Just to note, "pin-compatible" doesn't always imply compatible. There are occasionally differences in power requirements and such that prevent the chips from working properly.
-Clive
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Intel Core 2 Quad in January
- Thread starter MacRumors
- Start date
- Sort by reaction score
Just to note, "pin-compatible" doesn't always imply compatible. There are occasionally differences in power requirements and such that prevent the chips from working properly.
-Clive
Before you all get your hopes up too much I just thought I'd point up that this is a processor for high end desktop pcs. The iMacs are technically made up of laptop components and use laptop processors which have a different pin configuration. We're going to have to wait for Core 2 Quad laptop processors before we see any quad cored iMacs, I'm afraid. 
Before you all get your hopes up too much I just thought I'd point up that this is a processor for high end desktop pcs. The iMacs are technically made up of laptop components and use laptop processors which have a different pin configuration. We're going to have to wait for Core 2 Quad laptop processors before we see any quad cored iMacs, I'm afraid.
Also known as Penryn
Before you all get your hopes up too much I just thought I'd point up that this is a processor for high end desktop pcs. The iMacs are technically made up of laptop components and use laptop processors which have a different pin configuration. We're going to have to wait for Core 2 Quad laptop processors before we see any quad cored iMacs, I'm afraid.
Ah... so would this chip be pin-compatible with Conroe and not Merom?
-Clive
Noted. I haven't really read the review in much detail, but I believe Anandtech did a successful drop-in and didn't run into any major issues.
Regardless, I would be completely surprised if Apple DIDN'T introduce an 8-core Mac Pro at MWSF.
At least as an option, yes. I don't think they'll sacrifice the entire line for the grossly expensive chips. They may drop the price of lower-performing MacPros and further plug up the hole between them and the iMac.
-Clive
I am always amazed that someone, somewhere always rates a story with a faster CPU for less money to be a negative.
Is it really that surprising? This processor doesn't appear to be destined for any Mac. The iMac, Mac Mini, MacBook and MacBook Pro platforms all use Intel's mobile variants of the Core processors while the Mac Pro and Xserve platforms use Intel's workstation variants. The desktop variants have traditionally remained unused and I don't see this changing any time soon. People rate positive or negative for many different reasons, some people vote negative if they don't like the news, some people vote negative if they don't believe its true and some people vote negative if they think it doesn't bear any relation to Apple and Apple's hardware.
Grossly expensive? From what I read at AppleInsider the price of the incoming 2.33 Quad processor is the same that Apple was paying for the 2.33 Duo processor last year - essentially, they could replace the whole line with Quad Core chips and keep the same price points.
I could be mistaken.
A more interesting question pertains to the timing and coordination of these breakthrough events that are slated for 2007:
- Introduction of Leopard
- Santa Rosa chipset (including NAND memory for booting and prefetching)
- Utilization of Penryn processor (laptop Core 2 Quad) in the iMac line.
My feeling is that Apple would prefer to accompany Leopard with a strong iMac spec bump. Especially in case Leopard supports NAND for booting or prefetching, it would be essential to have at least one Mac model available immediately to demonstrate that feature. Which means that we should expect to see Santa Rosa-based iMacs in late spring (with slightly faster Core 2 Duos), with Penryn arriving no earlier than September 2007.
- Introduction of Leopard
- Santa Rosa chipset (including NAND memory for booting and prefetching)
- Utilization of Penryn processor (laptop Core 2 Quad) in the iMac line.
My feeling is that Apple would prefer to accompany Leopard with a strong iMac spec bump. Especially in case Leopard supports NAND for booting or prefetching, it would be essential to have at least one Mac model available immediately to demonstrate that feature. Which means that we should expect to see Santa Rosa-based iMacs in late spring (with slightly faster Core 2 Duos), with Penryn arriving no earlier than September 2007.
Hehe, Moore's law strikes again. It shall be noted that the clock speed is the same tho, but now you get 4 processors for the price of 2. I wish hard drives, RAM and removable media would evolve as fast. DVDs (even dual/double layer) barely have any space for backups or any kind of libraries on them, blueray might, but by the time those get main stream, they'll be outdated already. Technology sucks...
That's a good assessment. Intel haven't precisely specified when Penryn's due yet, have they? Its a shame they're not releasing the mobile and desktop variants at the same time like they did the Core 2 Duo, that was a wonderful bit of timing for Apple.
Okay, I should have clarified:
2 dual core chips >= 1 quad core chip
however,
2 dual core chips << 2 quad core chips
If Apple were to replace, chip for chip, all dual cores with quad cores, the price for processors used in the MacPro would nearly double.
If you're talking about replacing 2 dual cores for one quad core, then the price should decrease.
-Clive
Yes - gimme 30" iMac w/C2Q, up to 8GB RAM and nVidia 8900GTX 512MB video, DL BluRay burner, dual HDD bays and a DVI-DL *input* so it can be used as a 30" display by another system as well as a stand-alone computer.
I would so buy that tomorrow and probably twice.
FW800 and eSATA would be nice too, while I'm dreaming... I bet they could even fit a PCI-E expansion slot in there too.
Sorry, I was wrong. The quad cores are more expensive clock-to-clock. But not quite twice as expensive.
From AppleInsider:
That would mean that Apple could put together an 8-core 2.0ghz setup for the same price as a 4-core 2.66 setup, unless I'm misreading it.
To put things into perspective for you, my company still backs up data on 200 GB tapes. It's the most cost effective solution for them... and would be for a lot of backup solutions... the only con is that it's slower than all hell, winding through 200 GB to get to a certain spot in the tape. That's why they only use it for backup.
-Clive
To put things into perspective for you, my company still backs up data on 200 GB tapes. It's the most cost effective solution for them... and would be for a lot of backup solutions... the only con is that it's slower than all hell, winding through 200 GB to get to a certain spot in the tape. That's why they only use it for backup.
...Same here. Tape is the only real way to do massive data archives. I wish it weren't so, but it is. We have a fairly large SAN and have an evolving "live" backup built out of RAID storage. Then tape archives are made at regular intervals of that live backup for archival. It's pretty transparent other than the need to swap tapes, but the system tells us when to do it, usually twice a week and takes a few minutes.
By the time BluRay gets here with all 12 of its theoretical layers (they've only shown 4 layers in lab demos), it still will not have caught up to tape in capacity. ...Really sucks. I hate tape, but I have to love it because it's the only way. Hard drives aren't designed for long-term shelf storage and optical media degrades over time just as magnetic media does, often worse. And it has nowhere near the capacity. Tape systems are expensive to buy into in many cases, but they pay off over time. So, if you don't need tape or if running the numbers it looks too expensive (which means you don't need tape), then consider yourself lucky. For the rest of us and all those poor SOB's at companies much larger than mine, I pity you.
I will say that having the SAN with live backups on RAID is awesome though. This is the way many datacenters have gone lately. It's very reliable and I haven't had to pull anything from a tape in nearly 2 years since setting it up. The tape archives are just there for the extra fall-back cushion and that we can keep them off-site in case of flood, fire, or general human stupidity trashing our localized data.
I do believe Moore's law has met its demise. Simple physics results in processor speeds and capacities being limited by nanometer mask limits and power blow-thru limits. Hence the move to multi-core as a temporary workaround until we get "beyond silicon".
The benefits of multi-core are real even to so-called normal users. Have you bothered to look at the system preference pane showing tasks? Dozens of them. If you have 4+ cores (8, 16, 32?) it spreads them. ONE program even has many tasks. This is the world we live in now.
Some "tasks" require a lot of CPU time. Helps to have a spare laying around. Yes it still takes time. But so does dying!
Computers are a tool to do tasks. The evolution of the tasks themselves are more impressive than the evolution of the hardware, which is itself quite impressive.
In the minutia we have Robson, Santa Rosa, C2Q, and quite notably,
http://www.popularmechanics.com/technology/industry/4205068.html?do=print
PRAM
Rocketman
Hello? Anyone there? Anyone?
I take it that "PRAM" is different from the "P-RAM" that we're all used to (or at least those of us who used the Classic OS)? If not, it's sad that it's taken this long to catch on.
-Clive
It would be nice if they'd just figure out how to spawn their process in a single thread rather than hogging the entire cpu when they perform a background task like checking for the need for updates. (Symantec, you listening?)