I almost forgot my second vintage computer: An Athlon 800, sitting on an EP-8KTA2 motherboard.
I was foolish enough to try to upgrade this veteran with a spare Athlon XP 1800+ CPU, but no go. It did start up but would not run very long, later I read that even with BIOS upgrades, there is no way to make XPs run on that board, so I stopped trying.
I feel terribly un-ecological for using this host and may well stop doing so in the future. Those early Athlons don't even support SSE. My Pentium III should be really useful once we get optimized apps with SSE, but I don't think an MMX optimized apps are worth the hassle to write them. It does support 3DNow!, tho. Hmmm...Akos, any plans :-)?
I have to differ with you regarding MMX. When Akos made them for S4 it made a big difference on the K6's, and I would guess there are gains to be made for Athlon Classics as well.
I don't recall whether he did any 3DNow! for them, though.
My PII Deschutes is crunching happily a 30 hours SETI WU, a 100 hours QMC WU and a 200 hours Einstein WU using App. 4.31. Must be a monster WU. BOINC is alternating between the three according to share (Einstein 300, SETI and QMC 100). Deadlines are July 31 for QMC, August 8 for SETI and August 9 for Einstein. Crossing fingers.
Tullio
Well, assuming the estimates are in the ballpark it should make it.
I would imagine it will have to EDF QMC at some point since that looks like the tightest deadline result you have to deal with.
Alinator
They did not start at the same date. QMC is already 91% completed, SETI 48%, Einstein 6%. SETI has the easiest deadlines. QMC's quorum is 1, so I think I can make it.
Tullio
I have to differ with you regarding MMX. When Akos made them for S4 it made a big difference on the K6's, and I would guess there are gains to be made for Athlon Classics as well.
I don't recall whether he did any 3DNow! for them, though.
Alinator
MMX uses integers only, right? Would be a real pain to rewrite the code in this way, and I have no doubts that Akos can do that, it's just a question whether this is worth the effort for a very limited numbers of CPUs that would benefit from this.
I have to differ with you regarding MMX. When Akos made them for S4 it made a big difference on the K6's, and I would guess there are gains to be made for Athlon Classics as well.
I don't recall whether he did any 3DNow! for them, though.
Alinator
MMX uses integers only, right? Would be a real pain to rewrite the code in this way, and I have no doubts that Akos can do that, it's just a question whether this is worth the effort for a very limited numbers of CPUs that would benefit from this.
CU
BRM
Hmmm... Good point.
OTOH, since my K6's got a good boost well, and they don't do 3DNow!, Akos found something in there for them.
The really amazing thing was when Akos first started poking around to squeeze more performance out of the app, AFAIK it was a 'blackbox' from his POV. To be able to do that is a sure sign of someone who knows their stuff cold! ;-)
However, I'm positive that I still have the original PIII 500Mhz Slot 1 that it originally came with. If Tullio wants it, I'll be glad to drop it in a post box and let it take a slow boat to Italy! I'll be very jealous, of course, since I've not had a chance to return there myself since mid-2000. :(
Thank you very much, but I am not a hardware guru. I prefer downloading some software source program and compile it on my Linux box. My PII is good enough to do such things and watch NASA streaming TV. Speaking of nostalgia, once I crossed Kansas on a Greyhound bus going from St.Louis, where I was a student, to Los Angeles, where I have some relatives, and back. It took me 60 hours to go there on old Route 66 and another 60 to come back. Happy days!
Tullio
First of all, my condolences for having to travel across western Kansas and eastern Colorado in a bus! I hope you brought something good to read! ;) Ironically, I've taken this long to respond because I was on business in Colorado Springs for the last 2 weeks.
Quote:
The only problem he might have would be if the heatsink is not of the appropriate size and type. It's actually relatively easy to swap heatsinks in a lot of cases though.
Tullio, I think you're worrying too much about the "hardware" end of things. This should be as simple as unplugging the old processor, plugging in the new one and booting back up. I honestly can't remember if SSE will have to be compiled into the linux kernel or not, but unless you really tweaked your kernel, it's probably in there already. WORST CASE is that you'd have to flash the BIOS on your motherboard, which is as simple as formatting a bootable floppy, putting the new BIOS from the manufacturer on it and following a couple of instructions to make it complete.
P.S. If Gary asks, I might just get those 600's 'topless' and take some risque pictures of that pair for him! ;)
I almost forgot my second
)
I almost forgot my second vintage computer: An Athlon 800, sitting on an EP-8KTA2 motherboard.
I was foolish enough to try to upgrade this veteran with a spare Athlon XP 1800+ CPU, but no go. It did start up but would not run very long, later I read that even with BIOS upgrades, there is no way to make XPs run on that board, so I stopped trying.
I feel terribly un-ecological for using this host and may well stop doing so in the future. Those early Athlons don't even support SSE. My Pentium III should be really useful once we get optimized apps with SSE, but I don't think an MMX optimized apps are worth the hassle to write them. It does support 3DNow!, tho. Hmmm...Akos, any plans :-)?
CU
BRM
I have to differ with you
)
I have to differ with you regarding MMX. When Akos made them for S4 it made a big difference on the K6's, and I would guess there are gains to be made for Athlon Classics as well.
I don't recall whether he did any 3DNow! for them, though.
Alinator
He did. They were almost as
)
He did. They were almost as fast as the SSE one.
Then you're really interested in a subject, there is no way to avoid it. You have to read the Manual.
My PII Deschutes is crunching
)
My PII Deschutes is crunching happily a 30 hours SETI WU, a 100 hours QMC WU and a 200 hours Einstein WU using App. 4.31. Must be a monster WU. BOINC is alternating between the three according to share (Einstein 300, SETI and QMC 100). Deadlines are July 31 for QMC, August 8 for SETI and August 9 for Einstein. Crossing fingers.
Tullio
Well, assuming the estimates
)
Well, assuming the estimates are in the ballpark it should make it.
I would imagine it will have to EDF QMC at some point since that looks like the tightest deadline result you have to deal with.
Alinator
RE: Well, assuming the
)
They did not start at the same date. QMC is already 91% completed, SETI 48%, Einstein 6%. SETI has the easiest deadlines. QMC's quorum is 1, so I think I can make it.
Tullio
Any optimization would be
)
Any optimization would be great.Whole compsquad is waiting.
3x Pentium I(may be more);Pentium II;K6 and Pentium III celeron(not sure with that).
MMX had problem with performance because CPU had to switch to MMX part core,so it couldn't do at the same time non-MMX and MMX code.
RE: I have to differ with
)
MMX uses integers only, right? Would be a real pain to rewrite the code in this way, and I have no doubts that Akos can do that, it's just a question whether this is worth the effort for a very limited numbers of CPUs that would benefit from this.
CU
BRM
RE: RE: I have to differ
)
Hmmm... Good point.
OTOH, since my K6's got a good boost well, and they don't do 3DNow!, Akos found something in there for them.
The really amazing thing was when Akos first started poking around to squeeze more performance out of the app, AFAIK it was a 'blackbox' from his POV. To be able to do that is a sure sign of someone who knows their stuff cold! ;-)
Alinator
RE: RE: However, I'm
)
First of all, my condolences for having to travel across western Kansas and eastern Colorado in a bus! I hope you brought something good to read! ;) Ironically, I've taken this long to respond because I was on business in Colorado Springs for the last 2 weeks.
Well, a picture is worth a thousand words, so take a look at the entire CPU assembly and , the label showing specs.
Tullio, I think you're worrying too much about the "hardware" end of things. This should be as simple as unplugging the old processor, plugging in the new one and booting back up. I honestly can't remember if SSE will have to be compiled into the linux kernel or not, but unless you really tweaked your kernel, it's probably in there already. WORST CASE is that you'd have to flash the BIOS on your motherboard, which is as simple as formatting a bootable floppy, putting the new BIOS from the manufacturer on it and following a couple of instructions to make it complete.
P.S. If Gary asks, I might just get those 600's 'topless' and take some risque pictures of that pair for him! ;)