S5R2

Akos Fekete
Akos Fekete
Joined: 13 Nov 05
Posts: 561
Credit: 4527270
RAC: 0

RE: Well Gary, "Barcy"

Message 62311 in response to message 62309

Quote:
Well Gary, "Barcy" should be a "Barcelona", which is afaik a new quad core CPU. Dunno any more about it, I'm afraid, and I have no idea what exactly an "Agena" is (except for Atlas-Agena which would be wildly off topic...)


Agena would be the desktop processor version of Barcelona. ( AMD roadmap )

Brian Silvers
Brian Silvers
Joined: 26 Aug 05
Posts: 772
Credit: 282700
RAC: 0

RE: but some excessive

Message 62312 in response to message 62306

Quote:

but some excessive huckstering and pontificating on another forum reminded me how much I hate Intel arrogance (even if they do make good stuff)!

I think I know who you're talking about... ;-)

ExtraTerrestrial Apes
ExtraTerrestria...
Joined: 10 Nov 04
Posts: 770
Credit: 578456871
RAC: 195567

Sorry if this was asked

Sorry if this was asked before, but I haven't been following the E@H forum lately.

Akos, what about the S5R2 hot loop and SSE2? I know SSE1 and 3DNow brought a huge speed increase to the former applications, so the code seems to be fine for vectorization. These two work on 64 Bit, whereas you can do 128 Bits with SSE2. There's just not much benefit in doing so with P3/4/M and AXP/64. However, Core 2 D/Q and K10 would benefit greatly from such a codepath as they can process 128 Bit vectors in one cycle (OK, send and retire one such command per clock per unit :) instead of two cycles for the former CPUs. I see great potential for optimizations here, which is hardly used, if at all, by any BOINC app today (or did I miss some fancy new project?)

Regards, MrS

Scanning for our furry friends since Jan 2002

F. Prefect
F. Prefect
Joined: 7 Nov 05
Posts: 135
Credit: 1016868
RAC: 0

RE: Sorry if this was asked

Message 62314 in response to message 62313

Quote:

Sorry if this was asked before, but I haven't been following the E@H forum lately.

Akos, what about the S5R2 hot loop and SSE2? I know SSE1 and 3DNow brought a huge speed increase to the former applications, so the code seems to be fine for vectorization. These two work on 64 Bit, whereas you can do 128 Bits with SSE2. There's just not much benefit in doing so with P3/4/M and AXP/64. However, Core 2 D/Q and K10 would benefit greatly from such a codepath as they can process 128 Bit vectors in one cycle (OK, send and retire one such command per clock per unit :) instead of two cycles for the former CPUs. I see great potential for optimizations here, which is hardly used, if at all, by any BOINC app today (or did I miss some fancy new project?)

Regards, MrS

Right over my head, but helpful and interesting nonetheless. I have 1 single core 64bit 3400+, but the rest are all Athlon XPs from 1.1 to I believe 1.8. Are they going to be worth spit considering the codepath??

Gary

In the beginning the Universe was created. This has made a lot of people very angry and been widely regarded as a bad move.....Douglas Adams

Donald A. Tevault
Donald A. Tevault
Joined: 17 Feb 06
Posts: 439
Credit: 73516529
RAC: 0

Well folks, for awhile it

Well folks, for awhile it looked like I was going to avoid the seeming curse of the Athlon64 processors with the new 'R2's. Everything was going fine until just over a half-hour ago. Now, two 'R2 units on my new 6000+ machine have prematurely ended with the infamous "compute error" messages. On my 3500+, my P-III's and both of the pre-64-bit Athlon-class machines, they're doing just fine, though. (I haven't received any 'R2's for my old Pentium II machine yet. It's currently crunching on a pair of 'RI units that won't finish until tomorrow night.)

I have to wonder. . .

Hopefully, the program developers are able to gather enough debug data from us to get this straightened out.

P.S. As much as I usually like to blame Windows for things, this time I can't. The 6000+ machine is running Kubuntu Linux.

F. Prefect
F. Prefect
Joined: 7 Nov 05
Posts: 135
Credit: 1016868
RAC: 0

RE: RE: Well Gary,

Message 62316 in response to message 62311

Quote:
Quote:
Well Gary, "Barcy" should be a "Barcelona", which is afaik a new quad core CPU. Dunno any more about it, I'm afraid, and I have no idea what exactly an "Agena" is (except for Atlas-Agena which would be wildly off topic...)

I don't know if it's off topic or not. I may need a rocket for a processor to run some of this new stuff.

Agena would be the desktop processor version of Barcelona. ( AMD roadmap )


In the beginning the Universe was created. This has made a lot of people very angry and been widely regarded as a bad move.....Douglas Adams

Annika
Annika
Joined: 8 Aug 06
Posts: 720
Credit: 494410
RAC: 0

lol That sounds good. As long

lol That sounds good. As long as it helps... You might have to be careful about heat, though ;-) those rockets do tend to get a little warm...

F. Prefect
F. Prefect
Joined: 7 Nov 05
Posts: 135
Credit: 1016868
RAC: 0

RE: lol That sounds good.

Message 62318 in response to message 62317

Quote:
lol That sounds good. As long as it helps... You might have to be careful about heat, though ;-) those rockets do tend to get a little warm...

Not to worry. Have a piece of that heat dissapating material used on the shuttle glued right on the CPU. Never fails. Gets red hot and ya still touch it with your finger. Ouch. Sometimes.

Gary

In the beginning the Universe was created. This has made a lot of people very angry and been widely regarded as a bad move.....Douglas Adams

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6588
Credit: 317913559
RAC: 390421

RE: RE: lol That sounds

Message 62319 in response to message 62318

Quote:
Quote:
lol That sounds good. As long as it helps... You might have to be careful about heat, though ;-) those rockets do tend to get a little warm...

Not to worry. Have a piece of that heat dissapating material used on the shuttle glued right on the CPU. Never fails. Gets red hot and ya still touch it with your finger. Ouch. Sometimes.

Gary


[aside]
I once did an order of magnitude calculation on that, with some published dynamic data on a Shuttle re-entry. I got about 5 MegaWatts of heat dissipation per square meter of heat tile area. That is sustained for about twenty minutes during those long swooping S-curves over the Pacific before a US continental landing. Man what a substance that material is !!
[/aside]

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Dave Mickey
Dave Mickey
Joined: 3 Mar 06
Posts: 3
Credit: 264525
RAC: 0

I hope they do something

I hope they do something about the ratio of runtime to
deadline to get it back to close to where it used to be.
Besides old, slow machines having a problem with that, so
do newer machines that run a low resource allocation to e@h.
With BIG units coming in, BOINC has to prioritize them
up, with the result of not honoring my chosen resource
shares, and cutting heavily into the time allocated to
my primary project, s@h.

I'd like to keep e@h as a backup project, but given the current
runtimes/deadlines, I won't be able to.

Dave

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.