Even with that SSD it will still boot slower than a Commodore 64
And you know this to be a fact because you own a recent rMBP and have checked your facts?
Last edited:
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.
New 15-inch Retina MacBook Pro Lives Up to Apple's Claims, Reaches SSD Throughput Speeds of 2GB/s
- Thread starter MacRumors
- Start date
- Sort by reaction score
And you know this to be a fact because you own a recent rMBP and have checked your facts?
As much as I dislike the inability to swap out parts on the retina line, I'm thoroughly impressed by how Apple has made the lifetime of Macs longer over the years.
By that I mean this:
I had a 2009 13" MBP for 5 years, ran like a champ until some soda killed it. Now, I have the 2015 13" MBP 512GB with 16GB of RAM. With these SSD speeds and the RAM i have in it, there is very little likelihood that this machine wouldn't be able to last an average consumer 6+ years and still be more than fast enough for the average user.
By that I mean this:
I had a 2009 13" MBP for 5 years, ran like a champ until some soda killed it. Now, I have the 2015 13" MBP 512GB with 16GB of RAM. With these SSD speeds and the RAM i have in it, there is very little likelihood that this machine wouldn't be able to last an average consumer 6+ years and still be more than fast enough for the average user.
Samsung SSD SM951
A lot of my work involves copying files from one Mac to another, this is reason enough for me to get the 1TB version, it'll more than saturate the Thunderbolt 2 connection which just about reaches 1400MB/s
If only they'd popped this into the Retina iMacs at the same time. I'm more interested in SSD speed than processor increases to be honest, day to day speed is far more affected by drive reading and write - it makes me laugh how people on here only consider "faster" to be a more powerful CPU, they probably barely do anything that totally saturates the CPU very often unless they really are doing video encoding.
It's most likely the Samsung SM951. It's a M.2 SSD that gets exactly those speeds.
Indeed, there are many facets to performance. Random I/O anyone? Personally I want them to increase the max RAM limit with Skylake to at least 32GB. I have a mobile lab that just needs more and more resources, and I can't always remote back into my home lab so I have to be able to take the whole lot with me.
I can just about get it working in 16GB, but then it's pretty much useless. 32GB is okay, it works great with 64GB available. And before anyone asks, yes I really do need this much RAM. Multiple VM's need RAM more than they need CPU.
I would be interested to see random I/O benchmarks of one of these drives though when someone gets their hands on one.
This. This all the way. As a non-power user, I'm more than happy to purchase an over-specced system, whose speed and power I won't make much use of initially, because that's the computer that will still be putting up a fight 5+ years from now when the software I use finally starts to catch up with current processing power.
And you know this to be a fact because you own a recent rMBP and have checked your facts?
He's teasing you
Also from memory a C64 turned on in a couple of seconds.
He's teasing you
Also from memory a C64 turned on in a couple of seconds.
My memory of those was tapes though it looks like there were disks too...it was a long time ago. Had a Tandy, too!
----------
SSDs can fail, too. They can wear out (which turns out to be something that can be demonstrated but the demonstrations mostly impress with just how many petabytes of writes it takes to actually wear one out - TechReport had a long series that was aimed at testing to destruction) and they can also experience sudden, catastrophic failure - both of which can happen to any other electronic device.
On a related point, there was an interesting article on Anandtech recently about SSDs potentially failing from being left unpowered for very extended periods, which IIRC is a theoretical failure mode for HDDs as well but I have powered up HDDs that have been left offline for a decade just to see what happens and have not had any issues with them...
But wouldn't it be super if they kept it at the same size, and utilized the shrinking battery size to, I don't know, put more battery in?
I don't think I can use my 2013 15" rmbp for more than 1.5 hours doing my normal work before it's dead. Hell, even just giving a presentation over apple TV drains it in well under 2 hours.
Q: power consumption with 2 graphics cards
Am I correct that I have no control over whether the integrated card is in use during a particular application?
Is it the need for speed in an app that triggers going to the discrete card? Or does the developer look to see if there is a dedicated card and if so, use it--power consumption be damned!
Finally, is there a setting to turn off the dedicated card--in favor of longer battery life?
Thanks, in advance for the lesson.
Am I correct that I have no control over whether the integrated card is in use during a particular application?
Is it the need for speed in an app that triggers going to the discrete card? Or does the developer look to see if there is a dedicated card and if so, use it--power consumption be damned!
Finally, is there a setting to turn off the dedicated card--in favor of longer battery life?
Thanks, in advance for the lesson.
Not really, a full days charge is fine for me. If I really wanted the extra battery let me plug in a USB-C battery pack when I actually need it with the extra weight that comes with it.
If you are blowing through the battery at 5x the rating of the battery you must be pushing past the intel graphics and activating the discrete graphics and running the CPU flat out.
Either way just give me the option of getting a battery plate that I can put in the bag beside the computer when/if I actually need it.
I also wonder whether the display is set to max brightness? I calibrate my displays for photo use and find that most are far, far too bright for that use. On the other hand, I'm not trying to use them outside in bright sun, either.
Many people are saying that they would never ever pay over 2K dollar on a laptop with AMD graphics card in it. Is there any particular reason for this? How is nvidia better than AMD or vice versa. Please explain thanks.
When I look at mobile workstations, I barf everytime a vendor offers only NVIDIA and move on. Which means most of them.
More like 12 months. Skylake won't make Q3 2015 and most likely won't be in full production until Q2 2016, irregardless of Intel's leaked charts.
Thanks
Appreciate it. No Mac yet, but when it gets here...
So it sounds as though rather than the app seeing what is available--and deciding whether to use it--the OS checks out what the app needs and gives it the resource (I.e.,graphics card)
Appreciate it. No Mac yet, but when it gets here...
So it sounds as though rather than the app seeing what is available--and deciding whether to use it--the OS checks out what the app needs and gives it the resource (I.e.,graphics card)
Last edited:
Not only watching movies, but dealing with Office apps or browsing on Safari. When I had a HDD on my older Macs, there was a noticeable slowness due to the browser being caching/loading every thumbnail and every html page. A simple page required a hundred of disk reads/writes. With a SSD, the speed bump on these operations improve the browsing experience. However, when you're loading a page, you do not feel the difference between loading something at 12ms or 3ms (e.g., loading 3MB of images, decorations, css, etc).
In short, try a better argument for refuting my point. You're talking like if everyone used FCPX in a daily basis.
His argument is valid. If you're not doing anything that requires performance beyond your slowest Mac, you won't see the difference. That doesn't mean there isn't a significant one.
You're talking about processing power. I was talking about I/O performance. Web browsing IS a challenging task, i.e., you really feel I/O bottlenecks when they exist. Think about a 3MB html page, full of thumbnails, external javascript and CSS files. Now imagine this being loaded in a machine with a HDD:
Safari will check for Last-Modified HTTP headers. Then it will check its database for comparing with cached files. If stored files are as recent as the remote file, Safari will load the files from cache. Otherwise, it will replace older files with newer ones. Depending on the page complexity, it will require hundreds or thousands of I/O operations, and it WILL be perceptible to the end-user.
Now think on the same procedure on a SSD. If a single web page sums a total of 10MB in files, it really doesn't matter if you load/store them at 250MB/s or 1000MB/s. The difference between on or another will be a dozen of milliseconds. It will be barely perceptible at best.
You'll only feel the difference of a 1000MB/s and a 250MB/s performance on very demanding I/O operations like RAM swapping (that is, if you have like 1GB to be stored to/loaded from disk). Then you'll note that in the first case the operation takes around 1s while in the second case it will take like 4s.
If only there were a way to do a search for this kind of information. Somebody should come out with some sort of "search engine".
----------
No, it isn't. Not compared to web serving, video editing, video gaming, simulations, etc. There are many things that are much more demanding than web browsing, and web browsing is generally fine on a basic system these days. Take the lowest end of the current Mac Mini for example. It's fine for web browsing.
The initial post which started this "thread" was:
Originally Posted by bigjnyc
My 2009 C2D is still holding its own
Then another poster said it was not. Then I said "no, it really is", and so on.
It's an endless discussion. Of course that you can prove a performance improvement from, e.g., a 2.2GHz Broadwell-U processor and a 2.3 Broadwell-U processor. Will it be noticeable? Well, if you're doing an operation which takes minutes to finish, then yes, the latter will perform around 4% faster.
Going back to the SSD discussion, the 250MB/s vs 1000MB/s difference is only clearly noticeable when swapping RAM, playing/recording huge multitrack audio projects or doing video editing at full hd and higher resolutions. Sporadic load/store operations of data as big as 50MB will, at best, be barely noticeable when comparing 250MB/s and 1000MB/s access speeds.
----------
It's perhaps fine, but you'll note browser caching being populated/accessed. Even with a SATA-I SSD, caching will be imperceptible. I did a test before I put a SSD (SATA-III in a SATA-II bus) on my 2010 Mac Mini.
Firstly, I dropped 16GB of RAM on it and started browsing. It was perceptible laggier than doing the same on my 2009 Macbook or my 2013 rMBP. Then I installed a SSD and page loading speeds were practically undistinguishable between the 3 machines and still are comparing to my wife's 2015 MBA.
I really think it's fantastic having faster SSDs, but they REALLY doesn't put a SATA-I SSD to shame in a lot of common tasks (that is, tasks that any computer user does). It starts being perceptible when you need transferring like >100MB at once. Then you'll feel some latency. Otherwise, there is no perceptible difference between 5ms and 20ms.