So, the idea now is that rather than a new MP at WWDC, which, after all, is sure to disappoint, the preferred outcome is for Tim Cook to be fired and for Apple to license OSX to HP? I suppose anything is possible. I wonder if HP would even sign up for such a deal, the last time that HP did a deal with Apple, it was a contributing factor to the HP CEO's firing.
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.
2016 nMP
- Thread starter ManuelGomes
- Start date
-
- Tags
- nmp
- Sort by reaction score
- Status
What I'm saying is that the HP with the same underclocked Radeons and CPU as the MP6,1 would be very close in power draw. (If you don't need the second GPU, and a significant percentage of people don't, it could be lower.)
Comparing Pascal GPUs to the Radeons is absurd - but even then the HP might come out ahead. It might have higher kilowatts during the task - but it could complete the task faster and drop to idle sooner. More kilowatts, but fewer kilowatt-hours (kWh). http://www.energylens.com/articles/kw-and-kwh
And kilowatt-hours determines efficiency, not kilowatts.
You seem to be pretty good at finding things on the interwebs, find some actual kWh figures for performing a reproducible task.
Last edited:
Specs, lots of specs....numbers, numbers & more numbers. At the end of the day its more than just numbers, but the actual performance of doing the actual work that matters. Don't care if they under-clock, overclock or de-clock as long as I continue to get faster results like advertised. I'm still getting 50% faster results over my 2009 Mac Pro.
What if you could get 100% or 200% faster with a "Mac Edition"?
(And I know that the correct answer is to figure out what is "fast enough" for you. For some people, the difference between 100% and 50% wouldn't be worth $100 extra. For others, it would be worth $1000 extra.)
What if you could get 100% or 200% faster with a "Mac Edition"?
(And I know that the correct answer is to figure out what is "fast enough" for you. For some people, the difference between 100% and 50% wouldn't be worth $100 extra. For others, it would be worth $1000 extra.)
If $100 is that big of a difference, they probably would not be buying a Mac in the first place.
Probably true, but if you're locked in to Apple OSX applications it could be significant.
I assume that we're not talking about people reading email and surfing, but people who are generating income from the system and need to look at value vs. cost. Sometimes the more expensive (and faster) system is the better value.
You should look at the dual GPUs in the mac pro as a single, large GPU. When Apple devised the Mac Pro, they budgeted 250 W of power towards the GPU. When the mac pro was released, no single GPU could touch the combined 7 TFLOPS of SP compute and the 1.6 TFLOPS of DP compute of the dual D700s. In fact this combination is still competitive against a Fury X (8.6 TFLOPS/0.5 TFLOPS) or the Titan X (6.1/0.2). They could have gone with a 290X, but then we would be stuck at 5.6 SP/0.7 DP TFLOPS.
Of course you could argue that not all tasks can utilize the dual GPUs but anything properly coded on OS X to use OpenCL or Metal should have no problem with this. The nice thing about compute tasks for the GPU is that they are already embarrassing parallel so throwing more GPUs at the problem is a perfectly fine solution.
I think the move to 2 smaller GPUs is supposed to be forward looking. Big, hot GPUs are going to be harder and harder to make. For example the first GPUs out on 14/16 nm are smaller than on previous nodes. As we get down to 10 and then 7 nm this will be a bigger and bigger issue. The cost per transistor is no longer going down, which is why we see a moderately sized Nvidia GTX 1080 at enthusiast prices and the P100 is on the order of $10k a chip.
It seems like this point has to be brought up over and over again. If your needs exceed a single high end Xeon CPU and dual GPUs such as the D700 then you probably are getting into computer cluster territory anyways. Given how prolific (and relatively affordable) these linux based clusters and compute boxes are these days I would say Apple has taken the right approach.
Do you really consider a 2013 vintage 12-core Xeon "high end" today? "High end" today is a pair of faster E5-2600 v4 22-core Xeons.
The MP6,1 when introduced was a mid-range system. (Dual mid-range GPUs, but one display-only and the other compute-only.) Today, almost three years later, it's mostly an embarrassment.
And as far as Linux clusters go, there are few cases where one would choose multiple boxes over a single, more powerful box. Unless your workflow is embarrassingly parallel, the intra-system communications and synchronization consume a significant amount of the overall performance. I have several multi-node (144 core and larger) Spark clusters. For many of the jobs, limiting the job to 8 to 12 cores is quite a bit faster than sending it to 144 cores. Go figure.
Any links, please?
If you got this info from WCCFTech... http://wccftech.com/wwdc-2016-what-to-expect/
Another Narcisist Apple-Hardware Hater, seems he wasn't here at 2013 WWDC when Apple Launched: Macbook Pros, Macbook Air, the new Mac Pro, Airport Extremme, etc...
For me the red flag is they make no mention of macOS (which has basically been confirmed that OS X will be rebranded) so I'm going to take that article with a grain of salt.
Also, anyone following Apple for the last 5 years knows iPhone cycles have shifted to the fall (I think the last summer iPhone was possibly the 4?).
With that said, I wouldn't be surprised if there was a spec bump in the iPad Pro, but overall iDevices are wild cards and not expected, same with iMacs. I would expect to see MacBook Pro, Macmini, and MacPro news... If not I'm going to be very weary about Apple Hardware, which I'm sure we can all agree have been neglected, not only by supplier components, but even the occasional bump in specs (more ram, better drives, etc. in the past it seems that there was something to throw at to avoid stagnation, which I think is where most of the Hardware line is at the moment).
Yes, WWDC is geared towards software and developers, and Apple pretty much dictates iOS apps are developed on Apple Hardware, especially when selling those numbers with metal and other Apple technologies, the developers are nearly if not more excited then us for some raw power and hardware upgrades.
We are having two debates here. The first is on the choice of form factor for the mac pro and the other is on how long it has since its been updated.
Of course I think the mac pro should have been updated by now. At the very least gotten Haswell-E which would have brought DDR4 and more cores. On the GPU front, updating to Hawaii would have been probably very modest at best, given how hot it is. GPUs like AMD Fury and Nvidia's Maxwell are much more graphics focused than compute focused.
On the form factor, both GPUs can be used simultaneously for compute. You are incorrect that only one can be used for compute. Since its been so long since an update it is overpriced for what it is, but 7 TFOPS of compute performance is not an embarrassment.
This is a good rule of thumb for anything from WCCFTech.
I am hoping Apple has been holding back waiting to make a splash at WWDC. The tech should be in place to make substantial updates to the macbook pro, mac pro and introduce an external retina 5k display.
How many times do we need to repeat that only a Haswell-EP or Broadwell-EP processor are candidates for the Mac Pro update unless there is a major redesign and refactoring?
*-EP, not *-E
At WWDC Apple at least Will launch 2 new Macs: Mac Pro and MacBook Pro 15.6|16 and very likely all-new rMacbooks 14/16, maybe later this year they also launch updates for the Mac mini and iMac and a new Thunderbolt retina Display on an special off-schedule event even before iPhone event or close later.
At least the Mac Pro and Macbook pro are safe bet for updates the next WWDC, also no single leak on the fate of the smaller MacBook Pro 13.3
Reasoning, we have enough leaks about the Mac Pro to be confident its close to an very important upgrade (VR-...), also are enough leaks on the upcoming Macbook Pro (AMD Polaris)to be confident about its launch.
About the Retina Macbook 14/16, apart from some Chinese leakers talking about with mixed accuracy records we have nothing more substantive to be as 100% confident as with the pro's Macs.
I don't believe we will see a new thunderbolt display until the renewed iMac launch, but still a possibility for WWDC.
At least the Mac Pro and Macbook pro are safe bet for updates the next WWDC, also no single leak on the fate of the smaller MacBook Pro 13.3
Reasoning, we have enough leaks about the Mac Pro to be confident its close to an very important upgrade (VR-...), also are enough leaks on the upcoming Macbook Pro (AMD Polaris)to be confident about its launch.
About the Retina Macbook 14/16, apart from some Chinese leakers talking about with mixed accuracy records we have nothing more substantive to be as 100% confident as with the pro's Macs.
I don't believe we will see a new thunderbolt display until the renewed iMac launch, but still a possibility for WWDC.
And on 13 June, come back here for "Much Wailing, Rending Of Garments, And Gnashing Of Teeth" when none of this happens.
Be prepared for the most disappointing MacWorld SF of recent memory.
if a 7_1 is shown, there will be much wailing here.
if it's not shown, there will be much wailing here.
either way, prepare for much wailing here.
It will be a keynote of biblical significance....
Last edited:
There's good macro reasons for a Mac Pro update at WWDC. Apple needs to get a Mac sales spike going, so they're likely to do a bunch of Mac updates. All of the models are long in the tooth, so there's a nice overall base of pent up demand, and the MP puts it over the top. Further, AMD is hungry and will let Apple do another multi-year pre-buy of parts at a big discount like they did last time (which is why Apple never updated the GPU after the nMP launch). This lets AMD announce an immediate revenue boost and market share increase with the launch of the new GPU line, which otherwise would take several quarters to show real results (again, just like last time). WWDC is the last good time for Apple to do an announcement before we get into summer months and then into the iPhone announcements in the fall. Now, Apple is fully capable of doing none of this, but it does make it seem more likely than not. In any case, we'll know for sure in a few weeks.
ill call it broadwell ep when you stop calling wwdc Mac world
Cringe worthy website.......UGGGHHH.
[doublepost=1463290374][/doublepost]
How hungry do you think they are? If they have this high end cpu custom upgrade....something AMD hidden from us....then......dog food!- Status