines : Apollo, Data General, Digital, HP, IBM, MIPS before SGI, Motorola, NeXT, SGI, Solbourne, Sun...
Re: R5000PC to R5000SC?
There is a serial eeprom on the board with boot information for the processor. At least the clock multiplier and the cache size are configured by this information. I successfully moved the eeprom data from a 180mhz board to a 150mhz board and replaced the processor by a 200mhz version. If there is any interest I may publish the eeprom data from both SC processor versions.










- Raion-Fox
- Donor
- Posts: 1598
- Joined: Thu Jan 30, 2014 5:01 pm
- Location: near King George, Virginia
- Contact:
Re: R5000PC to R5000SC?
Thanks diegel for the info. I dont possess an eeprom writer, but it may be a fun experiment sometime.






I am probably posting from yangxiaolong, HP Z230 with Xeon E3-1230v3, 16GB RAM, GeForce 750ti, and running NetBSD and Windows 8.1 Embedded.
Owner and operator of http://irix.cc
Who is online
Users browsing this forum: No registered users and 2 guests