Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Always bet on economics. It's how things work nowadays. Factor in the design leverage/properties of this thing and it's still a winner in today's market place handily. Why upgrade now? Because some fan boys are angry? Not likely! Just my .02. :cool:

I totally agree that they gauge economics, not just new parts coming down the pipe, but we don't really know the actual sales of the device. Apple's not going to break apart its sales figures just to give us an insight, and given that a "successful" Mac Pro would be a small portion of even the desktop revenue Apple brings in, I don't think you can reliably divine sales from a bump here or there.

Opinions on forums like these are pretty much always going to be a vocal minority that doesn't necessarily track with the "common person", even if in this case that "common" person is more likely an AV professional with cash to spend. But as for extrapolating from what forums and the online crowd and even what you might see at a trade show funnels into actual sales? It's a waste of time.
 
I totally agree that they gauge economics, not just new parts coming down the pipe, but we don't really know the actual sales of the device. Apple's not going to break apart its sales figures just to give us an insight, and given that a "successful" Mac Pro would be a small portion of even the desktop revenue Apple brings in, I don't think you can reliably divine sales from a bump here or there.

Opinions on forums like these are pretty much always going to be a vocal minority that doesn't necessarily track with the "common person", even if in this case that "common" person is more likely an AV professional with cash to spend. But as for extrapolating from what forums and the online crowd and even what you might see at a trade show funnels into actual sales? It's a waste of time.

For sure, this is probably their least popular selling product outside of accessories, so realistically it is going to get a smaller amount of attention.
 
For sure, this is probably their least popular selling product outside of accessories, so realistically it is going to get a smaller amount of attention.

Obviously they saw enough of a market to redesign the power PC concept. That should speak volumes. If having the greatest processor is the goal then why change anything? Fact of the matter is they saw a market that was being undeserved with lots of potential. That's why they spent so much in redesign.
 
I give it about 50/50 chance of announcement either in person or via web at the Oct 16th event.

I've heard a few people say chances are low because there have been no rumors, but it's worth noting that previously there were not usually leaks or hard rumors about announcements. Even big announcements, like the WWDC 2013 announcement were not leaked beforehand.
 
I give it about 50/50 chance of announcement either in person or via web at the Oct 16th event.

I've heard a few people say chances are low because there have been no rumors, but it's worth noting that previously there were not usually leaks or hard rumors about announcements. Even big announcements, like the WWDC 2013 announcement were not leaked beforehand.

Exactly. I mean we haven't seen any leaks about the Mac Mini either, regarding supply chains ramping up or new parts leaking out or anything. And that's likely to be updated in 8 days. Same goes for the retina iMac.

I give it a 50/50 shot as well. The parts are available and it can be done rather easily, so I don't see why they couldn't do it this soon. If you look at all of the Intel lines that they didn't skip regarding the Mac Pro, it seems that with the exception of Sandy Bridge, the Mac Pro was updated within 3-4 months of the new chips dropping into retail. I do attribute the long gap the last time around to figuring out what to do with the machine, deciding on a redesign, and then moving forward with it.

I'm breaking this mythical refresh down into two time periods: From now until the end of the year, and mid-January into April. Don't think they'll wait that long to do it, but we'll see.
 
I'm breaking this mythical refresh down into two time periods: From now until the end of the year, and mid-January into April. Don't think they'll wait that long to do it, but we'll see.

Yeah, if it doesn't get rev'd at the event, there is a slim chance we could see a rev in November. Past that, January would be the next time.
 
Unless Apple does a big about-face - the MP6,2 will have Haswell-EP.

Haswell-E chips are the Core i7-5xxx series, which do not support ECC memory.

I was expecting the Xeon versions in the nMP. I just want some kind of haswell-e support in the kernel because it'll help my hackintosh build.
 
They better refresh it. I want native Haswell-e support.


Why? What benefits do you see? I'm not sure there's much to look forward to... AnandTech recently did a number of benchmarks comparing Haswell-E with equivalents from Ivy and Sandy Bridge, and in a couple cases it was a few percent better than Ivy, and in a couple cases it was a few percent worse (a regression to Sandy Bridge). Even though Apple will use the EP variants, the benchmarks are indicative of a mixed bag when it comes to any performance improvements.
 
I was expecting the Xeon versions in the nMP. I just want some kind of haswell-e support in the kernel because it'll help my hackintosh build.


I see.

I guess for everyone else looking forward to Haswell... What about it are you looking forward to? Have you looked at benchmarks for your workflow and are you seeing anything worth looking forward to?
 
Why? What benefits do you see? I'm not sure there's much to look forward to... AnandTech recently did a number of benchmarks comparing Haswell-E with equivalents from Ivy and Sandy Bridge, and in a couple cases it was a few percent better than Ivy, and in a couple cases it was a few percent worse (a regression to Sandy Bridge). Even though Apple will use the EP variants, the benchmarks are indicative of a mixed bag when it comes to any performance improvements.

grantley-unixbench-multithreaded-whetstone_w_600.png


http://www.tomshardware.com/reviews/intel-xeon-e5-2600-v3-haswell-ep,3932-7.html

When you look at the improvements based on the core counts increasing, multicore benchmarks really start to shine on the new hardware. If you've got multicore workflows, Haswell-EP is a very nice upgrade. If you're on single core workflows, it's pretty meh.

Not a huge upgrade to everyone, but a significant enough of an upgrade to some (especially users of Apple apps) that Apple has reason to upgrade. That, and new GPUs.
 
Image



http://www.tomshardware.com/reviews/intel-xeon-e5-2600-v3-haswell-ep,3932-7.html



When you look at the improvements based on the core counts increasing, multicore benchmarks really start to shine on the new hardware. If you've got multicore workflows, Haswell-EP is a very nice upgrade. If you're on single core workflows, it's pretty meh.



Not a huge upgrade to everyone, but a significant enough of an upgrade to some (especially users of Apple apps) that Apple has reason to upgrade. That, and new GPUs.


That's indeed impressive improvement and the first benchmark I've seen that shows anything like that. The benchmarks AnandTech ran were not so flattering.
 
That's indeed impressive improvement and the first benchmark I've seen that shows anything like that. The benchmarks AnandTech ran were not so flattering.

If I were you, I would read the entire Tom's Hardware article on Haswell-EP. AnandTech were focusing on Haswell-E Core i7 parts, which will not be used in the Mac Pro. Specifically pay attention to the sections where they show and note where Haswell-EP is a pretty nice leap up from IvyBridge-EP in terms of the total package.
 
That's indeed impressive improvement and the first benchmark I've seen that shows anything like that. The benchmarks AnandTech ran were not so flattering.

Core for core it's pretty meh, but with the core counts moving up across the board, it would make it a pretty great upgrade, even if the cores were clock for clock the same speed as Ivy.

I don't know if Apple will use the higher core counts, but it's hard to see them not doing so, even if they don't go all the way up to 16 cores.
 
That's indeed impressive improvement and the first benchmark I've seen that shows anything like that. The benchmarks AnandTech ran were not so flattering.
I'm not sure why these multi-threaded benchmark results are at all surprising since 2690v1 is a 8-core part, 2690v2 is a 10-core part, and 2690v3 is a 12-core part.

Besides, aren't the E5-2xxx processors for dual CPU workstations? Wouldn't Apple use the E5-1xxx v3 variants like they do in the current nMP's? If so, moving from 1xxx v2 to 1xxx v3 doesn't seem to net you additional cores (and additional multi-threaded performance) like the move from 2xxx v2 to 2xxx v3 does.

I think the current significantly higher cost of DDR4 memory will be what pushes any nMP updates into next year. Or maybe not...I just checked prices and although it's 2x more expensive, it's not too bad as a percentage of total system cost.
 
Last edited:
nMP refresh

I'm not sure why these multi-threaded benchmark results are at all surprising since 2690v1 is a 8-core part, 2690v2 is a 10-core part, and 2690v3 is a 12-core part.



Besides, are the E5-2xxx processors for dual CPU workstations?


Hmm... Too much wine this evening :) I had overlooked that... You're correct so it's actually consistent with other benchmarks I've seen. Disappointing. It's actually a bit damning that the v3 10-core shows a 20% improvement over the v2 8-core in that benchmark with 25% more cores.

----------

Core for core it's pretty meh, but with the core counts moving up across the board, it would make it a pretty great upgrade, even if the cores were clock for clock the same speed as Ivy.



I don't know if Apple will use the higher core counts, but it's hard to see them not doing so, even if they don't go all the way up to 16 cores.


The problem is that pricing hasn't really changed with v3. I think the only part that is s bit cheaper is a new 8-core that's clocked fairly low so it's not much to get excited about. All the other v3 CPUs comparible to v2 parts have the same pricing which means Apple can't offer more cores for the same price thanks to Intel.
 
The problem is that pricing hasn't really changed with v3. I think the only part that is s bit cheaper is a new 8-core that's clocked fairly low so it's not much to get excited about. All the other v3 CPUs comparible to v2 parts have the same pricing which means Apple can't offer more cores for the same price thanks to Intel.

The 1660v3 offers a much cheaper entry point to the 8 core than v2 did. That should help Apple offer a better deal on the 8-core. The 4-core and 6-core should be pretty unaffected and be very small upgrades relative to v2. It will be interesting to see where Apple goes with the 12, 14 or 16 options on the top end too.
 
The 1660v3 offers a much cheaper entry point to the 8 core than v2 did. That should help Apple offer a better deal on the 8-core. The 4-core and 6-core should be pretty unaffected and be very small upgrades relative to v2. It will be interesting to see where Apple goes with the 12, 14 or 16 options on the top end too.


Right. They could offer an 8-core for about $1K less than current or they could offer a faster 8-core for the same money. They could offer the 18 core on the top end, but the CPU alone costs somewhere around $4500. :eek:
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.