[url]http://forumz.tomshardware.com/hardware/AM2-newegg-ftopict186050.html[/url]
am2s are out

The Conroe (Core 2) Extreme specs and price have been reduced.

The Core 2 Duo E6800 (2.93 Ghz) 4MB L2, 1066 FSB
will become the Extreme version
Core 2 Extreme X6800 (2.93 GHz) 4MB L2, 1066 FSB $999

So the Core 2 systems will use a maximum of 1066 FSB,
only the Woodcrest (Xeon DP 51xx) systems will use 1333 FSB.

AMD announced a new desktop platform, 4x4.

Two system makers will deliver systems in the 2nd half of 2006.

It will use two AM2 sockets, each with a dual core X2, for a total of 4 cores.
The sockets will be next to each other with a HT link connecting the
two sockets.

It will use regular DIMMS not ECC.

Each dual core chip will have its own memory though the other chip can
also access it.

[QUOTE=dsouza123]AMD announced a new desktop platform, 4x4.

Two system makers will deliver systems in the 2nd half of 2006.

It will use two AM2 sockets, each with a dual core X2, for a total of 4 cores.
The sockets will be next to each other with a HT link connecting the
two sockets.

It will use regular DIMMS not ECC.

Each dual core chip will have its own memory though the other chip can
also access it.[/QUOTE]
With four cores, there will be a lot of factors to worry about. Worrying about how to turn them in, that is. :razz:

Compared to a regular single core Athlon64 running 32-bit Windows XP Pro
a 4x4 running 64-bit Windows XP Pro should have about 8x the TF throughput.

4 cores * (2x Trial Factoring performance in 64 bit mode) = (8x TF work)

A true Trial Factoring monster !

4x4+
 

[QUOTE=dsouza123]Compared to a regular single core Athlon64 running 32-bit Windows XP Pro
a 4x4 running 64-bit Windows XP Pro should have about 8x the TF throughput.

4 cores * (2x Trial Factoring performance in 64 bit mode) = (8x TF work)

A true Trial Factoring monster ![/QUOTE]

Of course Intel are working on quad core chips.

But then AMD can counter with 4x4+ in 2007 which is quadcore PLUS quadcore = 8 cores.

That will be the real TF monster!

Just to note to use 4x4 it may insist on FX2 processor.

This could be that those chips could have TWO hypertransport links, one to the other processor, one to the chipset. Therefore I don't think you can use 4x4 operation mode with just ANY AM2 processor.

IMO, move to 4x4 shows one thing: this year AMD in order to beat Intel's dualcore Conroe will need 4 cores... which only makes sense if 4x4 will cost roghly the same as a single Dualcore Conroe, which I seriously doubt :no:
What will be really interesting to see from AMD is Torrenza, HTX, and of course next-gen K8L CPU :smile:

Possibly best summary I've seen on conroe:

[url]http://forumz.tomshardware.com/hardware/Collection-Conroe-Data-Core-Duo-Core-Extreme-ftopict183765.html[/url]

SuperPI 1M down to 10.140secs, with conroe@5GHz. No wonder AMD is coming out with the 4x4. Things are looking mighty interesting.

Anyone see a benchmark of prime95 on conroe yet?

-- Craig

here's one - looks interesting :)

[url]http://img152.imageshack.us/my.php?image=2400prime5cg.png[/url]

[QUOTE=nucleon]
Anyone see a benchmark of prime95 on conroe yet?
[/QUOTE]

kind off...

1024k FFT 15min test time:

Dualcore Opteron 170 clocked @3.2GHz:
[url]http://img153.imageshack.us/my.php?image=nohsprime320015044gz.png[/url]

lower bound: 5*4000 iterations/(15m*60) = 22.2 iterations per second <=> 45ms per iteration
upper bound: 6*4000 iterations/(15m*60) = 26.7 iterations per second <=> 37.5ms per iteration

[url]http://img242.imageshack.us/my.php?image=prime271313250sz.png[/url]
Conroe E6600 clocked @2.7GHz
lower bound: 9*4000 iterations/(15m*60) = 40.0 iterations per second <=> 25ms per iteration
upper bound: 10*4000 iterations/(15m*60) = 44.4 iterations per second <=> 22.5ms per iteration

[QUOTE=kaeptn_kork]here's one - looks interesting :)

[url]http://img152.imageshack.us/my.php?image=2400prime5cg.png[/url][/QUOTE]

nice...
Now I'm interessted how much the performance drops if 2 clients are running... Conroe has a shared L2-Cache.