Preview DC Optty 866 @ 2.2Ghz

P

Pirox

Weaksauce
Joined
Jul 13, 2003
Messages
71
Could this have been intended to be in the multiprocessing forums?

anyways It's about....

1x 248(1 core) vs 1x 866 (2 cores)
Dual-CPU 248 ( 2 cores) vs 1x 866 (2 cores)
and of coure Dual-CPU 248 (2 cores ) vs 2x 866= Dual dual-cores :) (4 cores)

which you can find Here

and Here..

Main page starts Here

Enjoy... :p

Notice the last scores at the end of the page :

CPU Mutil-Media Test:

Opteron 866 x 1 Opteron 886 x 2
Int :
42,244 84,615
float :
45,524 91,648

A whooping > 100% increase! that's just sick... :oops:


Need i say running dual-cores on a 4-way machine(8 cores) or 8-way (16 cores)? :p
 
Dual core AMD chips are so gonna walk allover dual core p4's. I know intel needs to introduce something to give the illusion of some kind of forward advance in cpu speed but sandwiching 2 cpus with a failed architechture together is just a bandaid on a mortal wound.
 
True, what they really need is new architecture, maybe something of the likes of their Mobile processors.
 
the pictures of intel's dual core really puts things in perspective.. that's one long skinny die :eek:
 
That's strange, in the review, 2 dual-core chips outperformed 1 dual-core chip in a benchmark which isn't multithreaded...
 
if it's running at the same speed, it should..

i have a reasonable idea in my head as to why it's faster, but i can't put it into words.. not yet at least. give me a little bit of time to overcome my lack of eloquence ;)
 
Thats hot. I still cant wait for my dual 2GHz DC procs.... mmmm... 8GHz in 1U. Mmmmm...

Its good to see the chips scale perfectly.
 
Yea, but you can use those extra 5.9GHz to fold for the [H]orde. And those are AMD GHz, not Intel GHz.
 
Hmmm... That is going to rip the XEON market wide open. Even Dell will have to think twice now. In all systems where calculations are of an issue and where the software is multithreading these CPUs will ROCK.

And iirc HP has allready opened for upgrading the old Opterons with DC...
 
(cf)Eclipse said:
hehe, too bad in things like gaming, 5.9 of those "8ghz" will be wasted ;) :D
Click to expand...

it wont be that much longer before game developers can no longer ignore multithreaded apps
 
Oh they would like to do it today allready. BUT the problems with programing doesnt increase proportionally to the number of threads but geometrically to the number of threads. Therefore i thing there has to be a lot better tools out there b4 we see any multithreaded "real" games. (Not to mention that the market is still a couple of years away).
 
Unreal Engine 3 is multithreaded, so when that comes out and shows up in games, the DC chips will shine. Unreal looks too amazing to pass up on, and with a 2cpu DC platform behind it and a dual next-gen graphics card, it should run just fine.. :D
 
serbiaNem said:
That's strange, in the review, 2 dual-core chips outperformed 1 dual-core chip in a benchmark which isn't multithreaded...
Click to expand...

IO, etc were offloaded, so it makes sense. But to make sense out of your post, which score are you talking about?
 
(cf)Eclipse said:
i think the 2x866 vs 1x866
Click to expand...

Well, obviously, but I was wondering which bench specifically he was wondering about. It really doesn't matter which, though. Any IO or interrupt offloading is going to speed up the main thread a tiny bit.
 
Mostly the winbench benchamarks. What bothered me was that 2DC chips outperformed 1DC chip when 1 Thread was being tested. It didn't make sense that one core couldn't handle the same amount of background info as could 3. Since the performance of the first core (running the benches) shouldn't be too heavily influenced in either case.

Edit: On another note, it looks as though a DC AMD chip running at the same frequency may increase game performance as opposed to the Intel DC chip game performance decrease. :D
 
serbiaNem said:
Mostly the winbench benchamarks. What bothered me was that 2DC chips outperformed 1DC chip when 1 Thread was being tested. It didn't make sense that one core couldn't handle the same amount of background info as could 3. Since the performance of the first core (running the benches) shouldn't be too heavily influenced in either case.
Click to expand...

The only thing this means is that the benchmarks are not entirely CPU dependent. The CPU dependent bench, SuperPi, was precisely the same on the 2xDC and the 1xDC.

The other Benches have more IO and interrupts, that's all. There's perhaps a 1% difference between the 2x DC and the 1x DC. It's not a big deal, and entirely within the realm of possibility. I'm guessing more interrupts than IO.
 
serbiaNem said:
Unreal Engine 3 is multithreaded, so when that comes out and shows up in games, the DC chips will shine. Unreal looks too amazing to pass up on, and with a 2cpu DC platform behind it and a dual next-gen graphics card, it should run just fine.. :D
Click to expand...

Cool...

I cant wait to make my video editing MACHINE with dual core opties
 
Can someone give me a breakdown of all of this? or some pics and stuff? I cant go to the linked sites here because the marine corps blocks alta vista for whatever reason so I cant access anything.... thanks
 
You must log in or register to reply here.
Back
Top