Core 2 will be a gaming revelation. (real data in thread)

The most impressive fact is that it's probably using exactly 1/2 th power 8O

Tis looking towards a very good holiday season indeed... Happy Holidays!

My worry is now that Intel has put too much focus on gaming performance. I can understand their efforts in this area since they have been lagging behind for quite a while, but it appears they've let their multitasking performance slip. That's kind of ironic considering they're the ones pushing so hard with dual core processors at low price points. I hope the bit-tech data on weak multitasking performance is corrected by the time the product finally chips. Their data is a month old anyways so we'll have to see what new stuff comes out when the NDA lifts.

The Register has also released their preview.

http://www.reghardware.co.uk/2006/ [...] ore_2_duo/

And they mention a Kentsfield pushing PowerPoint slides. How nice.

Even less than that as unlike the Netbursed based designs, the measured power doesn't exceed TDP. Which probably has to do with excellent power management features(EIST, which is not same as Pentium D's EIST).

Wasnt Prescott a gaming revolution also? Isnt that was the Intel ad said. Lets hope its something this time.

Hey the Acton_Man became a Action'o saurus "Roar!!! . Lols what did you get some engineering sample Core 2 DuoEE or 4 X 4 Action last night dude?? Haha '4 X 4 Action' Just coined a new term for computers and just repeated a old one for 4wheel drive '4 X 4 Action'

Whatever you're taking, stop taking them.

Actually my post should be a lesson to some people (fanboys, and people that can be decieved by a bunch of numbers that can be speculated easily). Oh by the way I actually over-estimated you and expected a better reply than "The rest of your post is boring." I wont post anything in reply to you. I'm so sorry to take your time.

Nah. Too much reasoning and no intel bashing.

Just stop posting about Core 2 chips already, this makes me sad for the fact that my current system would be nothing compare to it and makes me all excited inside knowing that my upgrade this Christmas would make my pc a gaming monster, thus giving me some mixed emotions.

8)

judging by intel's past stuff, it would be reasonable to question their statements and benchmarks, but not since they went israel

Isnt anyone considering that the PEE 965 has 4 logical units? Hyperthreading does actually work, anyone who has used a HT and non-HT intel chip can atest to this.

So i for one am not suprised by the apparent similarity in multi threaded beanchmarks.

think of it as comparing a GPU with 8 pipelines vs one with 16, even if the 8 are individually faster than the 16, the 16 will better cope with data that exceeds the envelope ideal for 8.

What do u guys think?

This is a good detail to bring to the table, effectively the comparison of multi-thread/task (there is a difference ) results at first glance are disappointing, the comparision was made to an 965 EE which has HT, so 2 physical cores but 4 logical cores crunching on the same thread.

Add on top of that PC Mark05 uses CPUID to loadup optimized compiles for various CPU's to report a score, and now you have a bench optimized for P4 and it's multicore environment up against a Core 2 duo which has not been recompiled. It does not surprise me that we would see such a descrepancy, but in test 2 of 3 it is disconcerting -- in this case, I agree with XtrPsy (what does that handle stand for ?>???) more data would be useful, and a compare to other non-HT dual core would be preferable.

It is indeed a concern, but then again, Netburst does have its merits in applications that are sensitive to clockspeed, but is decompression one of them? i will try to find a comparison between a Pentium D and AMD X2 in file decompression and post it here.

We are taking preliminary benchmarks way too seriously. Here's a reason why PCMark 05 is a poor real life benchmark. If I see anything common with benchmarks of Core 2 Duo, its that its OK in synthetics but EXCELLENT in real life benchmarks.

One reason preliminary benchmarks is bad that not because they are rigged, but they benchmark such a few applications to be taken notice of.

Hyperthreading does bring benefits. People got disappointed with hyperthreading as soon as the Pentium Extreme with Hyperthreading came out. Which was a flawed viewpoint as Pentium Extreme didn't get much with hyperthreading, while Pentium 4 did.

Look here with Presler EE tests with HT: http://www.anandtech.com/cpuchipse [...] i=2658&p=7

Look at Smithfield EE tests: http://www.anandtech.com/showdoc.aspx?i=2410&p=3

First CPU with Hyperthreading: http://www.anandtech.com/showdoc.aspx?i=1746&p=6

Yes, I guess that has come to mind to those "HyperThreading" aware. Nevertheless, it's the difference margin in multi-threading which is a bit worrysome.
I suppose JumpingJack focused on two fundamental issues: Multi-tasking is not multi-threading & compiler optimization.
Apparently, Tim Smalley (from bit-tech.net) mixes up both terms and, where he mentions "multi-tasking", the boards show "multi-threaded" (which I believe is what is correct).
Maybe someone else would care to elaborate on multi-threading vs multi-tasking, on a deeper level; anyway, with a wider & [much] more sophisticated microarchitecture, being able to process more IPC than any NetBurst chip, multi-threading shouldn't have been an issue with Core; not because it will be much needed in Conroe (...) but, because it should shine on the server space, with Woodcrest.
Not wanting to jump into any preposterous conclusions (and, as David1C referred, these are preliminary benches), Woodcrest should be Intel's major concern, on what regards multi-threading performance.
Certainly, the comparison was made with a GHz above 2c Presler w/HT...


Cheers!