O2 MIPS R5000 vs QED RM5200
Brian
the QED RM5200 300 MHz? I've gone through the white paper at
http://www.sgi.com/Products/PDF/2510.pdf but it didn't offer much in the
way of comparing these two.
DZ
-= Just because you're not paraniod doesn't mean
everyone's *not* out to get you =-
G. Douglas
>
> Has anyone had a chance to compare the 200-300 MHz MIPS R5000 O2
R5000-200 Mhz. was the fastest R5000 ever shipped in an O2.
> against
> the QED RM5200 300 MHz?
The RM5200 is not a great deal different than the R5000,
I think it was just a die shrink (gives better yield per wafer,
and runs cooler per given clock speed, as the smaller geometry
uses less power. This enables it to be clocked faster, as the
smaller lines in the silicon don't melt, resulting in
catastrophic failure mode i.e. "letting the magic smoke out
of the chip".
So, the RM5200 performance is pretty much linear with clock
speed - a RM5200-300 will be approx. 50% faster than an
R5000-200 Mhz.
The R5000SC-180MHz. CPU has only .5MB secondary (L2) cache,
so the performance increase will be greater than just the linear
clock speed increase with either the 200 MHz. R5000 or 300 Mhz.
RM5200, as the latter two have 1MB secondary cache.
Greg Douglas
Reputable Systems
Neil Rothwell
>Has anyone had a chance to compare the 200-300 MHz MIPS R5000 O2 against
>the QED RM5200 300 MHz? I've gone through the white paper at
>http://www.sgi.com/Products/PDF/2510.pdf but it didn't offer much in the
>way of comparing these two.
Their isn't a 300 Mhz MIPS R5000 O2, I assume you mean compare 200 Mhz
R5000 O2 to 300 Mhz RM5200 O2.
Having upgraded my O2 R5k-180 to a R5200-300 I find CPU intensive tasks
are about 40% faster (probably about 32% faster for 200 -> 300).
With my CAD application (SDRC I-Deas) time to replay a complex 3-D solid
model (something I do a lot off - CPU intensive task):
O2 R5000-180 - 1155 secs
O2 R5200-300 - 700 secs
A saving of over 7 mins on an operation I do many times a day - well
worth it.
As the O2 does most of its geometry calcs on the CPU I see a similar
speed up in graphics performance as well.
Also have a R12-270 O2 here as well, same replay completes in:
O2 R12K-270 - 497 secs
Not *that* much better than the RM5200-300 but at least it gives me an
upgrade patch to R12K-400.
Also see Ian Mapleston's site http://www.futuretech.vuurwerk.nl/sgi.html
for lots of O2 benchmarks.
Cheers
-------------------------------------------------
Neil Rothwell
Freelance Mechanical Design Engineer
E-Mail: ne...@rothers.demon.co.uk
Home Page: http://www.rothers.demon.co.uk
-------------------------------------------------
Alexis Cousein
> Has anyone had a chance to compare the 200-300 MHz MIPS R5000 O2 against
> the QED RM5200 300 MHz?
The 300MHz R5K O2 *is* an RM5200-based machine -- or did you mean the
195MHz/300MHz/400MHz R10K/R12K O2s vs. the RM5200?
--
Alexis Cousein System Engineer
SGI Belgium and Luxemburg a...@brussels.sgi.com
Brian
Alexis Cousein wrote:
> The 300MHz R5K O2 *is* an RM5200-based machine -- or did you mean the
> 195MHz/300MHz/400MHz R10K/R12K O2s vs. the RM5200?
>
I mean the MIPS R5000 vs. the QED RM5200, since the only upgrade I've
seen for the O2 R5k is to the QED RM5200. I'd love to swap out my 4
200MHz R5k O2's (and single 300Mhz) for R10k's or R12000A's, but my
Edgemark rep. hasn't found anything to let me do that.
If the 300MHz O2 is an RM5200, then why does my one and only 300Mhz O2's
hinv report:
CPU: MIPS R5000 Processor Chip Revision: 10.0
FPU: MIPS R5000 Floating Point Coprocessor Revision: 10.0
1 300 MHZ IP32 Processor
and not "CPU: QED RM5200?"
Thanks,
Brian (DZ)
David Evans
Brian <dzb...@NOSPAMhome.com> wrote:
>
>If the 300MHz O2 is an RM5200, then why does my one and only 300Mhz O2's
>hinv report:
> CPU: MIPS R5000 Processor Chip Revision: 10.0
> FPU: MIPS R5000 Floating Point Coprocessor Revision: 10.0
> 1 300 MHZ IP32 Processor
>
>and not "CPU: QED RM5200?"
>
AFAIK either there's no way for the software to tell the difference between
the R5000 and RM5200 or there's no point in telling the difference because
they're functionally identical. The R5000/RM5200 thing isn't anywhere near
the difference between an R4000 and an R4400.
--
David Evans (NeXTMail/MIME OK) dfe...@bbcr.uwaterloo.ca
PhD Student, Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo "Default is the value selected by the composer
Ontario, Canada overridden by your command." - Roland TR-707 Manual
Alexis Cousein
>
> If the 300MHz O2 is an RM5200, then why does my one and only 300Mhz O2's
> hinv report:
> CPU: MIPS R5000 Processor Chip Revision: 10.0
> FPU: MIPS R5000 Floating Point Coprocessor Revision: 10.0
> 1 300 MHZ IP32 Processor
>
> and not "CPU: QED RM5200?"
Because the kernel engineer wrote the string this way ;). I can assure
you that's a QED RM5200.
Tarjei T. Jensen
Brian wrote
>I mean the MIPS R5000 vs. the QED RM5200, since the only upgrade I've
>seen for the O2 R5k is to the QED RM5200. I'd love to swap out my 4
>200MHz R5k O2's (and single 300Mhz) for R10k's or R12000A's, but my
>Edgemark rep. hasn't found anything to let me do that.
I don't think the R10k or R12K is a matter of just getting a new CPU board. I
think you have to change the motherboard as well.
Another alternative for the O2 is to go for a RM7000 CPU. It is a yet faster
R5000. Faster than the RM5200, but from the same company. I'm not entirely sure
that RM7000 is the right name for the CPU.
Greetings,
Frans van Hoesel
> Another alternative for the O2 is to go for a RM7000 CPU. It is a yet faster
> R5000. Faster than the RM5200, but from the same company. I'm not entirely sure
> that RM7000 is the right name for the CPU.
I would LOVE to, but SGI isn't selling those.
-frans
_______________________________________________________________
"Hi, my name is Frans, I'm your crazy scientist for today."
Frans van Hoesel hoe...@chem.rug.nl
http://hpcv100.rc.rug.nl/~hoesel
_______________________________________________________________
Avi Bercovich
>
> Tarjei T. Jensen <tarjei...@kvaerner.com> wrote:
>
> > Another alternative for the O2 is to go for a RM7000 CPU. It is a yet faster
> > R5000. Faster than the RM5200, but from the same company. I'm not entirely sure
> > that RM7000 is the right name for the CPU.
>
> I would LOVE to, but SGI isn't selling those.
yet. I just hope that they'll manage to price the bugger competitively.
I don;t think there is an O2 owner out there that wouldn;t want to
update his/her 180Mhz blue baby with a RM7000. If the upgrade 'll fall
in the 500-800 USD range it'll go like wildfire.
dreams...
avi.
Frans van Hoesel
>> I would LOVE to, but SGI isn't selling those.
> yet. I just hope that they'll manage to price the bugger competitively.
> I don;t think there is an O2 owner out there that wouldn;t want to
> update his/her 180Mhz blue baby with a RM7000. If the upgrade 'll fall
> in the 500-800 USD range it'll go like wildfire.
> dreams...
Remember the R5200 for the O2? at the time it became available
many people in the newsgroups were complaining that it was
too little, too late and too expensive.
I think you have a nice dream, a very nice dream, but it will stay a dream.
I personaly would have expected a version of R7000 somewhere in june
last year. The longer SGI waits, the harder it will be not to fall
into the category too little, too late and too expensive.
G. Douglas
> Another alternative for the O2 is to go for a RM7000 CPU. It is a yet faster
> R5000.
Can you say "vaporware" ????
Zack Hynes
"G. Douglas" <gdou...@reputable.com> wrote in message
news:3A7B8692...@reputable.com...
Yes I can say vaporware...the RM7000 is certainly not vaporware as I have
one.
Neil Rothwell
wrote:
Has your washing machine also got Crime graphics :-)
Frans van Hoesel
> Yes I can say vaporware...the RM7000 is certainly not vaporware as I have
> one.
can you than please tell us when and where you got it, and how much
you paid for it? (asuming you have a RM7000 working in an O2)
G. Douglas
>
> Zack Hynes <ko...@offramp.org> wrote:
>
> > Yes I can say vaporware...the RM7000 is certainly not vaporware as I have
> > one.
Prove it, or go away.
Zack Hynes
so you can resell it for ridiculous prices.
Zack
"G. Douglas" <gdou...@reputable.com> wrote in message
news:3A7CEA66...@reputable.com...
Brian
a O2 with a 180MHz R5K and not your RM7000?
link: http://ticking.clock.org/banana.html
If you really have a RM7000 O2, just put up the hinv.
-Brian
200MHz R5K O2
In article <Dodf6.797$054.2...@news.pacbell.net>, "Zack says...
Stephane Albi
wrote on 5/02/01 3:01:
I don't think the hinv will prove something. Here is mine...
CPU MIPS R15000 Processor Chip Revision: 0.2
FPU MIPS R15000 Floating Point Coprocessor Revision: 0.3
4 750 MHZ IP52 Processor
Main memory size 4096 Mbytes
Secondary unified instruction/data cache size 8192 Kbytes on Processor 0
Instruction cache size 512 Kbytes
Data cache size 512 Kbytes
FLASH PROM version 9.24
Integral SCSI controller 0 Version ADAPTEC 7880
Disk drive unit 1 on FW controller 0
CDROM unit 4 on FW controller 0
Integral FW controller 1 Version ADAPTEC 9980
On-board serial ports tty1
On-board serial ports tty2
On-board usb ports tty3
On-board EPP/ECP parallel port
CRM-IR graphics installed
Integral Giga-Ethernet ec0, version 2
Iris Audio Processor version A4 revision 0
Video MVP unit 0 version 1.9
AV AV1 Card version 1, Camera not connected.
Vice TRE
Stephane Albi
Zack Hynes
look like, but if you did you wouldn't need to see mine. I'm prolly already
in trouble for talking about it anyway, as Alexis Cousein pointed out. I
like my job at SGI and wouldn't want to lost it. :P
My page shows a R5000 O2 because that is what the webserver was. It's
actually been moved into an R5k Indy but I haven't updated the page yet.
Zack
"Stephane Albi" <stepha...@cybercable.fr> wrote in message
news:B6A43AD8.223EE%stepha...@cybercable.fr...
Tarjei T. Jensen
Zack Hynes wrote in message ...
>^5 Yes, hinv won't prove anything unless you know what it is supposed to
>look like, but if you did you wouldn't need to see mine. I'm prolly already
>in trouble for talking about it anyway, as Alexis Cousein pointed out. I
>like my job at SGI and wouldn't want to lost it. :P
It would be silly to deny that such a thing exists. We know that there are
RM7000 CPUs in O2s because we have seen the comments about it in the list of
fixes(6.5.8?) on the SGI web server. We have also seen the CPU it in the
roadmap.
What we don't know is when the CPU will be released and what effect it will
have on the performance of the O2. That will be exciting news even if it can't
compete with the raw power of a P3 or Athlon @ 1GHz.
In short: come out, we know you're in there!
Greetings,
G. Douglas
>
> ^5 Yes, hinv won't prove anything unless you know what it is supposed to
> look like, but if you did you wouldn't need to see mine. I'm prolly already
> in trouble for talking about it anyway, as Alexis Cousein pointed out. I
> like my job at SGI and wouldn't want to lost it. :P
I guess SGI has decided to forego the grammar test, due to
their "hiring blitz".
Did you get any SGI stock as a signing bonus?
Jon Leech
Aside from your misattribution of the quoted comment to Zack Hynes,
the poster thereof is not a native English speaker. Cut some slack and
save the baseless insults for the Yahoo message boards.
Jon
Vedran Hunjek
> If you care to read the original followup from Zack Hynes, you will see
> that you are in error about the misattribution.
>
> Re: the content of my followup, it was an attempt at humor.
> If I offended you or your employer, I apologize. Sometimes I
> forget to put those little smiley faces. Sometimes not.
>
> Greg Douglas
> Reputable Systems
what,
who is masturbating??
going to get my glasses 8OD..
v*