Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
gnasher729 said:
The only problem with Conroe vs. Merom in an iMac is the heat production. I am sure it is possible to keep an iMac with a Conroe cool enough. What I think is very hard to achieve is to keep it cool and quiet at the same time. I have never, ever _heard_ the iMac in my office, and that is really nice.

Of course it would be possible to put something into the Energy Saver Preferences like a "Keep Quiet" option. If selected, processing power could be cut down when the iMac gets too hot to be cooled down without making much noise. You would still have the potential to get full performance if you choose so.

Hmmm, it's a tough one. But I think the price vs. performance of Conroe being so much better than Merom will mean that Apple will try their best to cram one into the iMac. It would mean a more powerful computer that's cheaper for Apple to produce, and it is a desktop after all.

But you are right that the thermals of Conroe could be a problem. I'm not sure how much of a difference speed-step will make. Although it is the most power-efficient desktop chip out there, it consumes less power under load than Pentium D's at idle:

power_conroe.png


And it runs much cooler than the chips in the Macbook and Macbook Pro in normal computer cases:

conroe_temp_min.gif


25 degrees celcius with speedstep...

x6800_temp_max.gif


And 45 degrees at max clockspeed. And that's the 75W Conroe, the 65W one should be slightly cooler. Now I know this isn't in the iMac case, but if the chips run that cool in desktops then hopefully they shouldn't run too much hotter in an iMac...
 
Manic Mouse said:
Hmmm, it's a tough one. But I think the price vs. performance of Conroe being so much better than Merom will mean that Apple will try their best to cram one into the iMac. It would mean a more powerful computer that's cheaper for Apple to produce, and it is a desktop after all.

But you are right that the thermals of Conroe could be a problem. I'm not sure how much of a difference speed-step will make. Although it is the most power-efficient desktop chip out there, it consumes less power under load than Pentium D's at idle:

power_conroe.png


And it runs much cooler than the chips in the Macbook and Macbook Pro in normal computer cases:
conroe_temp_min.gif


25 degrees celcius with speedstep...

x6800_temp_max.gif


And 45 degrees at max clockspeed. And that's the 75W Conroe, the 65W one should be slightly cooler. Now I know this isn't in the iMac case, but if the chips run that cool in desktops then hopefully they shouldn't run too much hotter in an iMac...
Then all we're looking at is cranking up the current 180 watt power supply. I remember my iMac G5 2.0 GHz hitting 75-76º C at 100% load. The Rev. C iMac G5 was whisper quiet compared to my machine using the same 970FX chip. If Conroe doesn't break 45° C then it's not a thermal nightmare to put into the iMac. It's just a pain to power.
 
Ok, here are my predictions.

MacPro - woodcrest (available immediately, all dual duals, 3 levels starting at 2.0 GHz)
XServe - woodcrest (Available soon)
iMac - Conrow (It is the oldest Intel computer and has received 0 upgrades in 6 months. E6400 in the 17" and E6600 in the 20", we may see a small price decrease)
MacBook Pro - Merom (in 1 Month, Paris intro?)
MacBook - Yonah for now (Merom in about 4 months)
MacMini - Yonah (use existing 1.66 and 1.83 parts)

I would love to see the MacMini get Conroe parts.. it should reduce the cost. Heat may be an issue. The E6300 might work.

I would love to see a small tower, Video Card would be replaceable. Maybe room for a second 3.5" drive. Does SATA have the capability of driving 2 HDs from one controller. It would be similar size to the cube. Pricing about $100 more than the Mini. No expansion slots.

I've got to admit though, I really like my 1.66 GHz Mini. Small, quiet, and fast enough.
 
DVK916 said:
Allendale is not faster than Merom. Benchmarks show it is slower.
Links please. Both low end Merom and Allendale have the same amount of cache. Still, Allendale clocks higher and faster.
 
Eidorian said:
Links please. Both low end Merom and Allendale have the same amount of cache. Still, Allendale clocks higher and faster.


Who is talking about low end Merom. I am talking about higher end with 4mb of Cache.

I highly doubt apple would use a low end merom, when they can go with a higher end.

The MacBook and MacBook Pro will both get 4mb Meroms.
 
another update in Q1 2007? damn. So fast, but screw it, once the new MBP design is out, I'm ordering. I wish it was the 7th, but now most likly the 12 of September.
 
DVK916 said:
Who is talking about low end Merom. I am talking about higher end with 4mb of Cache.

I highly doubt apple would use a low end merom, when they can go with a higher end.

The MacBook and MacBook Pro will both get 4mb Meroms.

Just FYI: It seems that Merom will ship at speeds from 1.66 to 2.33 GHz, the 1.66GHz and 1.83 GHz have 2 MB cache, and 2GHz and faster all have 4 MB cache. Prices seem to be identical to the current Yonah prices (except the 2.33; there is no 2.33 GHz Yonah). So I would expect that all Yonahs will be replaced with Merom at same clockspeed, and that means 4MB cache except for the low end MacBook and possibly MacMini.
 
gnasher729 said:
Just FYI: It seems that Merom will ship at speeds from 1.66 to 2.33 GHz, the 1.66GHz and 1.83 GHz have 2 MB cache, and 2GHz and faster all have 4 MB cache. Prices seem to be identical to the current Yonah prices (except the 2.33; there is no 2.33 GHz Yonah). So I would expect that all Yonahs will be replaced with Merom at same clockspeed, and that means 4MB cache except for the low end MacBook and possibly MacMini.

Actually there is a 2.33 Yonah, but it just shipped recently.
 
Manic Mouse said:
But you are right that the thermals of Conroe could be a problem. I'm not sure how much of a difference speed-step will make. Although it is the most power-efficient desktop chip out there, it consumes less power under load than Pentium D's at idle:

The new version of Speedstep that is available from Yonah upwards seems to be quite good. The operating system can reduce the clockspeed, but at lower clockspeed it can also reduce the voltage, and power consumption is proportional to (clockspeed x voltage x voltage).

So my idea would be: A control that lets you set "maximum fan noise". If you need to render an hour of video, you should set it to maximum, if you are doing audio recordings with your Mac and want it silent you set it to minimum. That control also shows whether your clockspeed has been throttled, so if you prefer the Mac fast instead of quiet, you can have that as well. So you get the best of both worlds.
 
Just a thought:

If Apple had gone with AMD instead of Intel as many people demanded when the move to x86 chips was announced, wouldn't Apple and Steve Jobs look really stupid today?

Since Intel announced that their four core chips would be available in the last quarter 2006, there is another possibility now for Mac Pros: Just Conroe chips for the "low end" replacing the dual core G5s, and the quad core G5 remains the last PowerPC until it is replaced in the last quarter with a much cheaper quad core system based on Intel's Kentsfield quad core desktop chip. Depends on how urgent it is for Apple to finish the transition.
 
Eidorian said:
Then all we're looking at is cranking up the current 180 watt power supply. I remember my iMac G5 2.0 GHz hitting 75-76º C at 100% load. The Rev. C iMac G5 was whisper quiet compared to my machine using the same 970FX chip. If Conroe doesn't break 45° C then it's not a thermal nightmare to put into the iMac. It's just a pain to power.

I hope nobody's brought this up because I skipped a few pages of the thread, but...

I've noticed some things with regards to pricing.

The current 1.86 GHz Yonah in the 17" iMac costs $294.
The new 2 GHz Merom costs $294.
A 2.16 GHz Merom costs $423.
A 2.16 GHz Conroe costs $224.

A 2.16 GHz Conroe is a full $70 cheaper than the 1.86 GHz Yonah in the iMac today and $70 cheaper than the 2 GHz Merom Apple would use if they went with Merom. This would allow either higher profit margins or a price drop (or they could put the extra money into something else).


If there is a power supply problem- I'm sure it won't cost $70 to increase the power supply capacity a little.

If, instead, there is both a heat and power issue- a 2.16 GHz Conroe underclocked to 2 GHz is still $70 cheaper than a 2 GHz Merom and probably outperforms it, and can be advertised as a desktop processor and completes Apple's lineup.


I'm strongly hoping for Conroe in an iMac. I also hope the iMac gets updated at WWDC. I really don't want to wait anylonger to make the purchase, and the back to school deal expires in September two days after MacExpo Paris.


From what's been said, it looks like Conroe doesn't run too hot, it just sucks too much power. However, it still saves a lot of money to use, a little which can be put in to increasing the power supply, and the rest is pure profit for Apple. It also provides a huge leap in performance.

Apple can bump the iMac from 1.86/2 GHz to 2.16/2.4 GHz. The 2.4 GHz Conroe costs $107 less than the 2 GHz Yonah in the current 20" iMac, which could even spell a price drop, additional features, or just a huge Apple profit margin.
 
GFLPraxis said:
I hope nobody's brought this up because I skipped a few pages of the thread, but...

I've noticed some things with regards to pricing.

The current 1.86 GHz Yonah in the 17" iMac costs $294.
The new 2 GHz Merom costs $294.
A 2.16 GHz Merom costs $423.
A 2.16 GHz Conroe costs $224.

A 2.16 GHz Conroe is a full $70 cheaper than the 1.86 GHz Yonah in the iMac today and $70 cheaper than the 2 GHz Merom Apple would use if they went with Merom. This would allow either higher profit margins or a price drop (or they could put the extra money into something else).


If there is a power supply problem- I'm sure it won't cost $70 to increase the power supply capacity a little.

If, instead, there is both a heat and power issue- a 2.16 GHz Conroe underclocked to 2 GHz is still $70 cheaper than a 2 GHz Merom and probably outperforms it, and can be advertised as a desktop processor and completes Apple's lineup.


I'm strongly hoping for Conroe in an iMac. I also hope the iMac gets updated at WWDC. I really don't want to wait anylonger to make the purchase, and the back to school deal expires in September two days after MacExpo Paris.


From what's been said, it looks like Conroe doesn't run too hot, it just sucks too much power. However, it still saves a lot of money to use, a little which can be put in to increasing the power supply, and the rest is pure profit for Apple. It also provides a huge leap in performance.

Apple can bump the iMac from 1.86/2 GHz to 2.16/2.4 GHz. The 2.4 GHz Conroe costs $107 less than the 2 GHz Yonah in the current 20" iMac, which could even spell a price drop, additional features, or just a huge Apple profit margin.
You're the first one to bring this up. Conroe is well worth the money for its processing power. Getting a higher output power supply for the iMac shouldn't be to hard. So, I really do hope Apple somehow puts a Conroe in the iMac. :D

Oh and no underclocking please. :p
 
gnasher729 said:
...Since Intel announced that their four core chips would be available in the last quarter 2006, there is another possibility now for Mac Pros: Just Conroe chips for the "low end" replacing the dual core G5s, and the quad core G5 remains the last PowerPC until it is replaced in the last quarter with a much cheaper quad core system based on Intel's Kentsfield quad core desktop chip. Depends on how urgent it is for Apple to finish the transition.

Better be careful. I posted a similar idea in another thread and got flamed by a couple antagonistic people who have limited vision and are knashing for Woodcrest. I'm in agreement with you. I think having Conroes in the middle and lowend to replace the currently shipping Powermacs is feasible for Apple. Keep the G5 Quad until Kentsfield and maybe introduce a 3.0 Quad Woodcrest on the high-end workstation model to start a new professional line?

I can see why folks are clamoring for Woodcrest, but to me it seems a bit weird for Apple to adopt a chipset for 6 months or less. With Kentsfield shipping at the end of the year, why bother with Woodcrest now? If they would have begun selling last month when they first came out it would have made more sense. Now I'm thinking Apple is going to hold off simply because they haven't announced anything. Woodcrest has been out for around a month now, if Apple is/was going to use them, what's the hold up? I think they have been waiting for Conroe, not WWDC.
 
THX1139 said:
Better be careful. I posted a similar idea in another thread and got flamed by a couple antagonistic people who have limited vision and are knashing for Woodcrest. I'm in agreement with you. I think having Conroes in the middle and lowend to replace the currently shipping Powermacs is feasible for Apple. Keep the G5 Quad until Kentsfield and maybe introduce a 3.0 Quad Woodcrest on the high-end workstation model to start a new professional line?

I can see why folks are clamoring for Woodcrest, but to me it seems a bit weird for Apple to adopt a chipset for 6 months or less. With Kentsfield shipping at the end of the year, why bother with Woodcrest now? If they would have begun selling last month when they first came out it would have made more sense. Now I'm thinking Apple is going to hold off simply because they haven't announced anything. Woodcrest has been out for around a month now, if Apple is/was going to use them, what's the hold up? I think they have been waiting for Conroe, not WWDC.

You don't think Apple would get raked over the coals if they released towers that were slower than the last generation? Conroe is fast, but no way it beats a quad G5. And I don't think a promise of a quad machine later on helps public relations any.

Also, doesn't the kentsfield have the same limitation as conroe? That you can only use it in single processor configs? A woodcrest chipset would have a longer life since you'd use the same one for multiple cloverton configs.

Next gen, conroe gets you 2 cores, woodcrest gives you 2 chips for 4 cores.

Gen after that, kentsfield gets you 4 cores, cloverton gets you 2 chips for 8 cores. There's room for both chipsets for at least the next two generations, and I wouldn't be surprised if it continues beyond that.
 
milo said:
You don't think Apple would get raked over the coals if they released towers that were slower than the last generation? Conroe is fast, but no way it beats a quad G5. And I don't think a promise of a quad machine later on helps public relations any.

Also, doesn't the kentsfield have the same limitation as conroe? That you can only use it in single processor configs? A woodcrest chipset would have a longer life since you'd use the same one for multiple cloverton configs.

Next gen, conroe gets you 2 cores, woodcrest gives you 2 chips for 4 cores.

Gen after that, kentsfield gets you 4 cores, cloverton gets you 2 chips for 8 cores. There's room for both chipsets for at least the next two generations, and I wouldn't be surprised if it continues beyond that.

I didn't mean to suggest the Conroe as a replacement for the G5 Quad. I was thinking more in line with replacing the duals. At this time, the only thing that would come close to replacing the current Quad would be Woodcrest and that's why I mentioned a possiblity for a 3GHZ Woodie in the lineup. I do see a need for a Woodcrest Quad as a professional work station, now and in the future... I just don't see why they would need to put Woodcrest in the complete line-up. Not sure how Kentsfield would play into the long term plan. However, I see nothing wrong with using Conroe in the middle and bottom (non Quad) machines if those are going to continue. Yea, everyone seems to want Quad across the board, but for some people that would be overkill and overly expensive. If you are a web developer or motion graphics designer working at web resolutions, do you need a Woodcrest Quad? Once the currently shipping duals are gone, are we going to be forced with choosing between Woodcrest in a tower... or an iMac or mini?

Maybe it all boils down to the need for Apple to split the lineup as been suggested in other threads. They should have a couple medium towers with Conroe (call them Macs) and then 2 or 3 high-end workstations that are priced accordingly for the Macpro line. I would love to get a Quad Woodcrest, but for the most part, it would be over-kill. Most professional work I do can be done on a single chip / dual core.

It's going to be interesting to see what direction Apple reveals next month. I still think it's strange that Apple hasn't announced Woodcrest if they are indeed going with that solution. Why wait for WWDC if the chips are ready?
 
THX1139 said:
I didn't mean to suggest the Conroe as a replacement for the G5 Quad. I was thinking more in line with replacing the duals.

Maybe I misunderstood your post, I thought you meant releasing conroe machines and not shipping quads until months later. If that were the case, people would inevitably compare the new towers to the G5 quads, regardless if they were intended to replace those models.

I think the reason they haven't announced woodcrest towers is because they want to wait for WWDC, and because the line will be split between woodcrest and conroe. It wouldn't make sense to announce half the tower lineup, people would assume that was it and react accordingly.
 
milo said:
Maybe I misunderstood your post, I thought you meant releasing conroe machines and not shipping quads until months later. If that were the case, people would inevitably compare the new towers to the G5 quads, regardless if they were intended to replace those models.

I think the reason they haven't announced woodcrest towers is because they want to wait for WWDC, and because the line will be split between woodcrest and conroe. It wouldn't make sense to announce half the tower lineup, people would assume that was it and react accordingly.

Exactly! I think the orginal argument was someone saying that there was no way a Conroe was going into a tower or workstation. That the line-up was to be all Woodcrest because that was the only chip that would be feasible to use in a workstation. My counter point to that is - if it was all Woodcrest, they would be out by now. Waiting for WWDC says to me that Conroe is going in there somewhere. My guess is that the Conroe will be used to replace the currently shipping duals at a speed increase. Others argue that the proline will be all Quad... or at least all Woodcrest even if they use only one Woodcrest chip in the lower-end. My best guess is they split the line-up and use both processors. Woodcrest on top in Quad, Conroe on the bottom and middle. They will continue to offer G5 until MWSF to accommodate legacy users.

It's going to be interesting to see what path Apple chooses. The unknown element is that they "might" have something in secret development that will allow them to produce machines we haven't considered. Doubtful but fun to think about.
 
Milo's Analysis Makes The Most Sense To Me - Woodcrest Top, Conroe Mid and Bottom

milo said:
You don't think Apple would get raked over the coals if they released towers that were slower than the last generation? Conroe is fast, but no way it beats a quad G5. And I don't think a promise of a quad machine later on helps public relations any.

Also, doesn't the kentsfield have the same limitation as conroe? That you can only use it in single processor configs? A woodcrest chipset would have a longer life since you'd use the same one for multiple cloverton configs.

Next gen, conroe gets you 2 cores, woodcrest gives you 2 chips for 4 cores.

Gen after that, kentsfield gets you 4 cores, cloverton gets you 2 chips for 8 cores. There's room for both chipsets for at least the next two generations, and I wouldn't be surprised if it continues beyond that.
milo said:
Maybe I misunderstood your post, I thought you meant releasing conroe machines and not shipping quads until months later. If that were the case, people would inevitably compare the new towers to the G5 quads, regardless if they were intended to replace those models.

I think the reason they haven't announced woodcrest towers is because they want to wait for WWDC, and because the line will be split between woodcrest and conroe. It wouldn't make sense to announce half the tower lineup, people would assume that was it and react accordingly.
I believe this is the correct analysis. I am in full agreement with Milo. Good job M. :)
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.