Ivy Bridge Release Accelerated?

kirtar said:
That's what I thought too. It almost seems like they're doing it to try to make up for the current fiasco. Actually, after re-reading it, it's just an "unveiling," probably not a release
Click to expand...

Whatever Intel does, it does US no good until motherboards are available. Will the current 1355 socket be supported? What about the expected socket 2011? Will a new chipset be needed?
 
RM_Bulldog said:
i find it funny how people get so pissed defending their purchases
Click to expand...

Some people do defend purchases, but many are just being realistic. Expecting 30% gains is most likely unrealistic. 10%-15% has been the more conservative expectations. The question is how much more will it cost for a 10%-15% increase over SB.

I like to upgrade but bang for the buck is always nice, too.
 
jifffy said:
Some people do defend purchases, but many are just being realistic. Expecting 30% gains is most likely unrealistic. 10%-15% has been the more conservative expectations. The question is how much more will it cost for a 10%-15% increase over SB.

I like to upgrade but bang for the buck is always nice, too.
Click to expand...

Well, if in the die shrink they are able to improve IPC some we might get some decent improvements along with the sure bet that we will see higher clocks. Who knows, its too early to say for sure.
 
I'm more concerned with the die shrink/power consumption and on-board GPU performance more than overall CPU performance personally, as I would love to see one of these in the next iteration of the MacBook Air
 
savantu said:
The ramp up is underway with Q4 being the schedule for revenue shipments. I do not believe we will see IB pulled in, it would affect the tick-tock process. Secondly, Intel still has a lot of excess 32nm capacity and probably the process isn't yet amortized. Thirdly, SB doesn't appear to be threatened in any way by Llano and GPU-less 6 and 8 core models will handle BD without breaking a sweat.

Intel will spend more than $9B this year to bring 22nm online, which is a 80% increase compared to 2010. Moreover, they will bring to 22nm 4 FABs instead of the usual 3. That's a huge capacity being brought online. Why increase the risks and hurry up for a month or two?
Click to expand...

SB is native 32nm. the tock was the Nehalem 45-32nm. IB will be 22nm and the "tock" respective to SB. IB's successor will be native 22nm.

And if they do release IB on 22nm, it will probably be only high end/server. Notebooks/mainstream will probably continue to be 32nm as was planned if IB wasn't going to be released early.
 
Hey noob question here go easy on me. By reducing the nm of a cpu chip it makes it faster? someone explain how it works thanks!
 
ntherblast said:
Hey noob question here go easy on me. By reducing the nm of a cpu chip it makes it faster? someone explain how it works thanks!
Click to expand...

Longer answer:

You can do a number of things with a smaller process:

Add more transistors (might make it faster)
Increase clock speed (will make it faster)
Do nothing (won't make it faster, but will reduce power consumption/heat)

so yeah not necessarily...
 
ntherblast said:
Hey noob question here go easy on me. By reducing the nm of a cpu chip it makes it faster? someone explain how it works thanks!
Click to expand...

Normally reduces power consumption, and often allows higher clocks, but it isn't a given. Often-times additional tweaks are made to the architecture to take advantage of the smaller die size as well, also potentially increasing performance.

That being said, Intel has always (to my knowledge) increased performance when they've done a die shrink.
 
Forceman said:
Normally reduces power consumption, and often allows higher clocks, but it isn't a given. Often-times additional tweaks are made to the architecture to take advantage of the smaller die size as well, also potentially increasing performance.

That being said, Intel has always (to my knowledge) increased performance when they've done a die shrink.
Click to expand...

Prescott is an exception to that streak though.
 
There will likely be IPC improvements from moving to DDR3 1600 and having a larger cache at the very least. But obviously the tock updates do not improve as much in terms of IPC as the ticks.

In terms of the majority of Intel's sales though (not OCers) they can sell CPUs with a large performance boost simply by clocking them higher. Look at the room SB chips have now, combine that with the lower process size, they can always fall back on this if Bulldozer does push them.
 
I feel really dumb for asking this, but is the gist of better performing integrated solutions that "average joe" consumers will be thrilled with the results of increased performance and lower power draw and thus buy up laptops and consumer grade computers?

I realize this in the end benefits us due to die shrink but I'm having a hard time coming to terms with the first part actually spurring higher sustained sales.
 
MrSneis said:
I feel really dumb for asking this, but is the gist of better performing integrated solutions that "average joe" consumers will be thrilled with the results of increased performance and lower power draw and thus buy up laptops and consumer grade computers?

I realize this in the end benefits us due to die shrink but I'm having a hard time coming to terms with the first part actually spurring higher sustained sales.
Click to expand...

Well that's why I'm excited about it. I don't do any heavy gaming on my laptop but would like it to be somewhat capable for the occassional game. IB's integrated solution will allow me to do just that without adding unecessary bulk, heat, weight af a seperate, dedicated GPU.
 
I guess in the end, the [h] types only care about the latter half; smaller dies, lower power draw; which could be increased performance around 10-20%....

I can see the consumer side too, I am currently on a 17" mbp which gets around 4-8 hours real battery life which a few years ago was unheard of!
 
(from what i read) The good thing about Brazos is they put a good combination of CPU and GPU power, so the GPU has just enough processing power so that they're not bottlenecked by the CPU.

So I wonder whether Llano will have the same balance of CPU and GPU power like Brazos. and finally how will it compare to Ivy Bridge's on a system without discrete graphics.
 
Originally Posted by Forceman;
Normally reduces power consumption, and often allows higher clocks, but it isn't a given. Often-times additional tweaks are made to the architecture to take advantage of the smaller die size as well, also potentially increasing performance.

That being said, Intel has always (to my knowledge) increased performance when they've done a die shrink.
Click to expand...

ryan_975 said:
Prescott is an exception to that streak though.
Click to expand...

you forgot to mention the P4 chip that got the whole mess started. the 'willamette' on 423 bundled with RDRAM was trounced by PIII's. clock for clock this P4 cpu was horrible. the P4 was only able to surpass the PIII (@ 1000MHz) when its clock was near 2GHz.
 
aphasia said:
you forgot to mention the P4 chip that got the whole mess started. the 'willamette' on 423 bundled with RDRAM was trounced by PIII's. clock for clock this P4 cpu was horrible. the P4 was only able to surpass the PIII (@ 1000MHz) when its clock was near 2GHz.
Click to expand...

But they could scale up to 10ghz! :p
 
Ryan711 said:
SB is native 32nm. the tock was the Nehalem 45-32nm. IB will be 22nm and the "tock" respective to SB. IB's successor will be native 22nm.

And if they do release IB on 22nm, it will probably be only high end/server. Notebooks/mainstream will probably continue to be 32nm as was planned if IB wasn't going to be released early.
Click to expand...

Ticks are the die shrinks.
Tocks are new architecture on the matured process.

Nehalem was a tock at 45nm and Westmere was the tick 32nm shrink.

SB was the 32nm tock and IB will be a tick 22nm shrink.

Its not IF they do release IB on 22nm. IB is 22nm itself and will be a wide release. Intel itself said IB will make 4 core the minimum even on entry level desktop.
 
aphasia said:
you forgot to mention the P4 chip that got the whole mess started. the 'willamette' on 423 bundled with RDRAM was trounced by PIII's. clock for clock this P4 cpu was horrible. the P4 was only able to surpass the PIII (@ 1000MHz) when its clock was near 2GHz.
Click to expand...

Willamette wasnt a die shrink though.
 
Speaking of Intel slides, I heard that they think 11 nm is possible by 2015. That's INSANE!

That means, in only a few years, we may see the advent of nanoelectronics.
 
That's_Corporate said:
Speaking of Intel slides, I heard that they think 11 nm is possible by 2015. That's INSANE!

That means, in only a few years, we may see the advent of nanoelectronics.
Click to expand...

I wouldn't be surprised if they have plans for smaller manufacturing processes using new silicon hybrid materials or even completely new ones. Intel seems to be in their own bubble. While nvidia and amd struggle to get their graphics cards down to 28nm, intel is pushing 22nm and beyond.
 
Well of course. Intel has its own well established Fabs, like in Costa Rica, Malaysia and a bunch of others. AMD and Nvidia have to use TSMC in Taiwan...
 
Falkentyne said:
Well of course. Intel has its own well established Fabs, like in Costa Rica, Malaysia and a bunch of others. AMD and Nvidia have to use TSMC in Taiwan...
Click to expand...

Yea and TSMC/GF are struggling. I think that was implied when I said AMD/NV
 
Falkentyne said:
Well of course. Intel has its own well established Fabs, like in Costa Rica, Malaysia and a bunch of others. AMD and Nvidia have to use TSMC in Taiwan...
Click to expand...

Those are packaging facilities.

The FABs are in Oregon, Arizona, New Mexico, Leixlip ( Ireland ), and Qiryat Gat ( Israel ).
 
I had actually heard from HP that it may be pushed back a quarter due to the recall.
 
You must log in or register to reply here.
Back
Top