Why are you here?!
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.
Intel's Upcoming Coffee Lake Processors Up to 30% Faster Than Kaby Lake Chips Coming to Mac Notebooks
- Thread starter MacRumors
- Start date
- Sort by reaction score
The price points on the X series chips are better than I expected. $599 for a i7-7820X seems great to me. AMD's pressure with Ryzen is paying off. Now we just need Apple to release an iMac that is thick enough to support a 140W CPU.
Why are you? He has an opinion and you are trying to block what he has to say.
I believe so, or at least there is on retail drives. The difference may be a lot smaller, but they increase the price by a substantial amount more because they can. But I agree, I would like to see capacity put first, and hopefully we'll see that given the speeds
Closer to 5 years IRC.
AMD made a huge bet on Bulldozer architecture (and it's variants), unfortunately, they completely missed the mark and the Architecture *and it's derivitives) were so lackluster in performance that they were forced to almost immediately go back to the drawing board and start developing RyZen, I believe almost 5 years ago.
Apple does the same
The 13" Macbook Pro's (both touch and non touch) use Intel's 15w dualcore(4 thread) CPU, which are the identical CPU's that Dell and Microsoft are using
it's only in the 15" MacBook pro that you move up to a 65w quad core part.
AMD is forcing them to compete again. Intel has been slow-walking their CPU release schedule for a few years now. Ryzen comes out and all of a sudden we have two major releases in the same year. What a coincidence!
Wireless everything is a pipe-dream in a lot of "pro" scenarios
Some examples: And i'm sure many others can come up with their own
Large Scale database maintenance and administration: if I'm trying to copy a 100GB database to my local drives for maintenance, WiFi is too slow, nor even an option in many data centres that contain sensitive data.
Many photographers / Videographers are dealing with content that ranges in Gigabytes to even Terrabytes now. Transferring to the laptop over WiFi is slow and unreasonable. Nevermind nobody in their right mind is going to sit editting content over a wifi network and keep it "local"
So yeah, an Ideal future, we'd all be 100% wireless. But we're not in that ideal future. Today, There are very strong use cases where "the wireless future" isn't here.
If Apple believes in the Truly wireless future, and wants to push towards it, They will have to invent, or wait for someone else to invent a wireless standard that is capable of high enough data throughput. Considering that most AC wireless caps out at 500mbps, which in real practice isn't even 1/2 of wired ethernet. WiFi isn't there yet.
Ahh the Coffee Lake.. Perfect name for internals of MBP which serves no purpose. Sit in front of it, relax and drink a coffee at Starbucks.
Match made in heaven !!
Match made in heaven !!
The Atari 2600 was sold for 15 years (1977-1992). No one ever complained it was too slow. Yet the games consistently got better and better.
Today's machines are so fast, that no one will even notice the speed bump unless they are using extremely specialized software. Even then, the performance increases that could be gained by actually writing lean and tight code would far exceed the small hardware speed bumps.
It's time for coders to stop being lazy and plugging together blobs of fetid code and start producing lean, elegant code that will stand the test of time.
Today's machines are so fast, that no one will even notice the speed bump unless they are using extremely specialized software. Even then, the performance increases that could be gained by actually writing lean and tight code would far exceed the small hardware speed bumps.
It's time for coders to stop being lazy and plugging together blobs of fetid code and start producing lean, elegant code that will stand the test of time.
I doubt it, the way to Apple's heart is smaller, more battery efficient chips. So they can make their laptops thinner and thinner.
Stick to your principles and only buy a phone or laptop that has no ports or cables.
At this point I don't know whether you are trolling or just insulting the real Pros who need to manage huge data centers through their laptops. All the wireless technologies are utter trash when it comes to speed and reliability.
Wireless tech no matter how good it gets, will never be able to match 10Gbe networks. At least in the next 10-20 years.
We have to wait a long time for your dream to become reality.
No, not in a couple of months. Intel hasn't even finished testing the new chips yet, then needs to produce them and companies need to gear up for production as well.
yeah, NVME drives are a lot more expensive right now than SATA based SSDs.
some examples froom a quick newegg.ca search (I know it's retail and not necessarily indicative of OEM rpicing)
Samsung 850 2.5" SATA III SSD (500mbs rw) $480 $CAD
Samsung 960 m.2 NVME SSD (3200mbps/1900mbps) $650 $CAD
And while having NVME SSD seems like a truly fantastic idea, at some point you hit a point of diminishing returns on performance, especially when you're talking about mobile parts and that even on mobile, a standard SATA SSD tends not to be the bottleneck in performance (there's very little you're going to do that actually leverages that speed of the NVME storage)
I would much rather have an option to save $200 of storage costs, get the larger storage, than get the extreme speed that I can't actually leverage
[doublepost=1496163758][/doublepost]
I'm a little worried. looking into these new "X" series CPU's and intel seems to be taking previous generation Skylake, Increasing the TDP so that they can ramp up the clock speed and reselling them as new, But at the same time, requiring yet a new socket and chipset, which, because they're being deemed "enthousiast", i'm betting that the costs are being moved from the CPU to the chipset / socket.
So while the 8c/16t CPU might be a "steal" at 599, I imagine the extra parts you're going to need, the x299 Motherboard, plus sufficient cooling to handle the increased TDP will end up offsetting a good chunk of the difference.
Last edited:
Interesting right? Wasn't Intel switching to a tick-tock-tock model just a couple of years ago to milk out architecture lines?
Egggs-actly! Intel is trumpeting their engineering work here. There's a long road between testing this wafer on a bench and having your laptop operate Photoshop filters while you have 20 Chrome tabs open.
This not only has be put into production-type machines and tested with actual operating systems that will run on it, such as MacOS, Windows, FreeBSD, etc. But then it has to be tested for heat, efficiency, and other factors. It's a long road from here to there. It gets silly and conspiratorial that Apple is withholding this great tech from consumers out of perspicuous spite for their consumers.
Apple takes pride in releasing products which they have tailored to work properly together. Much work has to be done before MacOS 10.15 "Van Nuys" will be running on these chips properly.
And I'm fully taking in that Apple hypes their tailoring more than it fully matters in the real world. Windows is not so finely tailored, for example, but runs incredibly consistently on different makers' PCs. But compare that to the "LeeeeRoy Jenkins" model of Fedora, and both that and the Windows "good enough" performance is not what to expect from Cupertino.
I think there's furthe issues that Intel and any chip manufacturer are running into with dealing with the gate physics of Silicon.
Once the gates are too small, there's a lot of electron leak. the smaller the gate, the more the leakage, which means that intel can only get their CPU's on such a small process.
Intell has already pretty much stated that they believe 10nm as is might already be too small and that they'll need to find a better inhibitor than silicon if they wish to go smaller, which forced them to drag out the Tock cycle.
it'll be very interesting to see where Intel goes from here. First and foremost, AMD's re-entrance will force Intel to finally drop prices.
Many games / software in the 70s, 80s and somewhat 90s where written by a single person. Back in the 70s and 80s, memory was expensive as was CPU power, so developers were forced to write very efficient code.
Software today is often more complex too...
Today, memory is cheap, hardware is cheap so software does not have to be as optimal as back then. Modern compilers / runtime environments also do a lot of performance optimization.
In many cases, it is not a case of being 'lazy', it is also economics. Software development is expensive, software optimization is expensive, it is also complex. Software products need to ship.
Last edited:
You should get your facts right.
13" Touch-Bar models use 28W parts (with Iris 550 - there is no 15W Skylake that utilises Iris 550)
The 15" CPUs are 45Watt and not 65Watt (these would be desktop CPUs)
The non-touch bar indeed uses a 15W model (of a higher tier GPU), at the same Apple has been very upfront about it replacing the MacBook Air (which was using same-class CPUs).
As long as Apple is depending on Intel for processors Apple/Intel will have to play this game of catching up.
This is not true, the 13" MacBook Pro with the touch bar uses an Intel 28W chip with integrated Iris 550. The non touch bar variant is equipped with a 15W chip (lower clocked) and integrated Iris 540!