Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
You two have really been going at it here. Obviously you both have substantial knowledge about the CPU issues here, but may I kindly suggest you tone down the rhetoric a bit?

Seems you could be making your points without quite so much vitriol... Might make others feel more comfortable using this as a helpful forum for exchange of ideas. Just a thought...

Best to stay out of the middle unless you want to get pissed on.
 
Everything I listed was discrete with the exception of Intel's Graphics.

I got that. I was asking about the chipset because nVidia cannot make any and they got out of that business.


MacBook Pros CPU/GPU

Early/Mid/Late

E2006 C2D Yonah 31W + ATI Mobility Radeon X1600 30?W
L2006 C2D Merom 35W + ATI Mobility Radeon X1600 30?W
M2007 C2D Merom 35W + NVIDIA GeForce 8600M GT 22W
L2007 C2D Merom 35W + NVIDIA GeForce 8600M GT 22W
E2008 C2D Penryn 35W + NVIDIA GeForce 8600M GT 22W
L2008 C2D Penryn 35W + NVIDIA GeForce 8600M GT 22W
L2008 C2D Penryn 25/35W + NVIDIA GeForce 9600M GT 23W
E2009 C2D Penryn 25/35W + NVIDIA GeForce 9600M GT 23W
M2009 C2D Penryn 25W + NVIDIA GeForce 9600M GT 23W
M2010 i5/i7 Arrandale 35W + NVIDIA GeForce 320M 35W

This might be inaccurate.


The QC SB is 45/55W.

Are you sure that is not SB 35W + IGP 10W on the same dye that makes the 45W?
Arrandale has 35W with the IGP.
The DC SB beside the lowpowered at 17 W are already 25/35W, how do you get a QC at 35W IGP inclusive?
 
Last edited:
Ok, explain the physics to me. How does an overheating GPU also overheat the CPU?

Both overheat independently. The point is that the one with 20W lower TDP also overheats (the CPU). This means a 45W CPU in a MBP is unlikely.

The physics involve temperature gradients. Take a look at the thermal gradient between a die and a heat sink. There is a huge jump at that juncture. So, getting heat out of a chip is remarkably difficult.

So, one of you is right in that the better you spread the heat sources around within a laptop chassis, the better it is.

The other one of you recognizes that the total produced heat still has to be considered when designing a laptop with a low thermal conductive mass, restricted internal air movement and only a narrow slit to exhale the hot air.

So, both of you have a valid argument in that a MBP needs to be carefully designed and both factors (along with several others) needs to be weighed in any overall design.

I am very amazed that while my 2006 Intel MBP gets hot enough to burn my lap, it continues to run for hours on end at that temp and hasn't failed in all these years.

Peace bros.
 
LOL, that was your whole counter argument for my mid 2009 MBP example. You went on your 10 quote rampage about on die IGP's and on main board IGP's and never went anywhere with it. You never provided any source for your claims.
They were self contained rebuttals to your points. It's common knowledge, but if you want sources.

First paragraph details on die IGP http://en.wikipedia.org/wiki/Arrandale
Here is an article on intel's GMA IGPs that pre-date on die IGPs and are integrated into motherboards. This covers everything pre-Arrandale http://en.wikipedia.org/wiki/Intel_GMA

Well let's see. When two things are close together in a confined space... and one of those gets hotter, what does it give off? .... HEAT! And where does this spread? IN THE CHASSIS!

Well, two problems with this. Heat spread through the chassis (frame and body) would be convected to the outside air, not back into other components.

Why? This feeds into the second point. Air is a poor conductor of heat. This is also why heat is not going to magically jump to the CPU from the GPU directly. The logic board is out because it's mostly comprised of electrically non-conductive materials, which almost always implies thermal non-conductivity as well. This is why you have to have copper heatsinks to pipe the heat away. This is also a problem for the MBP because the CPU and GPU share a heatsink in the MBP.


The CPU is 35W while the IGP is 10W. Don't make me recite my mid 2009 MBP example again.
and the IGP, for SB, is on die. You cannot separate their TDPs because it's the same physical chip.


I got that. I was asking about the chipset because nVidia cannot make any and they got out of that business.


MacBook Pros CPU/GPU

Early/Mid/Late

E2006 C2D Yonah 31W + ATI Mobility Radeon X1600 30?W
L2006 C2D Merom 35W + ATI Mobility Radeon X1600 30?W
M2007 C2D Merom 35W + NVIDIA GeForce 8600M GT 22W
L2007 C2D Merom 35W + NVIDIA GeForce 8600M GT 22W
E2008 C2D Penryn 35W + NVIDIA GeForce 8600M GT 22W
L2008 C2D Penryn 35W + NVIDIA GeForce 8600M GT 22W
L2008 C2D Penryn 25/35W + NVIDIA GeForce 9600M GT 23W
E2009 C2D Penryn 25/35W + NVIDIA GeForce 9600M GT 23W
M2009 C2D Penryn 25W + NVIDIA GeForce 9600M GT 23W
M2010 i5/i7 Arrandale 35W + NVIDIA GeForce 320M 35W

This might be inaccurate.


The QC SB is 45/55W.

Are you sure that is not SB 35W + IGP 10W on the same dye that makes the 45W?
Arrandale has 35W with the IGP.
The DC SB beside the lowpowered at 17 W are already 25/35W, how do you get a QC at 35W IGP inclusive?

That's correct, the total TDP is 45W. The IGP, DMI and PCI express controller is responsible for about 10W of that.


The physics involve temperature gradients. Take a look at the thermal gradient between a die and a heat sink. There is a huge jump at that juncture. So, getting heat out of a chip is remarkably difficult.

So, one of you is right in that the better you spread the heat sources around within a laptop chassis, the better it is.

The other one of you recognizes that the total produced heat still has to be considered when designing a laptop with a low thermal conductive mass, restricted internal air movement and only a narrow slit to exhale the hot air.

So, both of you have a valid argument in that a MBP needs to be carefully designed and both factors (along with several others) needs to be weighed in any overall design.

I am very amazed that while my 2006 Intel MBP gets hot enough to burn my lap, it continues to run for hours on end at that temp and hasn't failed in all these years.

Peace bros.

I understand, I was being facetious.
 
Last edited:
asdf542 is in a time out so he/she can't see the forums. Guess we'll never see the end of this argument.
 
My original Unibody Macbook (non pro but looks like what they now call pro) is hanging in there still, and I don't have any desire to upgrade yet. I am actually impressed at how long this computer has been lasting me.

Although those i7s seem pretty nice....

Lol, I'm "hanging in there" with the late '06 2.33ghz MBP. Granted, I've got a dedicated GPU and you dont, but at least you can address 4GB of ram, while I only see 3 :(

I'll be eager to buy one of the Sandy Bridge i7's as soon as they come out.
 
asdf542 is in a time out so he/she can't see the forums. Guess we'll never see the end of this argument.

s/he can see them (even if just by logging out), but s/he can't respond. Given the insults, it's probably a good thing.

In any event, 35W is enough. Until Intel gets the TDP down, we're stuck with 2 cores.
 
What software right now lags on the current 2.8Ghz core i7 MBP?

Beside that I don't care about performance and I'm happy with my MBA C2D, the argument is not about the lag.

Most of people wants just the latest and greatest, works for cars, works for techs, even when they don't use them at the fullest, and sometimes is dumb.

However, there are many software that need the better, and for this I mean people that work with media; Final Cut, Logic, Maya etc. A better CPU reduce time of the extensive processing that these software go through, and off course if one just browse the internet won't benefit from either CPU/GPU.
 
What software right now lags on the current 2.8Ghz core i7 MBP?

I'd argue that even if no software currently benefitted from anything greater than a 2.8 GHz i7 (which isn't true anyway), it will eventually get there. Most people plan on using a computer for anywhere from 3-6 years so what it can handle now is irrelevant as the hardware will only become relatively weaker and weaker compared to what software might require in 3 years.
 
Most people plan on using a computer for anywhere from 3-6 years so what it can handle now is irrelevant as the hardware will only become relatively weaker and weaker compared to what software might require in 3 years.

That is a good point. My current laptop is about 3 and a half years old and I do expect to upgrade to a newer machine sometime in the next couple of years.

When I bought it, I thought it was really fast. Now I really long for a faster machine.

Buying Macs isn't cheap and there's many who want their purchases to last a long while.
 
Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_1 like Mac OS X; en-us) AppleWebKit/532.9 (KHTML, like Gecko) Version/4.0.5 Mobile/8B117 Safari/6531.22.7)

Ooooh!! I'm holding my breath!! i3 13", OSX on SSD, a few hours with Dreamweaver then THE WORLD!!!!! *Nyahahaha*
 
Isn't it best to wait for Lion

With the arrival of Lion expected late summer this year wouldn't it be better to wait for this before upgrading?

Lion is likely to be a more expensive upgrade than Snow Leopard, which was cheap compared to previous os updates. So as I'm looking to upgrade from my 06 MacBook to a MBP this year, would I not be better off waiting for Lion to come out and be a standard installation than buying as the MBPs are released and then paying approx £100 for the OS update.

Prices aren't likely to rise once Lion is the pre installed OS are they?
 
Same here

but id be happy if they added iX processor, dedicated GPU, SSD (like the air's, but with 2nd HD for storage) and No DVD.

i can dream :)

I have been thinking about the coming updates .... I wonder, will Apple wait to remove the optical disc drive from the MBPs until the hardware refresh after OS X 10.7 (Lion) is released? If Apple intends to give users a quick OS boot from an SSD and data storage on a mechanical hard drive, then it might make sense to wait until 10.7, especially if it needs to be built into the OS for user transparency/ease of operation. I realize many now have this setup via a mod, but this requires managing partitions with which the average user would be completely confounded. Thoughts?

[Edit: Unless Apple chooses to wait for chipsets free from the faulty SATA connectors, it seems there will be no dual-hard drive configurations in the upcoming refresh. From Laptop Mag:

Because the chipset error does not affect the first two SATA ports on the motherboard, Intel has agreed to resume shipments of the affected boards as long as OEMs promise not to use the other four ports. On notebooks, the only devices that attach to SATA ports are the hard drives, optical drives, and eSATA ports. Considering that most notebooks have just one hard drive, one optical drive, and no eSATA ports, most systems should be able to ship without a problem. Notebooks that have dual hard drives or eSATA ports can still ship as long as the vendor uses a third-party SATA controller card for those extra ports.

Since I doubt Apple will chose to utilize a third-party SATA controller for one refresh at this late time, I highly doubt we will see dual-hard drive configurations unless the refreshes are delayed until May/June. Consequently, I suspect Apple will retain the ODD at least for this next refresh.]
 
Last edited:
I thought all macs never had such problems, since they're made and designed by non human magical people. Nothing new here, a lot of Apple products have hardware issues, a lot, but that gets filtered out by the fan boys who claim that things like this only happen to PC. I've had 3 PC, 2 laptop, and 1 macbook, and only the MacBook has given me problems. Big and significant problems with that White MacBook model.
 
My.. can someone stop chrmjenkins and asdf542? Getting tired of skipping the conversations...

EDIT:
asdf542 is in a time out so he/she can't see the forums. Guess we'll never see the end of this argument.

Ohh GOOD!!!!!

I am waiting for the next generation MBP's, but I guess I'll wait until the Lion OS is released. Ivy bridge: well.. You can't wait for everything :D
 
Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

diamond.g said:
chrmjenkins said:
They are 25W chips with a 10W DMI/IGP.




Those were integrated on the mainboard, not the die. The first die integrated GPU was the one in Arrandale.
Arrandale is on chip, not on die. Sandy Bridge is on die.

Sorry, is there any difference?
 
With the arrival of Lion expected late summer this year wouldn't it be better to wait for this before upgrading?

Lion is likely to be a more expensive upgrade than Snow Leopard, which was cheap compared to previous os updates. So as I'm looking to upgrade from my 06 MacBook to a MBP this year, would I not be better off waiting for Lion to come out and be a standard installation than buying as the MBPs are released and then paying approx £100 for the OS update.

Prices aren't likely to rise once Lion is the pre installed OS are they?

I was in the market for my first MBP and waited until Leopard before pulling the trigger - as a result I ended up with a Santa Rosa 17". Shortly thereafter the unibodies came out. I should either have waited a bit longer, or not waited at all. (Not waited at all because waiting resulted in me not having an MBP for awhile, and the entire design changed right after I got mine anyway.)

True about saving a little money on the OS upgrade, but the amount is small compared to the price of the MBP, and many people now have multiple macs anyway that would need upgrades, and would just buy a family pack (if they are honest).
 
Arrandale is on chip, not on die. Sandy Bridge is on die.

Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

Sorry, is there any difference?

Yes, there is, what he said is more technically correct. The arrandale IGP is actually a 45nm die integrated onto the same package (the enclosure that houses the CPU and will have the heatsink applied to it). The arrandale CPU itself is 32 nm, so it's obvious they can't be on the same physical die. With Sandy Bridge however, it's all 32 nm and part of the same die that will actually be diced from the wafer.
 
Yes, there is, what he said is more technically correct. The arrandale IGP is actually a 45nm die integrated onto the same package (the enclosure that houses the CPU and will have the heatsink applied to it). The arrandale CPU itself is 32 nm, so it's obvious they can't be on the same physical die. With Sandy Bridge however, it's all 32 nm and part of the same die that will actually be diced from the wafer.

Just making a picky point here: if the process is 32nm, it's perfectly possible to have devices with 45nm drawn width. Of course the reverse isn't true. So it would be fine to have a CPU with minimum feature size of 32nm integrated with a IGP with minimum feature size of 45nm.
 
Does anyone think there will be a shorter update time between this upcoming Sandy Bridge release and the Ivy release due to the somewhat longer span between the last update and the Sandy Bridge soon to arrive? I don't use my computer for hardcore tasks except for occasional music production using Native Instruments "Maschine" But I am debating on waiting another cycle for the Ivy even though I've been holding off for this upcoming release.
 
Question for chrmjenkins

http://www.notebookcheck.net/Review-Acer-Aspire-7750G-Notebook.45861.0.html


This is a link to www.notebookcheck.net who review allsorts of new hardware. It is a review of the new Acer Aspire 7750G. This machine has the new Sandy Bridge i7-2630QM quad core processor and an AMD Radeon HD 6850M.

You will note towards the end of the review (under the heading Battery Runtime) that in idle the system consumes 12.9 watts. They conclude that the total system is more frugal than the previous i7-740 systems (are they correct in that assumption?).

The reviewed machine also does well for heat generation in the review, and also the review comments on it being thin for a cheap windows machine. Given how much Acer has managed to pack into this case, and how well it all comes together, might not Apple also be able to use the same cpu and gpu?

I appreciate Apple is more likely to go with a gpu from nvidia maybe a 420M or a 425M or even a 430M, these being closer to the 330 presently used. I doubt they would go for the GTX 460M which appears to be the more direct competitor to the 6850M.

But if Acer can get the quad Sandy Bridge cpu and a high end Radeon to work in a slim form why can't Apple use the same cpu and an nvidia gpu in the new MBP?

I note that the 330 gpu in present MBP's has been specifically down clocked by Apple compared to it's clock speed in many other PC manfacturers laptops (apart from Sony who pull the same trick). This downgrades it's ability but uses less juice. Could not Apple use the same down clocking on a 430M to fit it in a MBP with a quad Sandy Bridge cpu?

Is it the fact that when under load the Acer machine jumps up to 104 watts usuage? And the fact therefore that the Acer needs a large power adaptor?
 
I thought all macs never had such problems, since they're made and designed by non human magical people. Nothing new here, a lot of Apple products have hardware issues, a lot, but that gets filtered out by the fan boys who claim that things like this only happen to PC. I've had 3 PC, 2 laptop, and 1 macbook, and only the MacBook has given me problems. Big and significant problems with that White MacBook model.

Who said anything about Macs having problems...?
The only problem is with Intel's Sandy Bridge.
 
Just making a picky point here: if the process is 32nm, it's perfectly possible to have devices with 45nm drawn width. Of course the reverse isn't true. So it would be fine to have a CPU with minimum feature size of 32nm integrated with a IGP with minimum feature size of 45nm.

Thank you, I did not know that. I simply assumed because that they didn't that it wasn't possible.

Why wouldn't they? My guess is that it could be yield concerns or library incompatibility. I know that 32nm is Intel's second generation high k gate process, and rather than re-spin the IGP design it may have been advantageous to stick with the 45nm design as is.

Question for chrmjenkins

http://www.notebookcheck.net/Review-Acer-Aspire-7750G-Notebook.45861.0.html


This is a link to www.notebookcheck.net who review allsorts of new hardware. It is a review of the new Acer Aspire 7750G. This machine has the new Sandy Bridge i7-2630QM quad core processor and an AMD Radeon HD 6850M.

You will note towards the end of the review (under the heading Battery Runtime) that in idle the system consumes 12.9 watts. They conclude that the total system is more frugal than the previous i7-740 systems (are they correct in that assumption?).

The reviewed machine also does well for heat generation in the review, and also the review comments on it being thin for a cheap windows machine. Given how much Acer has managed to pack into this case, and how well it all comes together, might not Apple also be able to use the same cpu and gpu?

I appreciate Apple is more likely to go with a gpu from nvidia maybe a 420M or a 425M or even a 430M, these being closer to the 330 presently used. I doubt they would go for the GTX 460M which appears to be the more direct competitor to the 6850M.

But if Acer can get the quad Sandy Bridge cpu and a high end Radeon to work in a slim form why can't Apple use the same cpu and an nvidia gpu in the new MBP?

I note that the 330 gpu in present MBP's has been specifically down clocked by Apple compared to it's clock speed in many other PC manfacturers laptops (apart from Sony who pull the same trick). This downgrades it's ability but uses less juice. Could not Apple use the same down clocking on a 430M to fit it in a MBP with a quad Sandy Bridge cpu?

Is it the fact that when under load the Acer machine jumps up to 104 watts usuage? And the fact therefore that the Acer needs a large power adaptor?

Looking at that computer, it's actually thicker than the Envy 17 (1.33" vs. the 1.25" on the envy 17), so it's no surprise those components work in there. The 6850 is actually just a 5850 rebadge (sadly), so they have the same exact GPU. It's possible power consumption is slightly smaller (I've seen Nvidia rebadges vary slightly in performance specs that Nvidia quotes, so they may update screening specs in between runs).

As for the CPU, it's not secret that Sandy Bridge is better on power, and the idle consumption shows that.

You are correct in stating that the maximum load is what drives the design.

As for the GPU, I'd expect a GeForce 525M. It's rated TDP (25W) is right around what Apple commonly does at 23W. It's important to note that the next GPU, the 540M, is actually the same GPU with higher clocks at 35W. So, you still get the 96 shaders. Now, you might think, hey that's awesome, twice the shaders of the 330M. The problem is that they are radically different. One is a G92 shader while the 425M is the GF108 core. Performance should still be much better.

The problem with asking anything higher is that while the 17" would allow for more heat dissipation, all else constant in the current designs, Apple has never committed the effort to making the guts of the 17" differ from the 15" aside from ports. So, we're really limited by what they cram into the 15".

What is more interesting is what they decide to do with the 13". Board space is limited but they can no longer rely on the 9400M. They're going to have to rely on the Intel IGP or add a small discrete cheap to the board.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.