@wawa

not cool man!

_________________

@AmigaBlitter

Quote:

those two... really interesting...

These two... even better

Intel Celeron (P6) 1000 MHz 889.00
...
Freescale MPC5200B 400 MHz 1121.08

Some Coremark scores are simply weird.

@OlafS25

Checked the price of an i7-3930K lately? It's not entirely "gratis" either, you know.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@wawa

Quote:

wawa wrote: @ppc_addon is it you? the avatar seems to fit.

it's me

_________________
My Blog: http://linuxpowerpc.blogspot.com/
***
Linux PPC user!
IBM Intellistation POWER 275 - CRUX PPC (64bit) 3.0

@KimmoK

I just noticed something. The XEON E5 there is a dual-socket motherboard, so that's sort of cheating. The POWER7 is odd in that it's listed with 64 phtreads, but it's only scoring 32 times the score of the single-thread run elsewhere in the official list, so it seems to be figures in the right ballpark for a single 8-core (4SMT per core) chip.

All in all, quite impressive numbers from the T4240, IFF you have the software to utilise this sort of computing power. Quite a lot of things CAN be parallelled to this extent, but sadly enough games is one of the things that usually struggles. Allthough with the next generation consoles getting more generic cores SHOULD make game companies invest time and effort into parallell processing of games also. Something both Intel and AMD should be encouraging, given how per-core-performance is getting harder and harder to improve.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@olegil

One nice use case for multicore chip is video encoding.
My dualcore AMD3800+ managed to generate mp4 in 150fps speed (IIRC).
Now my 4 core FX4300 does almost 600fps.
(single core speeds are around 75fps and 150fps for each, so multicore advantage is huge)


I wish we had clustering aids built in AOS4, because IMO clustering is not easy enough to set up in mainstream.
It would be cool to be able to just to download "amigaclusternode" -app from x86 repository to enable my SAM to get access to networked x86 CPUs (autoconfig, kind of).

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@KimmoK

I would think the only thing worse than aids is clustering aids

But on a more serious note, I would think the REXX ports would be a good starting point for clustering (and before you say it, yes I advocate giving my PC REXX ports for remote control and scripting of my software). If you then have a remote-login REXX agent (sitting between the remote master and the local application ports) then bob should pretty much be your uncle.

Given enough high bandwidth links to a central (or distributed) data cache, obviously.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@KimmoK

Thinking a bit longer: Why does mp4 encoding scale so good? Does decoding also scale this good, or is there some operation done here between frames and blocks that is inherently easier to multiprocess only in one direction? To me it would seem that all the operations in video compression/decompression are either local within a macroblock, space-related or time-related. And I can't for the life of me see why analysing for similarities across space or time would be more easily "multiprocessable" than the reverse operation.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.