Core2Duo and SW

"core 2 Duo E6400 is represented by an engineering sample, which characteristics may differ a little from the release model..."

lies, damn lies, and benchmarks. we'll see how it shakes out in the real world.

arlan hopeless AMD fanboy

I read Anandtech's article on the Core technology.

Floating point performance is supposed to be stellar. It is really meant for improved server performance though.

Dale Dunn wrote:

Yes looks good but it is hard to know exactly what aspects of SW performance they tested. I see quad core is not all that far away. That would be cool for rendering and using SW at the same time...if I get

4gb :o) I see Intel and AMD are thinking of 32 cores eventually. I wonder if SW will be ever capable of semi using more than two. I really wish SW would say something useful about this topic.

I've been doing some reading at hardforum as well and it strikes me as good time to wait for a couple of months to see what is going to shake out of the tree.

Zander

neil wrote:

Exactly my thoughts but I am figuring I might actually wait about a year. I want get a decent run out of a Vista box. so I figure by then - Vista arrives with maybe sp1 SW 2008 Vista support -more parts multithreaded? hope so... Proper quad core arrives with low watts requirement and faster bus/ram Dual core graphics with DX10 support and HD capable.

"neil" wrote in news:e98vui$drk$ snipped-for-privacy@emma.aioe.org:

The scores given seem to be total times for the SPECapc benchmark, instead of normalised scores. They could have done a better job documenting that, for sure. If I'm right, then we know what they tested. Even if it's not the SPECapc test, it's still the only SW benchmark I could find.

uh well their test also included Wildfire and Matlab so I didn't know if that number actually reflects SW? maybe better or worse...still I would like one - just for improved rendering performance if nothing else :o)

"neil" wrote in news:e994gc$9oo$ snipped-for-privacy@emma.aioe.org:

Oh, I see your concern. If you skim through the article, there is a link to a spreadsheet with all the scores broken out individually.

And the SPEC benchmarks emphasize graphics performance.

Dale Dunn wrote:

i am not 100% on this, but i think they are dropping opengl support in vista. so, that means no more workstation video cards. what do we need workstation cards for? why, swx of course!!!

arlan

ok thanks found it - easy to miss... Those are quite impressive figures really. Looks like big price cuts will be on the way for AMD processors.

I am not really sure if SPEC is a good indicator of performance for multithreaded aspects of SW though is it?

I think the original plan was to support only a nobbled version of OpenGL

1.2 running underneath DX10? as a courtesy with obsolescence in mind but last I heard they pulled back from that because of the objections from game and hardware makers.

The bit I haven't quite worked out is the change in LCD connectors and spec to allow HD video to play.I hear that today's screens now won't play them. Anyone shed light on this?

I can't remember the name of the protocol, but it will be possible for the signal from the connector to the screen to be encrypted. If the flat panel does not support this copy protection protocol, movies will not play at HD resolution. The idea is to prevent pirates from recording the output signal in order to circumvent copy protections on the disc and in the player software. Currntly owned and perfectly good flat panel monitors capable of HD resolution will have to be upgraded. I don't know if Linux will support this. Probably not with an open-source player, but the DeCSS fiasco may be repeated. Apple could support this if they choose to, but I don't know their stance.

Here comes the OT rant:

It has absolutely nothing to do with the resolution and refresh rate capabilities of the display. It's more Digital Rights Management. DRM is the stuff that prevents you from doing what you want with the media you purchase. For example, if you buy music and download it, there is typically a limited number of devices you can play the music on, or a limited number of times you can copy the file to a CD or some such. All in the name of preventing piracy. This is also what "Trusted Computing" is about, if you've heard that phrase.

I'm not for piracy, but I'm against schemes that limit what I can do with the content I legitimately buy, or increase the cost of using it by adding silly DRM requirements. I don't see this particular new nuissance stopping any real pirating. Casual copying isn't done by tapping a VGA cable, and a determined pirate who could intercept the digital signal to one of these new monitors could also take the monitor apart and intercept the unencrypted signal to the glass panel itself. In other words, extra costs for the consumer to play HD media with absolutely no guarantee that professional pirates will be prevented from doing what they do.

Sony has issued a statement that they will not enforce this copy protection when it begins shipping. They did not promise that they will never do this. I suspect that all of the media vendors will take a similar stance, as new flat panels will eventually all support this. Once it has been determined that most user's displays support it, they'll start requiring the encryption. At least, that's how I think they'll play it.

Sony recently achieved some negative fame with rootkits on their music CDs.

Personally, I'll try to avoid copy-protected music if at all possible. first, on principle. Second, because it's inconvenient. I once got a free music download from Wal-Mart's on-line music store. I could only play the track on Windows media player, and I couldn't copy it to any other device. Also, it was not a very high bit-rate file.

That was enough to turn me off of copy-protected music. I have more than one computer around the house I like to play music from, I don't use Windows media player, and my mp3 player doesn't support acquiring licenses to play the music I buy. If I like an album enough to buy it, I'll get the CD. I'll have a hard copy, and I can rip the music to whatever format I want, to be played where I want. For downloading music, I use eMusic.com, which sells music (mostly older stuff and independent labels) as high-bitrate mp3 files with no copy protection. Unfortunately, you can't browse the catalog until you've signed up for at least the free trial membership.

I'm not even against digital copy protections, in principle. For example, I'd like to be able to send models to a customer that they couldn't send out for competitive quotes against us. I'd rather not do this if it prevents the customer using the model for legitimate purposes, though.

As for this silly encrypted display nonsense, I guess I'll just have to bend over.

I think not. That is, I think I agree. AFAIK, the only multi-threaded aspect of 2005 was in the Parasolid Kernel, and even that only for some operations. So, whatever multi-threaded aspect of performance you're looking for, this test won't reveal it. A lot of the reviews use 3DSMax or POVRAY, which support multi-threading quite well. MAybe one of those scores would be useful to you?

thanks..so another reason to hold off buying...encryption upgrades!...$$$$ :o(

• snipped-for-privacy@gmail.com:

That plan has been dropped already. Vista supports OpenGL as did Windowsxp...

Benjamin