Intel iMac CPU Swappable?

[jhu]

Let me expand my post:


-> 533 MHz Celeron (fastest chip in a SEC (Slot I) cartridge that supported the 66MHz bus)

😉

heh, that means you weren't looking hard enough. you could get socket 370 to slot 1 convertors for the coppermine and tualatin celerons. of course, you'd still be stuck on a 66 mhz bus.

 

[AidenShaw]

heh, that means you weren't looking hard enough. you could get socket 370 to slot 1 convertors for the coppermine and tualatin celerons. of course, you'd still be stuck on a 66 mhz bus.

I saw those from Asus and some other companies...but decided to stick with stock, warrantied items. (Plus, the price of the "dead end" Slot 1 Celeron was great.)

266->533 was a big enough jump, the next step up was to a P4.

I prefer to do one or two substantial upgrades, rather than replacing parts as soon as something slightly better comes out. That's why I'm keeping the 2.6GHz/400MHzFSB system until I can replace it with a dual or quad system - this guy surfs plenty fast (and I can always use my dual Xeon or one of the 4-ways or 8-ways at work for heavy duty stuff). BTW, the dual Xeon (Compaq) started life as a 1.8GHz, and was bumped to 2.8GHZ with HT when that socket went EOL.

Also, I've never touched a jumper or firmware for any of these changes - I simply pulled the old CPU(s) out, and put the new one(s) in.

I also prefer to put my money into I/O upgrades, such as U320 SCSI with multiple 15K drives. In most daily use, fast disk I/O is as important as a fast processor. It makes the finder "snappier"! 😉

 

[milo]

As far as I know I cannot even install a 3 Ghz P4 (or any 64 Bit P4), because it wohld need faster FSB or other sockets and it is not supported by the BIOS. I could install a 2.6 Ghz but that would not be worth the money.

You're in a situation that's not conducive to upgrading because your machine was on the tail end of that particular socket. End of it's life cycle.

But that doesn't make upgrading in general a bad proposition, you just have to plan for it. Start with a MB that has a socket that's early in its lifetime and will likely be supported for at least another year or two. It's great for the user, you can get a reasonably fast, reasonably cheap CPU to start, then after a couple years you can replace it with something quite a bit faster for pretty cheap.

 

[vgermax]

(It's Banias/Dothan, by the way)

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2627

The Yonah socket is still a 479-pin interface, however the pin-out has been changed once more, and of course Yonah won’t even physically fit into any current Pentium M motherboards.

Yonah (left) vs. Dothan (right) - Note the different pin-out

My bad... Banias/Dothan, Napa/Sonoma... dang codenames. Complete chipseet versus the CPU. Anyhow, true about the pin configuration. I had read that but it was lost in the midst of countless other useless factts. It came up in a lot of reviews of Pentium M desktop boards.

 

[steveh]

Considering how Macs have always had VERY limited support for processor upgrades, this is very interesting

Except for most of the PCI tower Macs, and the G3 and G4 desktop machines; which covers the upper end of the Mac line for the better part of ten years, you mean?

 

[jhu]

Except for most of the PCI tower Macs, and the G3 and G4 desktop machines; which covers the upper end of the Mac line for the better part of ten years, you mean?

i think what he meant by "limited" was no direct drop-in replacement as in the x86 world. in the mac world, upgrades are a hassle because of a lack of standard sockets.

 

[Val-kyrie]

Merom FSB

Intel just announced that the new Santa Rosa Platform for the Merom chip will have a FSB of 800 MHz, but this won't be released until March, 2007, well after the initial introduction of Merom.

See this link for more information.

The reason I am waiting for the next revision is to get ahold of a Merom chip. Even if they are pin compatible (yonah and merom) won't the initial lack of fast interconnect technology be revised by the time Merom is released. I don't think that Intel is so gutsy as to rely on antiquated bus architecture for an entire year. I bet (hope) they will have some CSI implementation when Merom is released, thus making a chip-swap between yonah and merom much less beneficial. Personally I am hoping that the intel northbridge will disappear similar to AMD's at which point I will buy a new power book.

Either that or I am wrong.



Deimo seems to think the latter.

 

[mongoos150]

It'd be amazing if I could put a merom chip into my iMac when they debut...😀

 

[~Shard~]

It'd be amazing if I could put a merom chip into my iMac when they debut...😀

From what I've read they are pin-compatible with Yonah, so that would be pretty sweet! And they'd even use less power too, from what people have been saying in this thread. There's still that EFI/BIOS potential issue though... it will be interesting to see how much flexibility Apple allows for, either by design or by "policy". 😉 😎

 

[cb911]

this is very interesting...

it would be great if you can swap in some new proc's. especially or the iMac (and Mac's in general) given their very limited history of upgradability.

oh - and XLR8 has a new look. looks good too! 😀

 

[mongoos150]

oh - and XLR8 has a new look. looks good too! 😀

😕

 

[~Shard~]

😕

Why, XLR8yourmac, of course! 😉 😀

 

[milo]

Except for most of the PCI tower Macs, and the G3 and G4 desktop machines; which covers the upper end of the Mac line for the better part of ten years, you mean?

Upgrading these machines was less simple and more expensive. Sure, you could upgrade but it took a third partry to make it possible.

Being able to simply swap out a CPU *is* far less limited. Not to mention that it could potentially be available in every mac, instead of being at the mercy of what upgrade manufacturers could cobble together.

 

[shamino]

Um yes it does. I've done this on several P3 laptops. As long as the BIOS supports the CPU speed and the clock speed it should work perfectly fine.

And even if it doesn't, you are sometimes OK.

It's been a while since I built a PC, but the last time I did, the clock speed is a function of the FSB speed and a multiplier. The multiplier is configured by 3 or 4 pins on the chip which are toggled on/off either with motherboard jumpers or firmware settings.

The problem with a processor that the BIOS doesn't recognize is that the mapping from the multiplier-pins to actual multipliers isn't the same for every model chip. (e.g. on-off-on may be a 3.5x multiplier on one chip and a 6x multiplier on another.)

If the BIOS supports the chip, it identifies it (via the CPUID instruction, I assume) and will use a lookup table to present you with a convenient list of available multipliers. It will also use this lookup table to figure out what to display for a clock speed on the POST screen.

If the BIOS doesn't support the chip, then you may not see an accurate list. You'll have to do your homework to figure out what it thinks your chip is so you can select the multiplier that sets the correct pins. And the speed displayed by POST may end up being wrong.

Of course, there are sometimes other things in the correct BIOS (like power management schemes) that can come to bite you even if you get the clock speed configured correctly.

I did this years ago (back in the 486 days) to put an AMD 120MHz chip (40MHz x3) into a motherboard that didn't support it. I had to figure out the jumper settings on my own, and the BIOS reported 100MHz (33MHz x3) on the POST screen, but it worked and benchmarks showed that it really was running at 120MHz.

On more modern systems (P2 and P3), I didn't have to do this, because BIOS updates were always available. (It's much easier to update flash than ROM.)

But having said all this, it's probably something I'll never have to do again.

Modern PC motherboards almost always have BIOS updates available to keep pace with all compatible chips. And there's no way I'd risk voiding my AppleCare to try it on a Mac.

 

[danman]

Useful upgrade path already...

Intel has already announced that the current Core Duos (Yonah) successor (Merom) will be pin-for pin compatible, and around 30% faster clock for clock.

..doh.. been said.