The Intel Powermac / Powermac Conroe / Mac Pro thread - Page 8

And you're not.

Yes I think Intel is going to have to swallow some bitter medicine and clear out the old chips at discount prices. Came across a link that shows that Intel expects 40% of the desktop chips they ship in Q1 07 to be Conroe. That looks like a pretty aggressive roll out of Conroe. I think Apple will be on that bandwagon as well. While I think Woodcrest will make an excellent high end powermac, I also think you will see Conroe on the lower end powermacs. Either both chips(Conroe and Woodcrest) are good or neither are good. Link below.

http://www.dailytech.com/article.aspx?newsid=1820

Decent read, but #1 it's totally invalid, and #2 it has nothing to do with the Nvidia SDI series.

Your wondering why I say it's invalid. The issue is with the timing on the article, what they scrutinized, and it's relevancy today, It was published in January and Nvidia did a huge update for those video cards in March to add most of the features that were talked about in the article that there weren't even drivers for when they did it. The card they used wasn't even supported, and neither were most of the areas the tests focused on until March. Talk about getting ahead of your self.

Here is a list of some of the added features to the card in the March driver update.

• WHQL Certified
• Adds support for GeForce 7900 GTX, GeForce 7900 GT, and GeForce 7600 GT
• NVIDIA PureVideo features and enhancements.
• Support for high definition H.264 hardware decode acceleration on GeForce 6 and 7 series GPUs.
• Support for high definition MPEG-2 inverse telecine.
• Support for high definition MPEG-2 spatial temporal de-interlacing.
• Adds mixed vendor support for NVIDIA SLI.
• TV-Out/HD-out support for NVIDIA SLI.
• Added support for VSync on Direct3D games when running NVIDIA SLI.
• Microsoft® DirectX® 9.0c and OpenGL® 2.0 support

Basically that whole write up focused on things that the card wasn't even running until March. You have to remember that anything you read on the web you have to verify for yourself, because people start listening to these clowns because they are writing on an "Extreme Tech" site; which really doesn't mean shit. If he was a good reporter #1 he should have done his research, and contacted both Nvidia, and ATI before doing the tests so they could have given him a heads up on what would, and wouldn't work using the setup he was using, and why. They (ATI, and Nvidia) would have also advised him on how to fix the problem, and in Nvidia's case they could have informed him the card he was using wasn't supported by drivers yet for 90% of the tests he was doing. They may have decided to give him a beta driver just so it would run.

And melgross, if that article is what you originally read to get you thinking that ATI was far and beyond Nvidia as the video king of all things I can see why you thought that.
If I hadn't remembered what a friend of mine in Florida (that makes TV commercials, and other related video productions) told me about how impressive his new Nvidia card got in just one driver update I probably wouldn't have thought twice about that article, and took it as total truth.

I think that it was Apple or IBM that invented vaporware, but I think that all of these Intel projects have been announced to Wall Street. That should be indicator enought that they are not vaporware, or they risk an FTC investigation. I doubt that Intel is that hard up, I think that the fact that the launch dead lines keep moving closer, is another sign that they are not vaporware. Probably wishful thinking on your friends part, I would probably do the same if I had that much invested in AMD. I also seem to remember that Steve said that looking at late '06 and especially '07 intel would have no competition in performance per watt, all indicators are that is it on or ahead of schedule. Good for us, I hope to buy an Intel designed MacDesk Pro.

You blather pretty well yourself.I've had this bookmarked, so I'll link it. If you're not happy with it, you can find more.

http://www.extremetech.com/article2/...1916993,00.asp

As far as I'm concerned, though, we've been going on about this long enough.

That's funny. I didn'r realise that you posted the same article, or I wouldn't have bothered.

Some of what they did there will have an effect on the video, quite true, but both companies have improved their drivers since then. Both cards were the latest cards shipping when the tests were done, with the latest drivers.

There isn't much point to test with a driver update that isn't available at the time of the test, because, as you surely know, these tests are for boards, and drivers that are actually available.

ExtemeTech happens to be a well respected site, even though you were not happy with their conclusions

I don't know your "friend". My friends do the same work, as did I. We all have our preferences.

I think we can lay this to rest.

Neither Apple nor IBM invented vaporware. That term was in use for a long time, and didn't refer to IBM

Don't forget that they are going to highlight things that go faster. And the reason they go faster may or may not be the processor (e.g. the chipset and memory). There are tests floating around that demonstrate that the iMac G5 outperforms the iMac Core Duo in some situations. The processors are more evenly balanced that Apple's advertising department would have you believe. The Core Duo's main advantages (in the iMac) are dual core, more efficient integer calculations, and a better chipset. Remember that Apple may have revved their PPC chipset by now.


This is absolutely the case. They need to demonstrate a clear improvement over the last G5-based towers. The chips that do that are coming soon and will bring x86-64, HT, and MP support with them. These are going to be some sweet machines.

Programmer,

While "Don't forget that..." is absolutely true, the presenter was actually saying that, although Apple likes some of the benchmark tests, users will notice (and are concerned about) changes in the performance of *applications* they use. (I couldn't agree more.) He had simply done the Handbrake comparison on his own to see how an extremely CPU intensive process performed. (If you have not run Handbrake take a look at it and you will see just how it can soak up processor power.) It really was not intended as a "scientific" test result. I found it interesting as my poor little G4 takes a lot longer than that.

Slightly off topic, but one very neat thing that was also shown was an emulator that will run (only on the Intel machines apparently) called Parallels which seemed to run Win XP rather well in the few little things that were demo'd. (It was simply shown as one possible solution to the VPC situation with the MacTels.)

Cheers8)

But there seems to be more thought that we will see the Woodcrest in at least one Tower.

It's funny, after reading nothing but Conroe, Conroe, we are now reading Woodcrest.

Hmm! Would be strange if it turned out that I started the whole thing in January.

http://computerworld.co.nz/news.nsf/...25714F000230BC

Repeated for the 60th time in this thread. . Woodcrest will be the only processor available at the time of the new PowerMacs release. It is also the only workstation processor intel will have at that time. The Conroe Is a desktop Processor. The Apple desktop is the iMac. Apple insists the PowerMac is the Apple workstation. Any questions?

Thank You!

I wish everyone else would get this

I am thinking a monster workstation demoed at WWDC 2006

Dual quad-core CPUs
Quad SLI capabilities

Mmm

Woodcrest is not a Quad core CPU. That's Clovertown, and it wont be available until early 2007.

At some point you lost your mind. Quad SLI exits, but it's not going to be in a Mac. Apple doesn't even have SLI capability yet; let alone upgrading to Quad SLI.

Didn't I see some article around recently about Intel demoing a unit with dual quad-cores in it?!?

As for quad SLI, never say never!

Could be secret Raycer-derived technologies!

(uh, if you do see my mind, could you send it back this way?!?)

;^p

There will be a Mac with Conroe in it. It will not be the iMac. Maybe a new line will be developed, maybe the low end powermac.

Yes, that was the Clovertown. It's due for next year though. They (intel) have a lot of urgent supply to meet from about 1000 computer vendors. They said they would be produced in this order. Woodcrest, Conroe, Merom, and I think they will stick to the order, but I think they will be released much closer together than earlier anticipated.


linky

'Santa Clara (CA) - Faced with increasing competitive pressure from AMD, Intel apparently has accelerated its launch schedule of its new Core microarchitecture. While Woodcrest, which will carry the Xeon 5100-series designation, was always scheduled for a Q3 introduction, chief executive Paul Otellini on Wednesday confirmed that the company will also be announcing the desktop processor "Conroe" (Core Duo E4000 and E6000 series) as well as the mobile CPU "Merom" (Core Duo T5000 and T7000 series) in the third quarter. All three variants will be shipping "in volume."'

And that's nothing. It's the Tigerton that comes after the Kentsfield that is the real kicker. This is where I hope Apple makes a break away, and makes one separate Mac pro design that uses this processor to it's fullest. *16 cores in all.

I read elsewhere that Tigerton is a Quad Core Quad Socket (*16 cores) version of the Clovertown with another **bonus, but I can't find the page again that spells it out from the beginning, The best part is Tigerton also (after intel said they were dropping it before) has brought back the **ODMC on die memory controller.
Tigerton

Which leads me to believe that this will trickle down into every other new processor after the Tigerton.

Intel needs their core processors out ASAP. Just came across this article where opteron just destroyed a paxville xeon. I don't think Intel can release the core chips soon enough. Link below
http://www.anandtech.com/IT/showdoc.aspx?i=2745

OK Apple was the first company to go to court and to then have the terms of what is and is not Vaporware revised so they can be properly found guilty.

Google->Vaporware Apple SEC

On May 30, 1991, a jury in northern California found that Apple
Computer, Inc., while promoting its Lisa computer and Twiggy disk drive
to potential consumers, had made overly optimistic statements that misled
investors under the federal securities laws. The verdict the jury rendered
would have resulted in damages exceeding $100 million. 2 Although the
trial judge ultimately set the verdict aside, 3 the point was made: potential-
ly huge securities liability can arise from statements made by corporate
officials to promote their companies' products. Nothing in the judge's
decision to set aside the verdict changed the fact that at another time and
place a jury could render an equally large verdict on similar facts.

Most securities litigation arises from securities filings describing
corporate earnings and projected financial performance. In light of Apple,
plaintiffs' class action attorneys who have previously combed annual
reports, 10-Qs, 8-Ks, etc. for overly optimistic statements following a
reported drop in earnings, now have a similarly powerful incentive to
scrutinize earlier product announcements and product press releases when
product performance does not measure up to product hype or when
products do not hit the market when promised. ~9 Even before Apple was
litigated, computer-related and other high-tech firms were more frequently
the target for securities class action suits than companies in other
industries, 2° perhaps because of the ubiquitous vaporware problem. 2t The
ruling in the Apple case thus demands that a new set of guidelines be
drawn up for the evaluation of the permissible boundaries of this product
hyping.

Mmm

(I could swear I posted this once already?!? Did we have a database failure here?!?)

Seems to me there are a lot of missing posts. Also was unable for most of the day to reply.

Must have because I replied to this earlier.

4 socket systems are are going to be $10k+ affairs. Doubtful that Apple will enter into such a market.


It's doubtful that Intel will go with an on-die memory controller for the foreseeable future either. The eweek article doesn't mention anything about it.

Tigerton will supposedly have a new point-to-point FSB - my bet is on octo pumping - that'll allow something like 1.6 to 3.2 GHz FSB data rates, but it doesn't appear to have an ODMC. An ODMC just needlessly complicates Intel's strategy of having a common socket platform between Xeons and Itaniums. In a year or two, Intel wants to be able to plug-and-flash-the-firmware-then-play Itaniums into Xeon systems. An ODMC on ICM CPUs would kill that strategy, and as far as I can tell, they haven't given up on Itanium yet.

It's true that Intel needs to have desktop and server ICM CPUs shipping, but AnandTech knows full well that Glidewell/Bensley platforms with dual independent FSB are coming in a month or so. This platform will bring 2 socket and maybe 4 socket Pentium 4 and Xeon systems performance parity with 2 and 4 socket Athlon64/Opteron systems.

Intel's last Netburst-based Xeon processor to ever ship will be an on-chip half clock 16 MB L3 cache dual-core CPU in 2H 06. I think that'll compete with AMD 2 and 4 socket systems of that time. At least in the power and heat are no problem category.

Once Woodcrest comes, yeah, whole new ball game.

Although remember the most likely reason Woodcrest is coming early (apart from Xeons being crucified by Opterons) is that the Workstation/Server platform is less heat sensitive than the Conroe and esp. Merom platforms (mobile). They are the first of a totally new process and running at 1333Mhz FSB I bet they wont be cool, they wont be "965EE P4" hot but they wont be cool. I would expect any 2 socket/4 core woodcrest Mac Pro to very similar in size/cooling requirements to the G5 quad (but please, dear god, let them have enough space for 3-4 hard drives)

So why do you say they wont be cooler? I thiought the whole line was restructured which ultimately gave it less power, and less heat, better performance.

I just meant that because they are running faster (3.0 and 3.33Ghz on a 1333FSB) there is going to be more heat, I thought they had a 80w or more TDP? Which isn't far off the G5...

They aren't the first there either.

This has been happening a lot. I'v already discussed it with them. so far, they aren't sure why it's happening.

85 watts. But, the cpu's they will be replacing are at 130 watt TDP. 85 compares to the Opteron.

Errr, do you mean the P4 or the G5 is 135watts?

Come on man! Remember when a basic Indy would run you at least US$10K

And the DCC software (of the time) to run on said Indy would easily be 2 or 3 times that! Just to get your toes wet!

Apple needs to provide a monster workstation for the Shake/Motion/FCP/3D profesionals who are more concerned with time rather than money

I am not saying Apple should plan a business model around what would have to be a extreme high cost, possible medium profit margin item, because that idea totally tanked out for Silicon Graphics in the long run; but they could do to cater towards the high end market and reap the marketing benefits

Quad sockets / Quad-core CPUs / Quad SLI QuadroFXs / Quad HDDs

The Quadfecta of future DCC workstations?!? With real-time 4K capabilities?!?

;^p

Oh yeah, Apple also needs to produce (purchase, polish & rebrand, whatever) their own 3D software Luxology always comes to mind

Again,

;^p

The Woodcrest.

That worked for MS. If they didn't do that, NT might never have been taken seriously as a 3D workstation. Strange that they got rid of it after they managed that.

Yeah, strange that, huh?!?

Almost like it was planned to switch DCC houses over to 'cheaper' Windows NT workstations, thereby starting the demise of Silicon Graphics

And yeah, Softimage was foisted off on Avid when Microsoft was done using it to begin raping & pillaging the DCC market

I have said time & again, Apple needs to become the next Silicon Graphics, just without the whole going out of business thing real soon now thing

This is in addition to the whole ruling the consumer market thing they got going so far

And they need to actually build the enterprise market, stupid bastards!

Like I say, Luxology is the place they ought to be

I've been saying the same thing. No one seems to listen apparently.

The Woodcrest is not 135 watts.


One of many resources say that woodcrest is only 80 watts.
todays XEON's are 110 watts. Paxville is 170 watts, Woodcrest is 40% less than Paxville.
AMD's Socket F Opteron is supposed to be 95 watts when that hits, but I think it's AMD's wishful thinking. Either way. Woodcrest is faster, and consumes less power.

more (old) woodcrest

looking for those write up's on Clovertown, Woodcrest, and Kentsfield?

Kentsfield is a Clovertown for your desktop (iMac) 4 cores.


Intel Woodcrest to ship in June, Conroe July, Merom August

I said that the Woodcrest is 85 watts.The current Xeons are 130 - 140 watts.

The problem with this site is that it doesn't carry part of the conversation over, so that sometimes an answer is out of sync with what you see.

The question was which chip was 85 watts. I see, however, that it came out wrong, because the entire post going back is zapped.