Power Challenge L, 12x R10000 CPUs
#1
Power Challenge L, 12x R10000 CPUs
My deskside Power Challenge L, recently revived:

Code:
everest 27# hinv -v
CPU Board at Slot 2: (Enabled)
  Processor 0 at Slot 2/Slice 0: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
  Processor 1 at Slot 2/Slice 1: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
  Processor 2 at Slot 2/Slice 2: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
  Processor 3 at Slot 2/Slice 3: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
CPU Board at Slot 3: (Enabled)
  Processor 4 at Slot 3/Slice 0: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
  Processor 5 at Slot 3/Slice 1: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
  Processor 6 at Slot 3/Slice 2: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
  Processor 7 at Slot 3/Slice 3: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
CPU Board at Slot 4: (Enabled)
  Processor 8 at Slot 4/Slice 0: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
  Processor 9 at Slot 4/Slice 1: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
  Processor 10 at Slot 4/Slice 2: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
  Processor 11 at Slot 4/Slice 3: 194 Mhz R10000 with 2 MB secondary cache (Enabled)
CPU: MIPS R10000 Processor Chip Revision: 2.6
FPU: MIPS R10010 Floating Point Chip Revision: 0.0
Secondary unified instruction/data cache size: 2 Mbytes
Data cache size: 32 Kbytes
Instruction cache size: 32 Kbytes
Main memory size: 2048 Mbytes, 2-way interleaved
MC3 Memory Board at Slot 1: 2048 MB of memory (Enabled)
  Bank A contains 64 MB SIMMS (Enabled)
  Bank B contains 64 MB SIMMS (Enabled)
  Bank C contains 64 MB SIMMS (Enabled)
  Bank D contains 64 MB SIMMS (Enabled)
  Bank E contains 64 MB SIMMS (Enabled)
  Bank F contains 64 MB SIMMS (Enabled)
  Bank G contains 64 MB SIMMS (Enabled)
  Bank H contains 64 MB SIMMS (Enabled)
I/O board, Ebus slot 5: IO4 revision 1
Integral EPC serial ports: 4
Integral Ethernet controller: et0, Ebus slot 5
FDDIXPress controller: ipg0, version 1
EPC external interrupts
Integral SCSI controller 4: Version WD33C95A, differential, revision 1
Integral SCSI controller 3: Version WD33C95A, differential, revision 1
Integral SCSI controller 2: Version WD33C95A, differential, revision 1
Integral SCSI controller 1: Version WD33C95A, single ended, revision 0
  Disk drive: unit 1 on SCSI controller 1
Integral SCSI controller 0: Version WD33C95A, single ended, revision 0
CC synchronization join counter
Integral EPC parallel port: Ebus slot 5
VME bus: adapter 0 mapped to adapter 21
VME bus: adapter 21

everest 28# uname -a
IRIX64 everest 6.2 03131016 IP25

everest 29# diskpatch -v
sc1d1l0:  Disk         SEAGATE ST336754LW      0003  Serial: 3KQ1M59V

Inside, left to right: MC3, 3x IP25, IO4, V/FDDI:

[Image: IMG_7310_sm.JPG]

The IO4 is fitted with a dual 1Gb/s FCAL mezzanine, and a triple channel SCSI mezzanine. There's also an Interphase VME FDDI board installed.

This system recently had some power problems, but I managed to fix them. The beast is back  Cool
(This post was last modified: 12-07-2021, 12:53 PM by jan-jaap.)
jan-jaap
SGI Collector

Trade Count: (0)
Posts: 1,048
Threads: 37
Joined: Jun 2018
Location: Netherlands
Website Find Reply
12-06-2021, 10:44 PM
#2
RE: Power Challenge L, 12x R10000 CPUs
Very nice system!
Irinikus
Hardware Connoisseur

Trade Count: (0)
Posts: 3,475
Threads: 319
Joined: Dec 2017
Location: South Africa
Website Find Reply
12-07-2021, 03:25 PM
#3
RE: Power Challenge L, 12x R10000 CPUs
Irinikus said it well:  very nice system, indeed.

I love seeing pictures of the deskside and larger systems.  I'm unlikely to ever acquire that kind of hardware myself (due to shipping) so photos are great.  (Here and in other threads.)

Plus, it must be very satisfying to revive one of those beasts!

SGI:  Indigo, Indigo2, Octane, Origin 300
Sun:  SPARCstation 20 (x4), Ultra 2, Blade 2500, T5240
HP:  9000/380, 425e, C8000
Digital: DECstation 5000/125, PWS 600au
jpstewart
Developer

Trade Count: (1)
Posts: 444
Threads: 6
Joined: May 2018
Location: SW Ontario, CA
Find Reply
12-08-2021, 01:07 AM
#4
RE: Power Challenge L, 12x R10000 CPUs
Thanks.

Back in 2009, we needed an Onyx2 for work, and it turned out that the easiest way to get one was to buy a pallet of them from a recycler. I made a deal with my employer: I would set up an Onyx2 for them, plus a spare, and the rest would be mine. So I ended up with 6 or 7 Onyx2's, and I traded one of them with a fellow Dutch SGI hobbyist for a pair of Challenge desksides. This is one of them. Initially it was a Challenge DM, so it was restricted to a single CPU. I posted on Nekochan how to remove this restriction -- ask me if you ever need to.

I normally care more for the systems with graphics capabilities (Silicon Graphics, you know), and the overall idea of my collection is to have one system of every family. So I got rid of my Origin 2100 in favor of the Onyx2, for example. But the Challenge has made me bend my rules a little. I guess it helps that it always been a rock solid system, and I haven't had much luck with the Onyx1. I'm a sucker for systems that need to be started with an ignition key, and I much prefer the deep hum of the big blower in these over the cacophony of fan noises of some other systems.

I have three sets of CPU boards for this one, and a system disk with matching IRIX installation:
* Usually I run it with 3x IP25 boards (12x R10000 @ 194MHz) and IRIX 6.2
* I have a set of three IP21 boards for it (6x R8000 @ 90MHz I think), and again an IRIX 6.2 disk
* Finally, I have a set of 3x IP19 boards (12x R4400 @ 250MHz), and an IRIX 5.3 system disk.

It will comfortably run IRIX 6.5 in all configurations, but what's the fun in that? I have an Origin 350 for that. This beast, with R4400 CPUs, is easily the fastest IRIX 5.3 system you can have except for a tall rack Challenge XL. Same for IRIX 6.2 and the IP25 CPUs. If you've got a job that's sufficiently parallel, it's a surprisingly powerful system.

The Challenge/Onyx have dual wide SCSI buses to the drive bays (20MB/s), with bus 0 single ended (tape drives, CDROM), and bus 1 HVD, high voltage differential SCSI, for hard disks. I have no idea why anybody would want to use HVD SCSI internally in a system. HVD is for long cable lengths, so external disk arrays, tape robots etc. It is dangerous: if you mix it with SE or LVD SCSI you will destroy something. Fortunately, it is possible to reconfigure a SCSI bus in the Challenge/Onyx from HVD to SE. You need to replace a red mezzanine card on the IO4 for a green one, and you need to reconfigure the jumpers in the drive bay backplane and all drive sleds, and replace a terminator. I've done that in this system. This way I not only got rid of the dangerous HVD SCSI, but I could also use a relatively recent Seagate Cheetah 15K.4 disk rather than the clunky and unreliable disks these systems came with originally.

Here's a photo of an IP25 CPU board. I believe this one is from my Onyx, though. I think it's one of the prettiest designs: the symmetry, the engineering that went into it (check those aluminum reinforcements!). They're pretty heavy.

[Image: IMG_7251_sm.JPG]
jan-jaap
SGI Collector

Trade Count: (0)
Posts: 1,048
Threads: 37
Joined: Jun 2018
Location: Netherlands
Website Find Reply
12-08-2021, 10:01 AM


Forum Jump:


Users browsing this thread: 1 Guest(s)