It’s possible that Apple will just milk the plastic MB machine with the 9400M for as long as possible...
Based on my assumptions in the previous post, it would indicate that Intel may continue to produce Penryn past the middle of 2010, which would mean Apple keeping the Mac mini and Macbook at Core 2 Duo with 9400M... While MacBook Pro goes Arrandale. iMac could have a mixture of Arrandale, Clarkdale and Lynnfield.
At this stage I think it's a 50% chance that Apple will outright disable the 45nm GPU in MacBook Pro Arrandales so it draws no power at all and just focus on managing the discrete GPU. As Dave mentioned above, 5 series ATI Mobility Radeons at 40nm, if used by Apple, have really low TDPs anyway so the Intel GPU is could be just unnecessary given additional resources needed to power it, as well as OS X coding needed to power-manage and switch between it and the discrete GPU.
Based on my assumptions in the previous post, it would indicate that Intel may continue to produce Penryn past the middle of 2010, which would mean Apple keeping the Mac mini and Macbook at Core 2 Duo with 9400M... While MacBook Pro goes Arrandale. iMac could have a mixture of Arrandale, Clarkdale and Lynnfield.
At this point in time that does seem the most likely.
It kills me too. Took me months to wrap my head around it... BTW looks like Intel is cleverly hedging its bets in 2010 with keeping the 45nm fabs churning while ramping up but not overdoing 32nm. And maintaining Core 2 and Nehalem architectures while preparing the Sandy Bridge architecture. Man, new employees at Intel probably spend their first week learning/ being force-fed all the codenames. At least, AFAIK, they won't be forced to say "Nehalem!" with gusto unlike... http://www.youtube.com/watch?v=7GM4Lt5k24s
So, here's where I'm confused. Does this mean Intel will provide new information on their CES Jan 7 2010 announcement about 4-core Arrandale and Clarkdale? Or, does this mean for at least half a year Intel will continue with the 45nm non-GPU Penryn, and Lynnfield (desktop) and Clarksfield (mobile) for 4 cores...? (In addition to continuing with Bloomfield)?
I have no idea. Intel’s nomenclature confuses me. You’d think that they’d go quad-core with the 32nm architecure but the only thing we have any evidence of is a sex-core triple-channel “Gulftown”, which will liekly find its way into the high-end Mac PRo next year but won’t even begin to be ready for anything else for awhile. Maybe the mobile segment will go from 2 to 6 cores on the 32nm, or perhaps we’ll have to wait for 22nm to see quad-core.
As I understand, Clarksfield battery life (a different thing from TDP alone) is horrible so Apple's Macbook Pro 17" will probably not use that. So the next refresh to the MacBook Pro processors would *all* be dualcores (4 threads on the higher-end) but *NO QUAD-CORE* ... at least until 32nm mobile quad-core processors come*
They save on the lack of a Northbridge and already have an 8 hour battery in the 17” MBP with a 35W Penryn. Even if the battery life is worsened by a great deal, portable workstations, which I think the 17” MBP falls into may warrant this as a BTO option. Since both Clarkfield and Arrandale use the same socket this looks to add to the likelihood and Apple can still claim that the 17” MBP gets x-many hours of battery life since the default builds will be Arrandale. Your guess is as good as mine as to what Apple will do. In the end I think it comes down to knowing the projected sales figures to determine the true likelihood.
I have no idea. Intel’s nomenclature confuses me. You’d think that they’d go quad-core with the 32nm architecure but the only thing we have any evidence of is a sex-core triple-channel “Gulftown”.......Maybe the mobile segment will go from 2 to 6 cores on the 32nm, or perhaps we’ll have to wait for 22nm to see quad-core.
Yeah, glad I'm not the only one a little confused. But from what I gather there is no leaked info on quad-core 32nm chips at all. This could change at the CES announcement in a few weeks time, and of course through the course of 2010.
If there are no quad-core 32nm chips for mobile, we won't have to wait for 22nm, Sandy Bridge is a new microarchitecture that will *start on 32nm* ~ and this is supposed to have 4 cores standard for desktop and mobile. As per the tick-tock strategy, it goes something like Core on 65nm, Core on 45nm, Nehalem on 45nm, Nehalem on 32nm, Sandy Bridge on 32nm, Sandy Bridge on 22nm... Here's the pictarz:
I suppose there has always been some overlap between the ticks and the tocks. Intel is showing it can keep up on a year-by-year tick-tock strategy. But I suppose they never promised each transition would be hard and fast, only they would introduce the tick or tock and then transition within... well, within whatever schedule they preferred.
But this "tick" of 32nm seems, at least starting out in 2010, *conservative* given it is only dualcores and they have yet to mention how they are going to transition the "big boys" of Bloomfield, Lynnfield and Clarksfield down to 32nm... if at all. If Intel decides that Bloomfield and Lynnfield are top-end desktop chips with decent enough power and thermals (which they are), they might not bring them to 32nm at all. If Intel decides that a quad-core mobile chip is not really needed in mainstream markets, they might not bring Clarksfield down to 32nm at all.
So Intel in all of 2010 could play the strategy of having the higher-end chips produced at 45nm for the enthusiast/workstation markets, and leave the 32nm chips for mobile, thin-and-light, and slim-desktop, mainstream-desktop segments. I guess it also depends on how many 45nm fabs Intel would keep running and what schedule there was for phasing out 45nm and bringing in 32nm. Additionally, since at 45nm Intel has generally a clear lead (though a higher price) over AMD in terms of performance, in both desktop and mobile, Intel may not be in a big rush to transition *everything* to 32nm by end of 2010.
In other words, unlike what I was assuming, in 2010 there may be no 32nm quad-core mobile CPU.
That is, in fact, entirely possible. Exactly zero leaked roadmaps show 32nm quad-core desktop or mobile processors from Intel in 2010 (or prior to the next "tock", Sandy Bridge). Just dual-cores (running four threads with hyperthreading) and hexa-cores.
There are some quad-core 32nm Xeons listed, but it's not clear if they will be made that way or will be hexas with two cores disabled. It's a big, complex chip, so that may be the case.
That is, in fact, entirely possible. Exactly zero leaked roadmaps show 32nm quad-core desktop or mobile processors from Intel in 2010 (or prior to the next "tock", Sandy Bridge). Just dual-cores (running four threads with hyperthreading) and hexa-cores.
it seems like everytime Intel is in the lead technology wise they mlk processes and drag out moving tech forward. It would be good to see vendors adopting a competitive AMD solution to keep Intel on it's toes. Of course this isn't the case right now so we are stuck with Intel draging ass.
Quote:
There are some quad-core 32nm Xeons listed, but it's not clear if they will be made that way or will be hexas with two cores disabled. It's a big, complex chip, so that may be the case.
the flip side here is that we simply haven't heard about all of Intels plans. Frankly though Atom is a sign to me of Intels lack of commitment. Atom is the last place we should be seeing yet another spin at 45nm. It is like they are just giving up on competeing with ARM.
In other words, unlike what I was assuming, in 2010 there may be no 32nm quad-core mobile CPU.
It doesn't look like there will be 32 nm quad-core Westmere.
Quote:
Originally Posted by nvidia2008
But this "tick" of 32nm seems, at least starting out in 2010, *conservative* given it is only dualcores and they have yet to mention how they are going to transition the "big boys" of Bloomfield, Lynnfield and Clarksfield down to 32nm... if at all.
Bloomfield is going to coexist with Gulftown in 2010.
it seems like everytime Intel is in the lead technology wise they mlk processes and drag out moving tech forward. It would be good to see vendors adopting a competitive AMD solution to keep Intel on it's toes. Of course this isn't the case right now so we are stuck with Intel draging ass.
the flip side here is that we simply haven't heard about all of Intels plans. Frankly though Atom is a sign to me of Intels lack of commitment. Atom is the last place we should be seeing yet another spin at 45nm. It is like they are just giving up on competeing with ARM.
Dave
Jeez, Sandy Bridge is Q4 2010. How much foot dragging can their be if Intel is still more or less on their tick-tock schedule. It's aggressive as hell and nice after how poorly they were executing just a few years ago.
Atom is at 45nm because it's not just mobile but also cheap. They'll go 32nm with Cedarview in 2011 which is about what their roadmap said in 2008. They haven't given up competing with ARM given they aren't really competing with ARM yet. That's like saying that ARM has given up on competing with Intel because there's no quad-core Cortex A9 yet.
They're going to converge on the netbook market for real right around when the A9 goes from dual core to quad and Intel goes Moorestown.
it seems like everytime Intel is in the lead technology wise they mlk processes and drag out moving tech forward. It would be good to see vendors adopting a competitive AMD solution to keep Intel on it's toes. Of course this isn't the case right now so we are stuck with Intel draging ass...
Quote:
Originally Posted by vinea
Jeez, Sandy Bridge is Q4 2010. How much foot dragging can their be if Intel is still more or less on their tick-tock schedule. It's aggressive as hell and nice after how poorly they were executing just a few years ago...
I agree with both of you, in a sense... Intel still wants to show it can "execute", as you say. So we will probably see Sandy Bridge on time and it should be impressive. But with Intel's dominance and Nehalem architecture proving to be really formidable against AMD, Intel is now enjoying the best of both worlds. They will milk 45nm for all it is worth while positioning Westmere as they see fit.
Intel is actually really strongly positioned in 2010, with their more-or-less locking out of other GPU and chipset manufacturers continuing at pace.
Maybe I'm getting old but I don't see needing a Core i7-level of CPU performance in my laptop or desktop in 2010. Most consumers and businesses also wouldn't. With the exception of a decent triple-core or quad-core CPU in a laptop at a Core i5-level of performance with great battery life.
What I am interested in and would like to see, more importantly, is that 250GB quality SSD at sub-$300 prices. That's the main bottleneck anyway in daily usage.
Additionally, I would like to see an entirely new form factor, ie. something like the tablet that will take us out of this keyboard-desktop-laptop paradigm.
See, that's the problem of too-dominant a company... for example, Intel and TSMC -- they can pursue their agenda and shape market demand around what they produce. If we had more competition, perhaps, we'll see $200 250GB SSDs, triple-core or quad-core in laptops, great ARM-based tablets, and a lot more GPUs from AMD and Nvidia at 32nm and below.
On that point, does anyone think that triple-core in a laptop is actually a great intermediate step for 2010? In desktops the differential is not so great, you'd go quad-core over triple-core in most cases. But for laptops, it would be a great transition CPU. Something like a Phenom II X3 or Athlon II X3 at 32nm, alongside a 40nm Mobility Radeon 5770 1GB VRAM... would be... sweeet. *sigh*
...They're going to converge on the netbook market for real right around when the A9 goes from dual core to quad and Intel goes Moorestown.
I wonder what the "netbook" market would be like at that time. Again, maybe I'm getting old, but Windows or OS X at anything less than 13" is totally not useable for me. It's not like I have big hands or poor eyesight either, I'm 31, 5'7" and don't wear glasses.
Perhaps a personal preference, but I say to Intel's Atom, "show me the useability!".
it seems like everytime Intel is in the lead technology wise they mlk processes and drag out moving tech forward. It would be good to see vendors adopting a competitive AMD solution to keep Intel on it's toes. Of course this isn't the case right now so we are stuck with Intel draging ass.
the flip side here is that we simply haven't heard about all of Intels plans. Frankly though Atom is a sign to me of Intels lack of commitment. Atom is the last place we should be seeing yet another spin at 45nm. It is like they are just giving up on competeing with ARM.
Dave
I think a lack of 32nm quad-cores, if that is true, indicates more the limitations of Intel's R&D budget and manufacturing capacity. Their resources are simply not unlimited, so they are focusing on their low-end volume seller and their high-end profit maker. The middle gets squeezed into 2011.
I think a lack of 32nm quad-cores, if that is true, indicates more the limitations of Intel's R&D budget and manufacturing capacity. Their resources are simply not unlimited, so they are focusing on their low-end volume seller and their high-end profit maker. The middle gets squeezed into 2011.
Isn't it weird how the low-end would do fine with Core 2 Duos? But Intel I guess has to push forward with selling the "latest and greatest" to the masses.
Core i3 and i5. Your new best friends in 2010.
[soup nazi] No quadcore for you![/soup nazi]
Us mid-end users I guess will benefit in 2011 or so with Sandy Bridge 32nm quadcores. \
I know, I'm contradicting myself, I said previously I didn't need quadcore. Though a triple-core would be a nice tasty nibble.
The lack of a 32nm Clarksdale mobile i7 is a bit of a disappointment. I think at 32 nm the clock speeds would be much higher and its performance would be excellent.
I still can't wait to see the mobile Arrandales. This may be the year for a new laptop for me. MBP, MBA ? I can't wait to see what Apple does with these machines this year.
Definitely looking forward to next version. I have never purchased a mac before but have run a hacintosh briefly to see if I'd like it. I'm ready to buy a 15" MBP as soon as Apple offers something over the currently horrible screen res. Something that does full 1080 would be nice. Could care less about the processors really.
As soon as they do, I'll be set. Otherwise by March / April I'll look into an E6500.
Definitely looking forward to next version. I have never purchased a mac before but have run a hacintosh briefly to see if I'd like it. I'm ready to buy a 15" MBP as soon as Apple offers something over the currently horrible screen res. Something that does full 1080 would be nice. Could care less about the processors really.
As soon as they do, I'll be set. Otherwise by March / April I'll look into an E6500.
Going from 900 to 1080 isn?t much, but I hope they don?t change it until they can offer full resolution independence as It?s already getting too small to read. Windows is much better at scaling than OS X but I hope Apple finally gets it done for 10.7 despite advertising it at least back to 10.5.
Isn't it weird how the low-end would do fine with Core 2 Duos? But Intel I guess has to push forward with selling the "latest and greatest" to the masses.
Core i3 and i5. Your new best friends in 2010.
[soup nazi] No quadcore for you![/soup nazi]
Us mid-end users I guess will benefit in 2011 or so with Sandy Bridge 32nm quadcores. \
I know, I'm contradicting myself, I said previously I didn't need quadcore. Though a triple-core would be a nice tasty nibble.
But the quad-core mobile Nehalem arch. processors have already been made. They're 45nm, but they're out on the market. And as for the low-end, Core 2 derived chips will still make up the majority of Intel's sales by volume for the next year.
However, smaller manufacturing process = more chips per wafer = less cost per CPU to Intel. So it's in their best interest to move their best sellers to 32nm first.
Comments
It’s possible that Apple will just milk the plastic MB machine with the 9400M for as long as possible...
Based on my assumptions in the previous post, it would indicate that Intel may continue to produce Penryn past the middle of 2010, which would mean Apple keeping the Mac mini and Macbook at Core 2 Duo with 9400M... While MacBook Pro goes Arrandale. iMac could have a mixture of Arrandale, Clarkdale and Lynnfield.
At this stage I think it's a 50% chance that Apple will outright disable the 45nm GPU in MacBook Pro Arrandales so it draws no power at all and just focus on managing the discrete GPU. As Dave mentioned above, 5 series ATI Mobility Radeons at 40nm, if used by Apple, have really low TDPs anyway so the Intel GPU is could be just unnecessary given additional resources needed to power it, as well as OS X coding needed to power-manage and switch between it and the discrete GPU.
Based on my assumptions in the previous post, it would indicate that Intel may continue to produce Penryn past the middle of 2010, which would mean Apple keeping the Mac mini and Macbook at Core 2 Duo with 9400M... While MacBook Pro goes Arrandale. iMac could have a mixture of Arrandale, Clarkdale and Lynnfield.
At this point in time that does seem the most likely.
They aren't confusing at all.
So, here's where I'm confused. Does this mean Intel will provide new information on their CES Jan 7 2010 announcement about 4-core Arrandale and Clarkdale? Or, does this mean for at least half a year Intel will continue with the 45nm non-GPU Penryn, and Lynnfield (desktop) and Clarksfield (mobile) for 4 cores...? (In addition to continuing with Bloomfield)?
I have no idea. Intel’s nomenclature confuses me. You’d think that they’d go quad-core with the 32nm architecure but the only thing we have any evidence of is a sex-core triple-channel “Gulftown”, which will liekly find its way into the high-end Mac PRo next year but won’t even begin to be ready for anything else for awhile. Maybe the mobile segment will go from 2 to 6 cores on the 32nm, or perhaps we’ll have to wait for 22nm to see quad-core.
As I understand, Clarksfield battery life (a different thing from TDP alone) is horrible so Apple's Macbook Pro 17" will probably not use that. So the next refresh to the MacBook Pro processors would *all* be dualcores (4 threads on the higher-end) but *NO QUAD-CORE* ... at least until 32nm mobile quad-core processors come*
They save on the lack of a Northbridge and already have an 8 hour battery in the 17” MBP with a 35W Penryn. Even if the battery life is worsened by a great deal, portable workstations, which I think the 17” MBP falls into may warrant this as a BTO option. Since both Clarkfield and Arrandale use the same socket this looks to add to the likelihood and Apple can still claim that the 17” MBP gets x-many hours of battery life since the default builds will be Arrandale. Your guess is as good as mine as to what Apple will do. In the end I think it comes down to knowing the projected sales figures to determine the true likelihood.
I have no idea. Intel’s nomenclature confuses me. You’d think that they’d go quad-core with the 32nm architecure but the only thing we have any evidence of is a sex-core triple-channel “Gulftown”.......Maybe the mobile segment will go from 2 to 6 cores on the 32nm, or perhaps we’ll have to wait for 22nm to see quad-core.
Yeah, glad I'm not the only one a little confused. But from what I gather there is no leaked info on quad-core 32nm chips at all. This could change at the CES announcement in a few weeks time, and of course through the course of 2010.
If there are no quad-core 32nm chips for mobile, we won't have to wait for 22nm, Sandy Bridge is a new microarchitecture that will *start on 32nm* ~ and this is supposed to have 4 cores standard for desktop and mobile. As per the tick-tock strategy, it goes something like Core on 65nm, Core on 45nm, Nehalem on 45nm, Nehalem on 32nm, Sandy Bridge on 32nm, Sandy Bridge on 22nm... Here's the pictarz:
I suppose there has always been some overlap between the ticks and the tocks. Intel is showing it can keep up on a year-by-year tick-tock strategy. But I suppose they never promised each transition would be hard and fast, only they would introduce the tick or tock and then transition within... well, within whatever schedule they preferred.
But this "tick" of 32nm seems, at least starting out in 2010, *conservative* given it is only dualcores and they have yet to mention how they are going to transition the "big boys" of Bloomfield, Lynnfield and Clarksfield down to 32nm... if at all. If Intel decides that Bloomfield and Lynnfield are top-end desktop chips with decent enough power and thermals (which they are), they might not bring them to 32nm at all. If Intel decides that a quad-core mobile chip is not really needed in mainstream markets, they might not bring Clarksfield down to 32nm at all.
So Intel in all of 2010 could play the strategy of having the higher-end chips produced at 45nm for the enthusiast/workstation markets, and leave the 32nm chips for mobile, thin-and-light, and slim-desktop, mainstream-desktop segments. I guess it also depends on how many 45nm fabs Intel would keep running and what schedule there was for phasing out 45nm and bringing in 32nm. Additionally, since at 45nm Intel has generally a clear lead (though a higher price) over AMD in terms of performance, in both desktop and mobile, Intel may not be in a big rush to transition *everything* to 32nm by end of 2010.
In other words, unlike what I was assuming, in 2010 there may be no 32nm quad-core mobile CPU.
That is, in fact, entirely possible. Exactly zero leaked roadmaps show 32nm quad-core desktop or mobile processors from Intel in 2010 (or prior to the next "tock", Sandy Bridge). Just dual-cores (running four threads with hyperthreading) and hexa-cores.
There are some quad-core 32nm Xeons listed, but it's not clear if they will be made that way or will be hexas with two cores disabled. It's a big, complex chip, so that may be the case.
That is, in fact, entirely possible. Exactly zero leaked roadmaps show 32nm quad-core desktop or mobile processors from Intel in 2010 (or prior to the next "tock", Sandy Bridge). Just dual-cores (running four threads with hyperthreading) and hexa-cores.
it seems like everytime Intel is in the lead technology wise they mlk processes and drag out moving tech forward. It would be good to see vendors adopting a competitive AMD solution to keep Intel on it's toes. Of course this isn't the case right now so we are stuck with Intel draging ass.
There are some quad-core 32nm Xeons listed, but it's not clear if they will be made that way or will be hexas with two cores disabled. It's a big, complex chip, so that may be the case.
the flip side here is that we simply haven't heard about all of Intels plans. Frankly though Atom is a sign to me of Intels lack of commitment. Atom is the last place we should be seeing yet another spin at 45nm. It is like they are just giving up on competeing with ARM.
Dave
In other words, unlike what I was assuming, in 2010 there may be no 32nm quad-core mobile CPU.
It doesn't look like there will be 32 nm quad-core Westmere.
But this "tick" of 32nm seems, at least starting out in 2010, *conservative* given it is only dualcores and they have yet to mention how they are going to transition the "big boys" of Bloomfield, Lynnfield and Clarksfield down to 32nm... if at all.
Bloomfield is going to coexist with Gulftown in 2010.
it seems like everytime Intel is in the lead technology wise they mlk processes and drag out moving tech forward. It would be good to see vendors adopting a competitive AMD solution to keep Intel on it's toes. Of course this isn't the case right now so we are stuck with Intel draging ass.
the flip side here is that we simply haven't heard about all of Intels plans. Frankly though Atom is a sign to me of Intels lack of commitment. Atom is the last place we should be seeing yet another spin at 45nm. It is like they are just giving up on competeing with ARM.
Dave
Jeez, Sandy Bridge is Q4 2010. How much foot dragging can their be if Intel is still more or less on their tick-tock schedule. It's aggressive as hell and nice after how poorly they were executing just a few years ago.
Atom is at 45nm because it's not just mobile but also cheap. They'll go 32nm with Cedarview in 2011 which is about what their roadmap said in 2008. They haven't given up competing with ARM given they aren't really competing with ARM yet. That's like saying that ARM has given up on competing with Intel because there's no quad-core Cortex A9 yet.
They're going to converge on the netbook market for real right around when the A9 goes from dual core to quad and Intel goes Moorestown.
it seems like everytime Intel is in the lead technology wise they mlk processes and drag out moving tech forward. It would be good to see vendors adopting a competitive AMD solution to keep Intel on it's toes. Of course this isn't the case right now so we are stuck with Intel draging ass...
Jeez, Sandy Bridge is Q4 2010. How much foot dragging can their be if Intel is still more or less on their tick-tock schedule. It's aggressive as hell and nice after how poorly they were executing just a few years ago...
I agree with both of you, in a sense... Intel still wants to show it can "execute", as you say. So we will probably see Sandy Bridge on time and it should be impressive. But with Intel's dominance and Nehalem architecture proving to be really formidable against AMD, Intel is now enjoying the best of both worlds. They will milk 45nm for all it is worth while positioning Westmere as they see fit.
Intel is actually really strongly positioned in 2010, with their more-or-less locking out of other GPU and chipset manufacturers continuing at pace.
Maybe I'm getting old but I don't see needing a Core i7-level of CPU performance in my laptop or desktop in 2010. Most consumers and businesses also wouldn't. With the exception of a decent triple-core or quad-core CPU in a laptop at a Core i5-level of performance with great battery life.
What I am interested in and would like to see, more importantly, is that 250GB quality SSD at sub-$300 prices. That's the main bottleneck anyway in daily usage.
Additionally, I would like to see an entirely new form factor, ie. something like the tablet that will take us out of this keyboard-desktop-laptop paradigm.
See, that's the problem of too-dominant a company... for example, Intel and TSMC -- they can pursue their agenda and shape market demand around what they produce. If we had more competition, perhaps, we'll see $200 250GB SSDs, triple-core or quad-core in laptops, great ARM-based tablets, and a lot more GPUs from AMD and Nvidia at 32nm and below.
On that point, does anyone think that triple-core in a laptop is actually a great intermediate step for 2010? In desktops the differential is not so great, you'd go quad-core over triple-core in most cases. But for laptops, it would be a great transition CPU. Something like a Phenom II X3 or Athlon II X3 at 32nm, alongside a 40nm Mobility Radeon 5770 1GB VRAM... would be... sweeet. *sigh*
...They're going to converge on the netbook market for real right around when the A9 goes from dual core to quad and Intel goes Moorestown.
I wonder what the "netbook" market would be like at that time. Again, maybe I'm getting old, but Windows or OS X at anything less than 13" is totally not useable for me. It's not like I have big hands or poor eyesight either, I'm 31, 5'7" and don't wear glasses.
Perhaps a personal preference, but I say to Intel's Atom, "show me the useability!".
it seems like everytime Intel is in the lead technology wise they mlk processes and drag out moving tech forward. It would be good to see vendors adopting a competitive AMD solution to keep Intel on it's toes. Of course this isn't the case right now so we are stuck with Intel draging ass.
the flip side here is that we simply haven't heard about all of Intels plans. Frankly though Atom is a sign to me of Intels lack of commitment. Atom is the last place we should be seeing yet another spin at 45nm. It is like they are just giving up on competeing with ARM.
Dave
I think a lack of 32nm quad-cores, if that is true, indicates more the limitations of Intel's R&D budget and manufacturing capacity. Their resources are simply not unlimited, so they are focusing on their low-end volume seller and their high-end profit maker. The middle gets squeezed into 2011.
I think a lack of 32nm quad-cores, if that is true, indicates more the limitations of Intel's R&D budget and manufacturing capacity. Their resources are simply not unlimited, so they are focusing on their low-end volume seller and their high-end profit maker. The middle gets squeezed into 2011.
Isn't it weird how the low-end would do fine with Core 2 Duos?
Core i3 and i5. Your new best friends in 2010.
[soup nazi] No quadcore for you![/soup nazi]
Us mid-end users I guess will benefit in 2011 or so with Sandy Bridge 32nm quadcores.
I know, I'm contradicting myself, I said previously I didn't need quadcore. Though a triple-core would be a nice tasty nibble.
I still can't wait to see the mobile Arrandales. This may be the year for a new laptop for me. MBP, MBA ? I can't wait to see what Apple does with these machines this year.
As soon as they do, I'll be set. Otherwise by March / April I'll look into an E6500.
Definitely looking forward to next version. I have never purchased a mac before but have run a hacintosh briefly to see if I'd like it. I'm ready to buy a 15" MBP as soon as Apple offers something over the currently horrible screen res. Something that does full 1080 would be nice. Could care less about the processors really.
As soon as they do, I'll be set. Otherwise by March / April I'll look into an E6500.
Going from 900 to 1080 isn?t much, but I hope they don?t change it until they can offer full resolution independence as It?s already getting too small to read. Windows is much better at scaling than OS X but I hope Apple finally gets it done for 10.7 despite advertising it at least back to 10.5.
Isn't it weird how the low-end would do fine with Core 2 Duos?
Core i3 and i5. Your new best friends in 2010.
[soup nazi] No quadcore for you![/soup nazi]
Us mid-end users I guess will benefit in 2011 or so with Sandy Bridge 32nm quadcores.
I know, I'm contradicting myself, I said previously I didn't need quadcore. Though a triple-core would be a nice tasty nibble.
But the quad-core mobile Nehalem arch. processors have already been made. They're 45nm, but they're out on the market. And as for the low-end, Core 2 derived chips will still make up the majority of Intel's sales by volume for the next year.
However, smaller manufacturing process = more chips per wafer = less cost per CPU to Intel. So it's in their best interest to move their best sellers to 32nm first.