OMG Intel 45nm........

[HitmanAgent47]

Now sources said intel delayed the 45nm because of issues in current mainstream mobos

source:
http://www.xbitlabs.com/news/m...e_Microprocessors.html

if this is true then my future upgrade to 45nm is in jeopardy.
i mainly chose the asus p5k cause of 45nm and now i dont know if it is capable.
maybe it is but with noise issue stated in the article

 

[smaky]

lets all jump to conclusions
NOW!

 

[Keysplayr]

To some, it will seem intel screwed up a bit I am sure. To others, it will seem intel is going an extra mile to ensure compatability with less than stellar motherboards.

Question:

Which one are you?

 

[sutahz]

Yes, lets assume the worst and throw a hissy fit, EdzAviator.

 

[AmberClad]

Originally posted by: keysplayr2003
Which one are you?

I think we know who he is/was/has been .

 

[JAG87]

Originally posted by: sutahz
Yes, lets assume the worst and throw a hissy fit, EdzAviator.

hes just reporting an article and stating his opinion, take it easy.

my 2c is that this is gonna cost intel a lot. with all the testing they do, i dont understand how things like these aren't caught earlier in production...

 

[BladeVenom]

Originally posted by: sutahz
Yes, lets assume the worst and throw a hissy fit, EdzAviator.

Good idea. Otherwise it could get boring.

 

[BlueAcolyte]

I note that this article says they are still releasing in February and March.

Also, WTF is FSB noise?

 

[Idontcare]

Originally posted by: BlueAcolyte
I note that this article says they are still releasing in February and March.

Also, WTF is FSB noise?

Not sure what FSB noise would be, but neither is the author of this quality article at X-bit labs as they keep calling it a "PSB" and never mention an FSB.

Now granted the author of course could have simply jsut been born yesterday and he has never heard of the term "FSB" despite it being used for, oh, the past 20 years or longer...or this could be yet another waste of interweb space "journalism" at work.

My take on all this is that there is good reason why Penryn don't work on "old" i680 boards...the same reason those old i680 boards don't support massive FSB overclocking with the original B3 quads...crappy signal/noise quality ratios...and guess what Yorkfields require for FSB?

 

[heyheybooboo]

Originally posted by: BlueAcolyte
I note that this article says they are still releasing in February and March.

Also, WTF is FSB noise?

I think an electrical engineer needs to 'splain that one to us.

Noise is like 'static' or a disturbance on the signal and could mean a bunch of different things.

 

[Killrose]

Originally posted by: BlueAcolyte
I note that this article says they are still releasing in February and March.

Also, WTF is FSB noise?

I'll venture a guess > intermod or trace problems with signals jumping the traces on the motherboard and trashing other signals on nearby traces.

 

[BlueAcolyte]

hmm, looks like no one really knows...

 

[BonzaiDuck]

Since we're all speculating . . .

Some of the "old" 680i boards were "ready" for 1,333FSB-rated CPUs. And some of the "old" 680i boards are Wolfdale-certified.

Other info about 680i incompatibility with the Yorkfield quad suggested that the chipset was compatible, but there was some "hardware" modification needed to the board.

Now I find that boards with P35 chipsets have problems with Yorkfield quads. And I still wonder, given Intel's spat with nVidia, that they somehow made a hardware differentiation in Yorkfield to thwart nVidia, or they just went to market too fast with the QX9650.

Who knows for sure?

IdontCare -- do you know anyone who was able to push a B3-stepping of Q6600 beyond 3.2 Ghz on air-cooling and VCore below 1.42V on ANY motherboard? Or on a NON-680i board with an INtel chipset?

I'm easy, though. If it turns out I can't move to WolfDale or Yorkfield eventually with my motherboard, I'll go "X38."

[In the room the women come and go . . . . speaking of Michaelangelo . . . . there will be time, there will be time . . . . .]

 

[PCTC2]

Originally posted by: BonzaiDuck
Since we're all speculating . . .


I'm easy, though. If it turns out I can't move to WolfDale or Yorkfield eventually with my motherboard, I'll go "X38."

I'm already going "X38" with the P5E3 Deluxe Wifi and the "790i" if and when it comes out. I'm tired of my 680i and I want a QX9650....

 

[aigomorla]

Originally posted by: PCTC2

Originally posted by: BonzaiDuck
Since we're all speculating . . .


I'm easy, though. If it turns out I can't move to WolfDale or Yorkfield eventually with my motherboard, I'll go "X38."

I'm already going "X38" with the P5E3 Deluxe Wifi and the "790i" if and when it comes out. I'm tired of my 680i and I want a QX9650....

rawr!

i already applied for stepup to 780i. My 680i is sitting in a box collecting dust, so i might as well spend the 100 dollars Board upgrade + shipping + Tax :T and get the new board. So far im hearing rumors that its a bit more stable then the 680i. So heres crossing my fingers. But a high end board like that needs high end hardware, so im going to have to debate if i really want to take apart one of my G0's and strap on 3 8800GTS's


LOL i would probably get flamed at if i remade my NAS with the 780i + 3 x 8800GTS. But i would no way in hell put another nvidia intel board in my main rig until i saw some long term stability on it.

 

[PCTC2]

Originally posted by: aigomorla

Originally posted by: PCTC2

Originally posted by: BonzaiDuck
Since we're all speculating . . .


I'm easy, though. If it turns out I can't move to WolfDale or Yorkfield eventually with my motherboard, I'll go "X38."

I'm already going "X38" with the P5E3 Deluxe Wifi and the "790i" if and when it comes out. I'm tired of my 680i and I want a QX9650....

rawr!

i already applied for stepup to 780i. My 680i is sitting in a box collecting dust, so i might as well spend the 100 dollars Board upgrade + shipping + Tax :T and get the new board. So far im hearing rumors that its a bit more stable then the 680i. So heres crossing my fingers. But a high end board like that needs high end hardware, so im going to have to debate if i really want to take apart one of my G0's and strap on 3 8800GTS's


LOL i would probably get flamed at if i remade my NAS with the 780i + 3 x 8800GTS. But i would no way in hell put another nvidia intel board in my main rig until i saw some long term stability on it.

LOL. A NAS capable of more FPS than my main rig. But why 3x GTS's? They aren't capable of triple SLI. just want to run 6x Monitors on your NAS? A usage graph on each with temperature readings running at 200FPS?

I'm waiting a couple months for the next nVIDIA chipset "790i".

 

[aigomorla]

Originally posted by: PCTC2
LOL. A NAS capable of more FPS than my main rig. But why 3x GTS's? They aren't capable of triple SLI. just want to run 6x Monitors on your NAS? A usage graph on each with temperature readings running at 200FPS?

I'm waiting a couple months for the next nVIDIA chipset "790i".

I heard the new G92' GTS's can do threeway SLI.

Also because none of my other setups has the room/cooling power to handle a loaded Yorkie, + 3 G92's on water.

Dude, only 2 of my rigs can handle that kind of heat load. My main and my NAS. And no thank you on my main. I can always migrate my NAS files on another cruncher Quad.


But anyhow, you honestly think NVidia is going to make the 790i less buggy AFTER they screwed intel over on SLI on intel chipsets?

I heard Intel didnt give NVidia all the CPU codes because of this on the new yorkies even.

Also i really dont require threeway SLI, but it looks hot expecially watercooled. And theres no point in that board unless you threeway SLI. :\

Dual 24's FTW! and Ditch SLI! <--- is still my original statement tho.

 

[Mr Fox]

Originally posted by: Idontcare

Originally posted by: BlueAcolyte
I note that this article says they are still releasing in February and March.

Also, WTF is FSB noise?

Not sure what FSB noise would be, but neither is the author of this quality article at X-bit labs as they keep calling it a "PSB" and never mention an FSB.

Now granted the author of course could have simply jsut been born yesterday and he has never heard of the term "FSB" despite it being used for, oh, the past 20 years or longer...or this could be yet another waste of interweb space "journalism" at work.

My take on all this is that there is good reason why Penryn don't work on "old" i680 boards...the same reason those old i680 boards don't support massive FSB overclocking with the original B3 quads...crappy signal/noise quality ratios...and guess what Yorkfields require for FSB?

"PSB NOISE" = Printed Circuit Noise this is caused by cost cutting in the fabrication process and it is noise that is generated by components that lack shielding on the circuit paths of the PCB.

Proper design and layering, keeps noise to a minimum but also adds cost.

 

[PCTC2]

Originally posted by: aigomorla

Originally posted by: PCTC2
LOL. A NAS capable of more FPS than my main rig. But why 3x GTS's? They aren't capable of triple SLI. just want to run 6x Monitors on your NAS? A usage graph on each with temperature readings running at 200FPS?

I'm waiting a couple months for the next nVIDIA chipset "790i".

I heard the new G92' GTS's can do threeway SLI.

Also because none of my other setups has the room/cooling power to handle a loaded Yorkie, + 3 G92's on water.

Dude, only 2 of my rigs can handle that kind of heat load. My main and my NAS. And no thank you on my main. I can always migrate my NAS files on another cruncher Quad.


But anyhow, you honestly think NVidia is going to make the 790i less buggy AFTER they screwed intel over on SLI on intel chipsets?

I heard Intel didnt give NVidia all the CPU codes because of this on the new yorkies even.

Also i really dont require threeway SLI, but it looks hot expecially watercooled. And theres no point in that board unless you threeway SLI. :\

Dual 24's FTW! and Ditch SLI! <--- is still my original statement tho.

Naw. The G92 GTS's only have a single SLI connector.

But I'm thinking just a dual-SLI with a Yorkie that at least boots with it on my 790i.

I'm going to be using a P5E3 for my secondary rig, running dual G92 8800GTS's running 2x 20" LCD's surrounding a main 30" LCD.

 

[JumpingJack]

Ok... read the story carefully ... if it is indeed true (the rumor mill is in over drive on this one), what Xbit says is that Intel is changing the packaging to ensure that it will be flawless in the 4 layer PCB boards..... in which case you have no worries.

Second the Asus P5K is a high end board, it is a 6-layer PCB. In this case you would have no worries....

Again, assuming this is true...

 

[JumpingJack]

Originally posted by: BlueAcolyte
I note that this article says they are still releasing in February and March.

Also, WTF is FSB noise?

Intel uses a parallel bus system, as such, there are at least one routing for each bit in the bus for each block in the transaction. For Intel 64 bits, thus 64 separate lines ...

An electron traveling down a wire must be pushed by a voltage, a traveling electron will also lose energy as it travels and it loses that energy by giving off electromagnetic radiation. A wire next to another can then act as an antennea, absrobing that radiation and transmitting it back into the electrical stream this generates noise or cross talk.

http://www.google.com/search?s...s+signaling+cross+talk

You will find many papers, methods and techniques by which people study to find ways of transmission of digital signals cleanly without noise or cross-talk betwen adjacent runs. This is just the physics of traveling electrons and the engineering bussing systems.

In fact, if you look closely at various motherboards, you will see in the top one or two layers the electrical wiring between sockets, in some cases you can see a wiggle in the trace or a loop back and forth, for example see this image:

http://www.virtual-hideout.net...imits/images/07_sm.jpg

Look at the line traces right around the P and the R of the word 'PRO', notice some are wiggled, and between the 5 and the P a few loop back against each other...

This is done for several reasons, one is to break up the potential cross-signalling that may introduce noise into the line, other reasons are to match the total impedance and balance out the load to the two devices it is running between.

Nonetheless, with a dual die packaging, Intel has to be mindful of potential EM interferiences within the package, also spec the right voltage and strengths of the signal to the package, and design such that noise from one bit line will not corrupt the signal from a different bit line....

Despite people's jokes and inuendo's that the MCM is simply gluing together, this is quite abit of engineering that goes into the actual package, one might even say it is innovative for package technology (but that type of person would be one who knows what they are actually talking about).

Jack

 

[nyker96]

really don't care why they doing this, as long as at the end of the day the cpu works with the new P35-E I bought that's all.

 

[j0j081]

sup Edz.

 

[JumpingJack]

Originally posted by: nyker96
really don't care why they doing this, as long as at the end of the day the cpu works with the new P35-E I bought that's all.

Read the article again, Intel is 'supposedly' delaying the launch to ensure that the CPU will work in the low quality boards. Your P35-E is NOT a low quality board, but since Intel is ensuring lowest common denominator you have no worries.

 

[WaTaGuMp]

Does the delay include the Wolfdale such as the E8400 etc or just quads?