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.
iMac Pro Appears to Include A10 Fusion Chip for Always-On 'Hey Siri'
- Thread starter MacRumors
- Start date
- Sort by reaction score
I just wouldn't want a workstation where I couldn't replace parts down the road, and I figure most people that need workstations don't want to replace their monitor every time. To each his own though, I get that there's a variety of different use cases out there.
I'm mostly just annoyed because my 2010 MP is on its last legs and I'm having to use my Windows box while I wait to see what Apple FINALLY does next year. No way I'm replacing with something that has poor thermals, can't be upgraded, and would cost me my beautiful 34" 21:9 display.
I feel that there are very few legitimate reasons to cram workstation functionality into an all-in-one other than to look pretty, so it's more than just my expectations. It's one of the most pure examples of form over function in recent memory.
Last edited:
A lot of businesses buy these things by the doZen. They never upgrade the internals because they depreciate the cost and replace after the depreciation period. In the mean time they want every desktop to have identical configuration and don’t want anyone opening them up.
Exactly. Everyone poo pooing the form factor seems to be people who buy one machine for themselves once every four or five years. That’s not the meat of the market for pro machines.
FakeA10FusionChip.kext
[doublepost=1512587817][/doublepost]
The same apps that run on the iPhone A10 with 2/3GB RAM while the iMac *Pro* A10 has 512MB?
[doublepost=1512587817][/doublepost]
The same apps that run on the iPhone A10 with 2/3GB RAM while the iMac *Pro* A10 has 512MB?
Is anyone else thinking that Apple may want its Mac's embedded ARM chips to take over some of the tasks currently handled by the Management Engine inside all Intel x86/-64 CPUs, or is it just me?
Actually, the potential ramifications of the Intel chip design snafu go way beyond just Apple. To really fix the problem, Intel will need to do a major redesign of its chips. Until that happens, correction must be implemented in microcode (BIOS/EFI level firmware and low level kernel software). This will induce performance hits on Intel based computers, anywhere from 3% to high estimates of 30%. So, in the meantime, until Intel can bring new chip design to market, hardware vendors will need to turn to non-Intel chip manufacturers for new computers if they are to roughly maintain performance levels to which their customers are accustomed. If Intel takes a nose dive, this could have a huge influence on the computer hardware market, both in sales and technologically. ARM and AMD remain the most visible alternatives to Intel, as other chip designs such as RISC have largely disappeared over the last 20 years with cheap Intel processors coming to the fore.
True, largely the only RISC architecture available right now. Sort of like RISC is now ARM. I think competition might be engendered with this Intel thing.
https://www.extremetech.com/computi...ttle-arm-and-x86-by-being-totally-open-source
Sorry, should have qualified my "disappearing RISC" statement. I had high hopes for DEC Alpha in the 90s, at which point only they and high performers like Cray were producing 64 bit technology.
@curtvaughan: If there are any vulnerabilities in the ME, then they'd be the one exploited by a ring (-3) rootkit, the one in zero-touch provisioning, SA-00075 'Silent Bob is Silent', PLATINUM, or SA-00086, not chip-wide ones like Meltdown or Spectre. I do, however, agree that horrible messes of nasty lie down those last two gnarly, knotty rabbit holes as well. In any case, Apple wouldn't be the first company to work around all of the former vulnerabilities by disabling the ME. I don't know of anybody else who might have tried to utterly replace the ME by offloading (at least some of) its functionality to a custom external chip, at least not off-hand…
Dell has been quite rapidly supplying BIOS updates to circumvent SA-00086. I updated my XPS-13 (running Linux) just several days ago from this website:
http://www.dell.com/support/home/us/en/19/drivers/driversdetails?driverId=GVNVJ
As to whether vendors (Apple included here) will choose to disable ME, the article you referenced indicates that it would be ineffective against SA-00086, as the vulnerability is apparently loaded so early in the boot process. To quote:
Effectiveness against vulnerabilities[edit]
None of the two methods to disable the ME discovered so far turned out to be an effective countermeasure against the SA-00086 vulnerability.[57] This is because the vulnerability is in an early-loaded ME module that is essential to boot the main CPU.[58]
It's worth noting in the case of my PC's running Linux, specifically with Dell (as it is the only one to have yet issued a BIOS patch addressing SA-00086), even with the BIOS update applied, and with updated kernels also meant to address the vulnerability, none of the three Linux distros on the XPS-13 (Ubuntu, Antergos, Mint) test as being completely secure. Tests can indicate whether the kernel patches have been completed, and tests can be made to detect BIOS vulnerabilities independent of the kernel, but the interaction between the OS and microcode in the firmware is as yet not clearly verifiable - leaving the machine to be declared "vulnerable".
The opera ain't over 'til the fat lady sings, and she has yet to make her appearance .... I've diverted somewhat from Apple specifically, primarily because Apple is less forthcoming about details of their patches and tests. Since they control both the firmware and OS kernel on their machines, I would assume they might have an easier go of fixing things in both EFI/BIOS as well as with the BSD kernel they deploy. As they also use Intel chips, I'm guessing they are running into similar firmware/kernel issues.
Guess I should have done more than skim the material I referenced before I did, huh? (Oops.) You make good points, and it sounds like you're more knowledgable about the situation than I am. I've mostly just been following Purism since they've been featured in Phoronix articles over at least the past six months, and I only found that other news site relatively recently. Sounds like things are more serious than I though, then…hmmmm. Still, if Apple intends to use its T-series chips to manage things like booting and power states, at least some of which are traditionally managed by the ME, could they use it to do more? What other ME functions could the T-series chips take over? Or are you saying that I'm grasping at straws here?
@RandomDSdevel: I'm rather thinking it is too early to tell exactly what direction computer vendors are going to take. It will possibly be a two-pronged approach, dependent upon how quickly chipset manufacturers, esp. Intel, are able to fix SA-00086 at the hardware level. Whether this will involve removal of ME in favor of some other chip level paradigm, or the alteration of ME to make it more secure, or just the removal of MINIX altogether at the firmware/hardware abstraction level isn't clear - and that's just from Intel's corner. That Intel's CEO sold nearly 200 thousand shares, down to the minimum allowed by law (250000 until a specified "grace period" is complete), some weeks before the public announcement of SA-00086 leads one to believe that there's no consensus for direction from Intel's top management. In any case, vendors (including Apple) have some immediate decisions to make:
1) move away from Intel in favor of AMD, ARM, or perhaps a yet-to-be designed and/or released CPU offering from other parties
2) how to best mitigate currently vulnerable Intel hardware issues with OS and firmware without crippling performance
3) how to best allocate software engineering R&D for future OS designs, depending on decisions made with 1) and 2)
To make a long story short, I don't think you are grasping at straws - it's just too early to answer your question, and it may just be a hypothetical, depending on how 1,2, and 3 take shape in the coming year. I see this as rather a disaster on Intel's part, particularly as they have been less than forthcoming until recently about the existence and function of ME in their chip designs for OVER A DECADE. That makes Apple's battery/performance snafu with iOS seem trivial.
As food for thought, check out Intel's rather depressing revelation, some of which goes back to November - with edits in December. Krzanich, Intel's CEO, began selling his stock on Nov. 29.
https://security-center.intel.com/advisory.aspx?intelid=INTEL-SA-00086&languageid=en-fr
1) move away from Intel in favor of AMD, ARM, or perhaps a yet-to-be designed and/or released CPU offering from other parties
2) how to best mitigate currently vulnerable Intel hardware issues with OS and firmware without crippling performance
3) how to best allocate software engineering R&D for future OS designs, depending on decisions made with 1) and 2)
To make a long story short, I don't think you are grasping at straws - it's just too early to answer your question, and it may just be a hypothetical, depending on how 1,2, and 3 take shape in the coming year. I see this as rather a disaster on Intel's part, particularly as they have been less than forthcoming until recently about the existence and function of ME in their chip designs for OVER A DECADE. That makes Apple's battery/performance snafu with iOS seem trivial.
As food for thought, check out Intel's rather depressing revelation, some of which goes back to November - with edits in December. Krzanich, Intel's CEO, began selling his stock on Nov. 29.
https://security-center.intel.com/advisory.aspx?intelid=INTEL-SA-00086&languageid=en-fr
Last edited:
Sorry, I sort of rambled on with TMI. The fallout from SA-00086 will have some interesting ramifications. If nothing else, it will put innovation and competition back into CPU chip design. That's been long needed, IMO.