SGI: Hardware

I have coming (if it ever arrives) an Octane (1) 250MHz R10000 Single.

Now, I purchased it for reasons that are not going to require super performance obviously, but you know what it's like, you're looking at upgrades before you've even plugged it in.

So I just want to confirm compatibility and if it's worth the £30 it will cost to upgrade it to a 300MHz R12000 CPU Single.

It's going to be used as a 24/7 YP(NIS)/Backup/File Server for my IRIX machines.

I do have the option of dual 195MHz CPU that's in my loft.

Thanks.. Up The Villa!

There's no problems with this CPU AFAIK on either revision of the Octane. No issues with IRIX or anything on the upgrade either! Have fun.

Cheers Team.. I reckon I'll go for it for that price.

sure, you can, have fun

i should have an R12K @ 400Mhz module, in case you want more upgrade

I wonder if you'll see much of a difference serving files and NIS

I wouldn't have thought so, especially with a small nis. I doubt either that I'll bench mark it, since the CPU will probably come before I've got round to setting up the yp ect.

For what it will be doing, I doubt it would be worthwhile, but for £30..

I suppose the other question would be, since I have dual 195 R10000's, would that make a difference?

I've bought the CPU now anyway.

Your 300MHz CPU will be faster.

uunix wrote: since I have dual 195 R10000's, would that make a difference?

it might be a little better in certain cases but obviously only with stuff that has proper smp support.
otherwise the 250 that came with the box is just fine for what you wanna do

If it makes you feel better, I've done that exact upgrade. See the original specs on my Octane here , compared to the current hinv:



And it most certainly runs much faster then my second Octane, which 2x175MHz R10Ks in it (and also has more RAM...I should move some over....)

It does armanox, I notice though you have Xbow 1.4, I have no idea what Xbow is installed in this machine yet.

Shouldn't matter for CPU upgrade at all.

The Xbow doesn't matter, but IIRC the system board does, you might need the 1467 (later, octane 2) for dual R12k 400 and up. If I could be arse to, I would search the forum and find out for you. *hint*

duck wrote: The Xbow doesn't matter, but IIRC the system board does, you might need the 1467 (later, octane 2) for dual R12k 400 and up. If I could be arse to, I would search the forum and find out for you. *hint*

AIUI the 0887 is OK for 300MHz R12K but not (always?) 270MHz - can't remember where I saw that asserted, but a search here for "octane 270mhz 0887" yields mixed (and informative) results for that and faster CPUs.

http://www.futuretech.blinkenlights.nl/ ... ading.html mentions a specific version of an R14K for the 0887 board, but confirms a single 300MHz should be okay on 0887.

TL;DR : 400Mhz R12000 works on '887' main board .

duck wrote: The Xbow doesn't matter, but IIRC the system board does, you might need the 1467 (later, octane 2) for dual R12k 400 and up. If I could be arse to, I would search the forum and find out for you. *hint*

OK, I'll bite .

I know the 270Mhz R12000 requires the 1467 main board. Some have said any other CPU including the 600MHz R14000 runs on the 887 board. Xbow matters only if you want to install VPro graphics.

One of the differences between the '887' and the '1467' board is that the newer board can run the SysAD bus at 120MHz rather than 100MHz. I can't be arsed to figure out which CPUs support this also, but if you're going to splash serious money on that dual-600 CPU you might as well replace the main board and get the best from it.

ajw99uk wrote: AIUI the 0887 is OK for 300MHz R12K but not (always?) 270MHz - can't remember where I saw that asserted, but a search here for "octane 270mhz 0887" yields mixed (and informative) results for that and faster CPUs.

http://www.futuretech.blinkenlights.nl/ ... ading.html mentions a specific version of an R14K for the 0887 board, but confirms a single 300MHz should be okay on 0887.

Well, I was guessing, my octane runs dual 300 on an 0887 quite happily.

jan-jaap wrote: I know the 270Mhz R12000 requires the 1467 main board.

This is an interesting quirk. Has anyone speculated as to why this is the case? Or should I heed my own repeated advice and use the search?

duck wrote:

jan-jaap wrote: I know the 270Mhz R12000 requires the 1467 main board.

This is an interesting quirk. Has anyone speculated as to why this is the case? Or should I heed my own repeated advice and use the search?

A quirk indeed. Jan-jaap mentions a difference in bus speed (or an additional speed?). Speculating only, I'd guess it might be a matter of what can be derived through sensible multipliers:
270 = 120 x 2.25 (equivalent to 225 = 100 x 2.25 on the old board)
360 = 120 x 3
but both would be awkward multiples of 100.

I'm no doubt reading too much into the multiplier, as 195MHz would be just as awkward with 100 or 120 as the clock speed and, according to this the 270 was using a 275MHz chip. This post speculates a PROM issue.

ajw99uk wrote: A quirk indeed. Jan-jaap mentions a difference in bus speed (or an additional speed?). Speculating only, I'd guess it might be a matter of what can be derived through sensible multipliers:
270 = 120 x 2.25 (equivalent to 225 = 100 x 2.25 on the old board)
360 = 120 x 3
but both would be awkward multiples of 100.

There is no 2.25 clock multiplier for R10000 and R12000. Compared to R10000, R12000 lacks the x1 and x1.5 clock multipliers, but has x4.5, x5, x5.5, x6, x7, x8, x9 and x10. Now 270 == 4.5 * 60, so this might be the configuration in use (rather than 3 * 90), which can't be achieved without a motherboard correctly feeding four bits of SysClkDiv at powerup time, rather than only three since the R10000 only needs three.

I can't speak for dual 195's vs single 300, but when I upgrade my octane from dual 360's to single 600 - even though the 600 was faster - the system sure felt much more "snappy" with the dual 360's, but maybe it was just me /shrug