Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Status
Not open for further replies.
I wonder what the 2016 MBPr's will have. I hope that it is at least 2.5gb/s read and 2.0gb/s write. I'll be a really happy man.

Really hoping that they will have NVMe SSDs, which is realistic given that the 6s and 6s+ both have NVMe SSDs. That is the biggest thing I'm looking forward to.
 
CPU inventory, uncertainty about what ports to include. Whether to drop Magsafe or offer a successor. Dedicated GPU or no (and if so which one, etc...). There are plenty of issues they could be hung up on. If it was so easy, I bet we'd have a refreshed MBP by now.

All of that could easily been done in the last year. The how and what of skylake is has been available public the last 6 month. With NDA's they could have had that data for eons, it is not like Apple just received a package from intel today with specs, technical data and a sample.
 
  • Like
Reactions: NJPT
Really hoping that they will have NVMe SSDs, which is realistic given that the 6s and 6s+ both have NVMe SSDs. That is the biggest thing I'm looking forward to.

Apparently the new retina MacBook is already NVMe enabled (but don't know if the SSD it runs is), so I'd say the chances of seeing NVMe on the MBP line are quite good.
 
  • Like
Reactions: cberge
my MBPr 13" 2013 I thought was stupid fast at 700 and 600 read and write speeds but man the 2015 one is double as fast. [...] I hope that [the 2016 MBPr's have] at least 2.5gb/s read and 2.0gb/s write.
I'm wondering when the first system without any Ram at all will appear. Current SSD speeds already surpass SDRAM speeds from ~20 years ago. While it will probably take some more time to reach DDR(2) speeds (let alone DDR3), I could imagine that entry level machines would do away with main Ram once those SSD's have become sufficiently cheap.

Instead all operations would be executed with the SSD doubling as main memory. Surely there are some implications, but still ... No more data transfers from storage to main memory, easily 128GB+ main memory (yes I know - who actually _needs_ that :p) and most of all: reduced footprint of the PCB and (potentially) reduced system cost (for the producer).

I'll go out on a limb and expect the market arrival of such systems within the next 5 years. I'm pretty sure they are already running in some labs.
 
Last edited:
RAM isn't about read/write speed, it's about fast accessing.
That's also the reason why you can't see any real difference in boot times between the average 400 mb/s SATA SSD and some 2000 mb/s PCI-Express ones.
 
This one pool of storage and ram dream is probably going to happen maybe one day - but 5 years? No.

And lab doesn't count. Pricing and reliability are important for a consumer device.
 
I'm wondering when the first system without any Ram at all will appear. Current SSD speeds already surpass SDRAM speeds from ~20 years ago. While it will probably take some more time to reach DDR(2) speeds (let alone DDR3), I could imagine that entry level machines would do away with main Ram once those SSD's have become sufficiently cheap.

Instead all operations would be executed with the SSD doubling as main memory. Surely there are some implications, but still ... No more data transfers from storage to main memory, easily 128GB+ main memory (yes I know - who actually _needs_ that :p) and most of all: reduced footprint of the PCB and (potentially) reduced system cost (for the producer).

I'll go out on a limb and expect the market arrival of such systems within the next 5 years. I'm pretty sure they are already running in some labs.

I highly highly doubt that happens anytime soon. SSD has much lower Access speed, even if its read speeds are relatively decent, and more importantly SSD's have a finite amount of reads and writes they can systain before they become too degraded to be useful. SSD has a write shelf-life so to speak. That right there is the biggest issue of all.
 
I'm wondering when the first system without any Ram at all will appear. Current SSD speeds already surpass SDRAM speeds from ~20 years ago. While it will probably take some more time to reach DDR(2) speeds (let alone DDR3), I could imagine that entry level machines would do away with main Ram once those SSD's have become sufficiently cheap.

Instead all operations would be executed with the SSD doubling as main memory. Surely there are some implications, but still ... No more data transfers from storage to main memory, easily 128GB+ main memory (yes I know - who actually _needs_ that :p) and most of all: reduced footprint of the PCB and (potentially) reduced system cost (for the producer).

I'll go out on a limb and expect the market arrival of such systems within the next 5 years. I'm pretty sure they are already running in some labs.

SSD as ram will never happen. They have limited write cycles, also every time you write something, a whole block has to be erased and rewritten (or moved). Also the latency of ram is about 10.000 faster than than a SSD (SSD 0.1ms vs RAM 0.000015ms).
 
SSD as ram will never happen. They have limited write cycles, also every time you write something, a whole block has to be erased and rewritten (or moved). Also the latency of ram is about 10.000 faster than than a SSD (SSD 0.1ms vs RAM 0.000015ms).

Sorry there is another kind of SSD know as MRAM with unlimited write cycles and retain data for more than 20yr w/o power, also are very fast, just prohibitive expensive for a while.

When MRAM being widely available will be possible to migrate from current computer pc architecture (essentialy kind Von Neumann), to more powerful architectures with wide massive parallelism (kind required for advanced AI).
 
Sorry there is another kind of SSD know as MRAM with unlimited write cycles and retain data for more than 20yr w/o power, also are very fast, just prohibitive expensive for a while.

When MRAM being widely available will be possible to migrate from current computer pc architecture (essentialy kind Von Neumann), to more powerful architectures with wide massive parallelism (kind required for advanced AI).
Sure, but thats like saying I cant wait till Quantum Processors come out. At the moment, its just a research technology.
 
Sorry there is another kind of SSD know as MRAM with unlimited write cycles and retain data for more than 20yr w/o power, also are very fast, just prohibitive expensive for a while.

When MRAM being widely available will be possible to migrate from current computer pc architecture (essentialy kind Von Neumann), to more powerful architectures with wide massive parallelism (kind required for advanced AI).

So assuming that's viable, that's a couple years away at least. Maybe for a MBP that's 2 generations away.
 
I'm thinking instead we'll definitively see something sooner, so probably on March.
Apple can't wait for such a long time, not now that the CPUs are released.
However, I really can't wait to see this thread closed ;) and see then the birth of Waiting for Kaby Lake MBP thread.
That day will be the day Skylake rMBP will be definitively available.
 
Maybe the silence in the media can mean a good thing, intel is probably working on production of the iris 580 chips and i imagine apple is by this time starting to round up the redesign

And for the hell of it and passing time, here's a picture of Mr Carmack I took last night in Amsterdam

image.jpeg
 
I'm thinking instead we'll definitively see something sooner, so probably on March.
Apple can't wait for such a long time, not now that the CPUs are released.
However, I really can't wait to see this thread closed ;) and see then the birth of Waiting for Kaby Lake MBP thread.
That day will be the day Skylake rMBP will be definitively available.


And when kaby lake MBP are expected to be released? march 2017?
[doublepost=1454630459][/doublepost]
Maybe the silence in the media can mean a good thing, intel is probably working on production of the iris 580 chips and i imagine apple is by this time starting to round up the redesign

And for the hell of it and passing time, here's a picture of Mr Carmack I took last night in Amsterdam

View attachment 614338


there is a legal weed in amsterdam, it that the reason you been there? :)
 
I think March is a given at this point despite the lack of leaks.

I think it was said earlier but shipping times for any custom config'd system has slipped to 4-6 days. It will be interesting to track this as we move closer to the event.

I'm thinking instead we'll definitively see something sooner, so probably on March.
Apple can't wait for such a long time, not now that the CPUs are released.
However, I really can't wait to see this thread closed ;) and see then the birth of Waiting for Kaby Lake MBP thread.
That day will be the day Skylake rMBP will be definitively available.
 
Status
Not open for further replies.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.