Core i7 LAUNCH CONFIRMED: 11/16

[T_Flight]

Ummm, yes...3000 level Chem? Hit it man. That's science. Science and math...well, chem could be considered math too in a way. Hit them books and rock on!

[MomijiTMO]

Essentially : Cover your ass and start early.

Thanks for the advice. It's pretty hard to stick with studying when all your friends who picked up a trade are buying their second car, going out every weekend, have moved out and travelled around the world. I'm pretty stubborn so to read things like this from people who know better does make me more motivated.

[Drwho?]

Little details for the Overclockers preparing their Core i7 stuffs ...
I got many questions about the QPI voltage.
This voltage inclused memory controler and QPI voltage. The default is usually around 1.15v. You got to understand that different technologies are used in this area, that means the power to voltage ratio is different.
you ll see that over 1.3Volts, you will not get much more room for clocking up air cooled. if you stay at 1.22Volts, you can archive already very high speed memory.
Notice that the QPI voltage is shared with the cache, when increasing this voltage, you increase massively the power, due to the number of transistors you are over volting. nothing wrong in the design, it is just the transistor count that is big on this parameter.

Enjoy the fun! and get ready.

thanks for sticking to the NDA ... remember, even the date is under NDA ;-)

[JumpingJack]

thanks for sticking to the NDA ... remember, even the date is under NDA ;-)

Well, we all know it is sometime in November ... Intel's quarterly conference call, Otellini specifically said, 'next month'...

[T_Flight]

Little details for the Overclockers preparing their Core i7 stuffs ...
I got many questions about the QPI voltage.
This voltage inclused memory controler and QPI voltage. The default is usually around 1.15v. You got to understand that different technologies are used in this area, that means the power to voltage ratio is different.
you ll see that over 1.3Volts, you will not get much more room for clocking up air cooled. if you stay at 1.22Volts, you can archive already very high speed memory.
Notice that the QPI voltage is shared with the cache, when increasing this voltage, you increase massively the power, due to the number of transistors you are over volting. nothing wrong in the design, it is just the transistor count that is big on this parameter.

Enjoy the fun! and get ready.

thanks for sticking to the NDA ... remember, even the date is under NDA ;-)

I laugh everytime I hear you say that now...the NDA is under NDA.

Thanks for the info on these voltages. I need to make a word doc with the heading "OK Voltages--OC Voltages--Holy $&*# Voltages---St00pid Voltages" and then put them into columns.

Looking forward to NDANDA lift day and then NDA re-lift day.

[bedlamite]

thanks for sticking to the NDA ... remember, even the date is under NDA ;-)

Lol, it's sometimes so hard to remember about that

[Marvin_The_Martian]

Lol, it's sometimes so hard to remember about that

Hey you do realize that every time this thread is bumped my hearth skips a beat right

I think PG is road testing these I7 cores under Anonymous team PL, would explain why their the 3rd most producing user suddenly and it does coincide with intels launch. It would also be very good pr for Intel, I don't think anything can beat being able to say your new to be released hw is capable of doing what that new 'user' does for the PandeLab.

who knows

[Movieman]

Pandelab has the GPU's.. These suckers will be huge in WCG.. HUGE!
I keep forgeting to sig that line;
"The NDA is under NDA!"

[ap4lifetn]

so movieman, where's ur preview on core i7?!!! how much of a beast is it?

[Marvin_The_Martian]

Pandelab has the GPU's.. These suckers will be huge in WCG.. HUGE!
I keep forgeting to sig that line;
"The NDA is under NDA!"

No that user is ONLY doing SMP wu's, and at an alarming rate to. Mike was the first I think who noticed them and posted here, though I read about it on the official forums before it didn't hit me as being important.

But seeing the time table, it could very well be, especially since Mike confirmed it's an SMP only user, who is serieusly kicking more behind then that can off whoopass from nvidia

http://folding.extremeoverclocking.c...t.php?s=&srt=1

There are two anonymous users, one is the default, the other is listed under pandelab. VJ made a statement they had hw donated to them to test or something for undefined time, and as Mike checked the user stats, all the projects ran are SMP work units.

Then I started to think.. what better way to really stress a new cpu then to give them to a cruncher lab and tell them to push them to their limits ( well ofcourse they could give them to you, but you're just one person ).

Probably I'm wrong, but there is a small small chanche I'm not and I would serieusly both be and giving a real big to Intel.

[Shintai]

Mmm...x86 GPU

Marvin_The_Martian: If it is GPU (Larrabee) you could run SMP client out of the box.

[Marvin_The_Martian]

Ow but isn't Larrabee not even taped out???

[DTU_XaVier]

Might be time to exchange the olde E6600... Been holding strong for 2 years now, through 4 different GFXs, so for fairness' sakes, it might be its turn now
Hope they outperform as badly as we've been told up until now...

Best Regards

[Shintai]

Ow but isn't Larrabee not even taped out???

Im sure it is since its been demoed and tested at 1Ghz.

Engineering samples to more people should also be out in late 2008. And thats now.

Larrabee "2" is also already in development with more job hirings.

https://intel.taleo.net/servlets/Car...=en&csNo=10000

[Marvin_The_Martian]

You know google searches for Larrabee tape out ended me up right here in a thread where you been very active

Idk what they have, I just know it's kicking some serieus ass

[Shintai]

You know google searches for Larrabee tape out ended me up right here in a thread where you been very active

Idk what they have, I just know it's kicking some serieus ass

Haha!

Ok let me help with the google...

http://beyond3d.com/content/news/565

Larrabee first silicon should be late this year in terms of samples and we’ll start playing with it and sampling it to developers and I still think we are on track for a product in late ’09, 2010 timeframe.

But yes, it is tapped out and in ES samples to various people. And its a folding monster. (But that was also expected). The only questionable thing is its "game" power.

[Marvin_The_Martian]

Yeah actually XS where one of the first few links I did get to the part about late 08 tape out but nothing definitive...

Man if this is really larrabee crunching here, everyone who is going to upgrade to I7 will just have to do the same when larrabee hits the stores as the output from whatever equipment they have is incredible

[T_Flight]

It's amazing man. Everybody gets wrapped up in these new releases, but Intel is never standing still. They just keep pushing what is the bleeding edge continuously.

BTW, I would love to have one of the Skynet Engineering Samples. I'm ready! Lay it on me!

[Drwho?]

Before this get miss quoted in the news papers ... this has nothing to do with Larrabee.

[T_Flight]

The plot thickens.

One thing to remember guys. It's possible somebody could have 6 GTX280's folding right now if they had a Sample from Asus on that P6T6. That would be extremely fast on the GPU client. I don;t think there would be anything faster unless there is some top secret hardware out there folding, or some kind of prototype hardware testing being done using Folding.

6 GTX280 would be extremely powerful. if they were tweaked they could get something like 48,000 PPD. That assumes 8000 points per card which is quite reachable.

[dinos22]

just had a look at our distributor pricing........it looks really REALLY depressing
you can see what RRP is listed at http://www.synnex.com.au/Default.asp...t=False&page=5

reseller margin is ok so sell price will be below that for sure but not by a huge way

i know its new tech and all but this pricing will have to come down or aussies just wont buy them in big numbers imo

[Booj]

just had a look at our distributor pricing........it looks really REALLY depressing
you can see what RRP is listed at http://www.synnex.com.au/Default.asp...t=False&page=5

reseller margin is ok so sell price will be below that for sure but not by a huge way

i know its new tech and all but this pricing will have to come down or aussies just wont buy them in big numbers imo

If we assume the 965 will settle around $1100 US price (which is a retail price, with distributor and retail margins built in) even with the pathetic Aussie $ then that equals $1800 give or take. Add in 10% tax we get around $2k. Then add a bit of extra margin and these should be going for $2100-$2200. Thats still a load, but even $2200 is a bargain compared to what they are charging.

Bloody price gouging importers.. They'll just blame the exchange rate, but hundreds will be going into the greedy buggers pockets. maybe the retailers here will make $50 off one lol

[dinos22]

Booj that price estimate of yours is still below wholesale buy price by some margin

[eva2000]

AUD$ retailer listed prices from one local retailer heh

Intel Core2 i7 920 2.66Ghz CPU, LGA1366 8mb, 45nm, DDR3 1066
CP-BX80601920 Pre Order $671.00

Intel Core2 i7 940 2.93Ghz CPU, LGA1366 8mb, 45nm, DDR3 1066
CP-BX80601940 Pre Order $1133.00

Intel Core2 i7 965 Extreme 3.20Ghz CPU, LGA1366 8mb, 45nm, DDR3 1066,
CP-BX80601965 Pre Order $2568.50

[dinos22]

when he sells the CPUs will will probably be able to go buy a cappucino from the enormous profit he's made