So why did Apple make the 5k iMac that I use daily at work?
As a display it's spoilt me and my home 1440p ultrawide G-Sync monitor no longer satisfy's.
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.
16-Inch MacBook Pro Will Reportedly Use Intel's 9th-Gen Processors With Up to 8 Cores
- Thread starter MacRumors
- Start date
- Sort by reaction score
So why did Apple make the 5k iMac that I use daily at work?
As a display it's spoilt me and my home 1440p ultrawide G-Sync monitor no longer satisfy's.
Intel's eDRAM has always been a separate die in the CPU package - often/always made on an older fab line at a larger geometry. It has also had few customers so it makes sense for Intel to postpone eDRAM variants when releasing a new generation.
physical function keys and a touch bar is redundant, and excessive. it's okay to not want a touchbar, there's nothing wrong with having model options with and without.
I suspect that the release will happen at some sort of "pro event". Probably when sales of the Mac Pro open. We might see iPad Pros as well.
The one other possible Pro-labeled machine is the 2 year old iMac Pro. It could get a new screen - maybe 32", some version of the technology in the XDR display? Are there new "medium socket" Xeons it could use?
The one other possible Pro-labeled machine is the 2 year old iMac Pro. It could get a new screen - maybe 32", some version of the technology in the XDR display? Are there new "medium socket" Xeons it could use?
Question for you all.
If I bought the 2019 Macbook Pro 15" 8 Core already in June... would it be smartest to try to sell it before the 16" comes? Maybe I can try to return? It's almost been 60 days though.
Thinking about it because...
#1 I'm a video editor and always looking to get best performance. (Although I dunno if it'll be any faster)
#2 I'm always trying to maintain value of the things I own so I can resell them at an awesome price. Worried that when 16" launches the 15" will be less popular and price will drop. Although, the con is that it may cost me a lil to upgrade like this.
Another question:
In the meanwhile, I've waited on getting AppleCare+ for my 2019 Macbook Pro (almost about to pass my 60 days) as I wanna see what the 16" brings and don't want to buy Applecare if I'm just gonna sell it soon. Should I just suck it up and get the applecare or wait?
If the 16" won't bring performance updates I might as well stick to my 15". Thoughts?
If I bought the 2019 Macbook Pro 15" 8 Core already in June... would it be smartest to try to sell it before the 16" comes? Maybe I can try to return? It's almost been 60 days though.
Thinking about it because...
#1 I'm a video editor and always looking to get best performance. (Although I dunno if it'll be any faster)
#2 I'm always trying to maintain value of the things I own so I can resell them at an awesome price. Worried that when 16" launches the 15" will be less popular and price will drop. Although, the con is that it may cost me a lil to upgrade like this.
Another question:
In the meanwhile, I've waited on getting AppleCare+ for my 2019 Macbook Pro (almost about to pass my 60 days) as I wanna see what the 16" brings and don't want to buy Applecare if I'm just gonna sell it soon. Should I just suck it up and get the applecare or wait?
If the 16" won't bring performance updates I might as well stick to my 15". Thoughts?
Since you've had it 60 days, you can't return it for a refund. Since Apple has already gone on record saying that the keyboard itself is warranted for 4 years from time of I purchase, I would just purchase the AppleCare+ warranty and use the machine for the next 4 years.
Is this a serious question? The 5k iMac is 27", the Retina MacBook Pro is 15". They both have a similar PPI. The MacBook Pro does not need a high PPI, it looks just like the 5k iMac does. It would be the 5k iMac becoming a 10k iMac - it'd just take a hell of a lot more processing power, actual power and heat to drive something that would look identical to your eyes as the 5k iMac.
Unfortunately, AMD isn't an option because they don't have a complete and fully competitive product portfolio for Apple to be able to make a wholesale switch. Not a ding against AMD, it is what it is. They are focusing on those key areas where they see the maximum growth potential and taking it right to Intel's door. If this strategy works, then they can begin filling out other product lines. However, they are still taking too long on the Navi rollout and this doesn't help matters, because NVIDIA isn't just sitting around picking their nose. I admire AMD's chutzpah, as they are taking on Intel AND NVIDIA, either one by themselves a formidable opponent, both together isn't a pleasant place to exist.
Also, at this juncture, if Apple is going to embrace A-Series CPUs in the Mac, they have taken AMD out of the CPU equation entirely and will simply continue down the road with Intel to the end. Cue Boyz II Men.
No . Some hardware fixes are rolling out before Ice Lake . There is a table here .
https://www.anandtech.com/show/13400/intel-9th-gen-core-i9-9900k-i7-9700k-i5-9600k-review/2
The 9th general is Coffee Lake Refresh . Intel has muddled where the fixes landed as the Whiskey Lake got tagged 8th gen .
Additionally , Ice Lake won’t fix everything either . Willow Cove arch in Tiger Lake is where Intel will lay down something more than band-aids and duck tape fixes .
[doublepost=1565559821][/doublepost]
Nothing anytime soon is going fixes for all the flaws . Some software/firmware is need regardless .
Genuine question. Why do you think other people should care about your imperceptiveness on a topic?
I understand I am risk life and limb by posting this, but since this entire family of CPU's support 128GB of RAM, can we PLEASE see an option for at least 64GB in a new MacBook Pro? I know, I know, every time there is a post like this half the world attacks the original poster saying "NO ONE NEEDS THAT MUCH RAM!" To that I simply say, "And you do not work in I.T. or you would never say that!" Personally I use a MacBook Pro all day long with 3-4 virtual machines running at the same time, which is necessary for the type of work I do. Other tasks such as 4K video editing also require or benefit from additional RAM. Again, I said AN OPTION, not the base model, so if you feel you do not need it, don't buy it, but recognize that some of us have such a need. I would love to be able to ditch my giant Dell laptop, but it has 64GB of RAM (and 128GB is now an option in the Precision line with these processors) but the 32GB in this MacBook Pro pukes when I attempt to launch everything I need.
I have no personal need for 64GB but I think it's perfectly reasonable for Apple to include that option, and I think it actually behooves them to do so because it would help shape the perception that the new MBP is a "Pro" machine.
128GB is very unlikely because that currently requires 4 DIMMs, and the recent MBPs only have room for two, and I don't see that changing.
I don’t think it is unreasonable for Apple to offer a BTO option for up to 64GB of DRAM. Perhaps they will even go off the hook and offer up to 128GB of DRAM, but I think that is highly unlikely given that even higher DRAM densities still require a fair number of chips to hit that 128GB of DRAM in a constrained devices size. Unless Apple makes the MBP a fair bit thicker to accommodate DRAM on both sides of the logic board. I have my doubts that they will go back to SOP-DIM slots, but anything is possible. I doubly doubt they will put 4 slots in, but I guess we have to see.
There is nothing wrong with wanting 64GB of DRAM, it really is highly dependent on what you do with your computer. Even for 4K video editing, 32GB is sufficient is 90% of the cases. The extra DRAM may help offload other tasks that you have running as well, but going to 64GB is not a cure all as some may believe with video editing or even photography.
Hopefully, both DRAM and GPU will get some love this time around. Four generations of essentially the same Radeon Pro GPU is disheartening even though we also have Vega 16 and Vega 20.
My suspicion is 64 GB will be a BTO option (still soldered) on the 16", which may also have a 32 GB base configuration, at least in higher-end models. Most similar thin and light workstations support 64 GB, but not 128 GB, which is left for the heavy models that support multiple drives and sometimes higher-end GPUs.
Honestly the monitor on a MacBook Pro is almost as thick as an iPad and that houses a battery, cpu and other such goodness. Why don’t they move the battery to behind the screen? That would give more room in the main unit for heat flow etc.
You can dm me for more ideas Apple.
You can dm me for more ideas Apple.
Stress on the hinge - even the convertible laptops that have to have some battery behind the screen (to enable tablet mode) keep as much of it as possible in the base.
Underutilized, yes, half baked, NO! I use mine every single day and I absolutely love it.
They didn't switch of the order of Comet Lake H and U releases. So I was off there.
They tossed them into the 10th. But to came up with a labeling system that is perhaps inspired by a drunken sailor in a strip club .. they have added an extra level of to complexity the product level decoder ring.
https://www.anandtech.com/show/1478...met-lakey-10th-gen-core-for-low-power-laptops
The newer 10nm implementation has 4 digits and ends with a Graphic level code . The tweaked 14nm with new memory controller has 5 digits and ends in U/Y and no graphics digit code. (more digits for the older product foundation. I suspect the more digits is coupled to the "more cores" a.k.a. bigger boobs. ) The fig leaf merging these into the same generation is that the memory type coverage is the same ( but not the speeds of memory ).
It is probably a really lucky thing for Intel that AMD is not trying to compete this year in this segment. They are shooting at different very narrow segments inside this segment with highly overlapping products.
If Intel slapped a new Memory controller on Comet Lake U then it seems likely they'd do the same for Comet Lake H .... unless following the Y track and just recycling the die with some small die tweaks and a new PCH coupling. It may just be consistent with the haphazardness of the gen assignments to put Comet Lake H into the 9th if it doesn't have the memory controller upgrade ( since perhaps that is the 'rational' factor covering the 10th designation. and just irrationally stuff Y into 10th because not looking for clarity. ). However, the H needs LDDR4 as much as the U does. It would be odd to confine all that backporting work to just one relatively narrow product. But its an odd situation that Intel is in.
The Tweakers leaked roadmaps do make mention of LDDR4 for U and not H. So maybe that is how H lands into the 9th. But for H to miss going into the 10th at this point it has to be missing something that the 10th are getting (or were suppose to get). But if H class is coming forward for an early MacBook 16" then 10nm is probably doing better than the original plan. ( These Comet Lake U are gap filling but the gaps are incrementally smaller than projected. )
Apple dropping the MacBook is suggestive of why Intel also didn't throw tons of memory augment work at the Y class.
( and perhaps why Apple upgraded the MacBook Air retina and still sitting on 'old' Y's to perhaps soak up the rest of the supply they contracted for and much better "off to college" timing. ). Also throw on top Intel putting lots of effort into Lakefield https://www.anandtech.com/show/14773/hot-chips-31-live-blogs-intel-lakefield-and-foveros ( if that is coming Q4 19 for some non Apple player(s) who is going to buy big then I could see them killing of a backport to Y for memory updates ). The Y won't do as well fending off Qualcomm (and ARM) on the lower end that Lakefield probably will.
P.S. The 1.1GHz base clock on the 6C Comet Lake U doesn't particularly bode well for the 10 core Comet Lake H. The incrementally higher core count is going to come at a significant loss in base clock. Apple could be skipping the 10C model if the trade-off and TDP (of let it off the chain a bit) don't match up with their new case and expectations.
Last edited:
They didn't switch of the order of Comet Lake H and U releases. So I was off there.
They tossed them into the 10th. But to came up with a labeling system that is perhaps inspired by a drunken sailor in a strip club .. they have added an extra level of to complexity the product level decoder ring.
https://www.anandtech.com/show/1478...met-lakey-10th-gen-core-for-low-power-laptops
The newer 10nm implementation has 4 digits and ends with a Graphic level code . The tweaked 14nm with new memory controller has 5 digits and ends in U/Y and no graphics digit code. (more digits for the older product foundation. I suspect the more digits is coupled to the "more cores" a.k.a. bigger boobs. ) The fig leaf merging these into the same generation is that the memory type coverage is the same ( but not the speeds of memory ).
It is probably a really lucky thing for Intel that AMD is not trying to compete this year in this segment. They are shooting at different very narrow segments inside this segment with highly overlapping products.
If Intel slapped a new Memory controller on Comet Lake U then it seems likely they'd do the same for Comet Lake H .... unless following the Y track and just recycling the die with some small die tweaks and a new PCH coupling. It may just be consistent with the haphazardness of the gen assignments to put Comet Lake H into the 9th if it doesn't have the memory controller upgrade ( since perhaps that is the 'rational' factor covering the 10th designation. and just irrationally stuff Y into 10th because not looking for clarity. ). However, the H needs LDDR4 as much as the U does. It would be odd to confine all that backporting work to just one relatively narrow product. But its an odd situation that Intel is in.
The Tweakers leaked roadmaps do make mention of LDDR4 for U and not H. So maybe that is how H lands into the 9th. But for H to miss going into the 10th at this point it has to be missing something that the 10th are getting (or were suppose to get). But if H class is coming forward for an early MacBook 16" then 10nm is probably doing better than the original plan. ( These Comet Lake U are gap filling but the gaps are incrementally smaller than projected. )
Apple dropping the MacBook is suggestive of why Intel also didn't throw tons of memory augment work at the Y class.
( and perhaps why Apple upgraded the MacBook Air retina and still sitting on 'old' Y's to perhaps soak up the rest of the supply they contracted for and much better "off to college" timing. ). Also throw on top Intel putting lots of effort into Lakefield https://www.anandtech.com/show/14773/hot-chips-31-live-blogs-intel-lakefield-and-foveros ( if that is coming Q4 19 for some non Apple player(s) who is going to buy big then I could see them killing of a backport to Y for memory updates ). The Y won't do as well fending off Qualcomm (and ARM) on the lower end that Lakefield probably will.
P.S. The 1.1GHz base clock on the 6C Comet Lake U doesn't particularly bode well for the 10 core Comet Lake H. The incrementally higher core count is going to come at a significant loss in base clock. Apple could be skipping the 10C model if the trade-off and TDP (of let it off the chain a bit) don't match up with their new case and expectations.
The Word Of The Day is: Hot Mess
EDIT (3:57PM EDT): Even worse...Intel gave the 15w U-Series Comet Lake LPDDR4 support, but omitted it from the Y-Series Comet Lake, where that extra reduction in power consumption and increased power state support would arguably be of much more use...again, forcing Apple (and other vendors) to choose a LPDDR3 or DDR4...it makes absolutely zero sense to have a hallmark 10th Gen feature (LPDDR4/x support) on everything but the CPU series that could use it the most. At this point, Ax-Series CPUs from Apple cannot get here fast enough for this particular segment of the market. Intel has lost the plot completely.
Love the following quote of yours, "They tossed them into the 10th. But to came up with a labeling system that is perhaps inspired by a drunken sailor in a strip club .. they have added an extra level of to complexity the product level decoder ring."
I feel as though Intel is simply throwing anything and everything at the wall, seeing what sticks and then moving on to the next target. It's just one constant recycle of 14nm after another, while 10nm circles the drain as Intel grasps to make back some sort of cash from the R&D and fab outlay on this turkey. Maybe SemiAccurate has an ax to grind, but, jeez, Intel is making it so easy. They just won't admit 10nm is a spectacular failure.
Last edited:
Depending up the quantity of Lakefiled they can produce later this year it isn't particularly even worse. If you didn't before you should look through the slide deck for the Lakefiled presentation. Lakefield is way smaller. If the Y-series is aimed at utlra, ultra light systems then smaller will turn out better. Running on 4 atom cores in "not much going on" mode will save more battery than LDDR3 vs LDDR4 ( although lakefiedl does have LDDR4 also). The point being the limbo down to ARM phone processor level is being handed to another line up with Celleluar modems in the reference desin ( which is another area Intel needs to finish figuring out because Apple is going somewhere else. )
For example
the one at the bottom is a Y-class reference board. The clamshell laptop Lakefield is second from top.
The Y class usage in the Mac line up at this pint is the MacBook Air Retina ( which isn't quite the same battery and size constraints as the last Macbook ).
It is not clear where the chicken and egg thing is here because Apple dropped the MacBook. Could be because they didn't have anything on Y decent enough to them to go to. Or perhaps because Apple didn't want Y , even with LDDR4 because had committed to fork the Mac line up with A-series at the low end. ( one port wonder future iPadPro processor meets one port wonder MacBook ). I highly doubt just LDDR4 would make Apple happy there versus the alternative they already committed to paying the development for. Intel is in deeper 'mud' at that point than just memory controller. Nor would a 'magical' 10nm necessarily get them out of that either.
It don't think LDDR4 is going to help with the Qualcomm 8cx systems coming in Q4 either. These are probably somewhat cherry picked benchmarks by Qualcomm but it does show Intel has an issue that isn't limited to Apple's A-series. ( Lenovo did demos with a future laptop around the same time at Computex ).
https://www.windowscentral.com/snapdragon-8cx-benchmarks
The Comet Lake Y probably works in more "up from the edge" systems ( and probably looking to enable cheaper logic board updates ) and the Comet Lake U Series probably match up better in systems where there is a discrete mobile GPU. Integrated graphics the Ice Lake U has the edge.
I don't think LakeField is going to be a computational big win but it will probably keep them in he game on those battery life metrics. Certainly is going to enable taking up less space ( so can insert more battery or "go thinner". ) .
Intel has outlined several of the parts of the initial 10nm attempt went wrong at this point. "Admitting" isn't really the problem. I think some folks are way too caught up in whether Intel is calling what they are currently implementing "10nm" versus 10nm revised" , "10nm+" , "10nm delta". etc.
Intel CPU line up has been throwing anything and everything at the wall for a long time. It just worked much better when AMD was shooting themselves in the foot most of the time and the ARM vendors were myopically focused on mobile phones and tablets. It also worked better when they were doing better as "keep it simple" on the incremental upgrades ( not piling on 4-5 major things at once into a new process and design. )
That fill semicustom processors into every product niche means there is more inertia and niches to fill and they have a limited number of designers to make new stuff to fill them. This "10nm" that isn't the fab+design libraries that was on the roadmap 4-5 years ago. They restarted some designs and they are gap filling with 14nm retreads in other places while probably also doing 10nm+/10mn++ updates too. Whether 10nm is a complete failure probably can't see until 2020.
7hm requires equipment that Intel didn't long term contract to buy and doesn't fall out of the sky if decide last minute to buy more. Similarly the more folks they assign to retrofitting stuff to 14nm the less folks they have to work on Willy Cove (the follow on to Sunny Cove) 10nm finally optimized design libraries and stuff likely targeted at 7nm.
AMD's current Zen 2 isn't particularly all that good at this very low end mobile space. It isn't exposed because AMD isn't trying to fill that space with new product. AMD probably needs another fab process (or two) to make it work. So not particularly in better shape than Intel is. It helps not to try to sell everything to everybody when have a smaller set of design teams.