We don't really fact check with Apple. Apple doesn't really talk to us. But I was wrapping up from WWDC and flying from West to East coast, so just haven't had much time to sit down.
arn
To fix the damn problem or fess up to poor decision making
Hey, no kidding! But it's either this or that, "I've got ants in my computer!" thread over in the macbook forum.if you work like you read forums, i don't think you'll have that job much longer![]()
is there no way that 10.5.7 could have done this?
Yes like I mentioned earlier in this thread, I don't think this sata cap should prevent people from buying the new mbp or install an ssd because you are still going to see some of the major benefits of it, particularly faster boot time and app loading time. even if apple releases a fix for it, the boot/app load time will be the same.
Interesting and not scientific at all - but I was moving some files via class 6 SD card to my new 13 inch MBP with Intel 160GB SSD and also moving them from the same SD card to my PC desktop which has a 300GB Velicoraptor 10K RPM drive (SATA 3GB Interface).
The specific files were ISO files of Office 2007 Ultimate (545KB), Windows 7 RC1 (2,471,656KB) and Vista Ultimate w/SP2 (3,167,396KB). So altogether about 5.9GB.
To try to make this as equal as possible I put the card into a a USB Sandisk MicroMate reader so that each computer was using the USB 2.0 interface.
PC moving 5.9GB from SD card to 300GB 10000RPM Velociraptor HD - 6 min 40 seconds
New MBP 13inch moving 5.9GB from SD card to 160GB Intel X25 SSD drive - 7 min 18 seconds.
Not sure what this tells us and these are large file - so for what it is worth.
Oh - this was done with Vista with SP2 on the PC and Windows 7 RC1 on the Macbook via bootcamp.
Really? So what other bottleneck there is? Because if an app opens in 8 seconds, I say there's still room for improvement. It could open in 5 with SATA 3 Gb/s, for all we know. Any time you have to wait for something to load, you'd wait less with faster data transfer. Boot time and launch times aren't binary, either slow or fast. Even if they're fast, they could be faster. It's another matter entirely whether you need it or not.
Okay, I think some people need to sit down and take a deep breath and consider some facts.
- We don't know if this is intentional or an error, and if it will be fixed in a firmware update. It may be that in a week this is all moot.
- Even for owners of the fastest SSDs, this doesn't make any real world difference. Why? Pretty simple - even if you can do 250 MB/s sequential transfers, you still need something to transfer to/from that can hit those speeds. With Gigabit Ethernet, you're hitting 100MB/s max, and almost certainly less. FW800 maxes out at 80MB/s generally. There aren't any faster interfaces available. If this were a Mac Pro where you could be transferring between multiple drives using SATA, then it would be a problem, but it's not.
The only time that it's going to matter is if you're doing a substantial amount of copying on the same disk, or if you're somehow generating >150MB/s worth of data (in which case you're most likely doing some sort of research, and using a Mac Pro).- Even when you're copying on the same disk, it's very rare that it would make any real-world difference. Say for example you're copying a dual-layer DVD image (~8.5 GB). Even assuming you could copy data on the same drive at the max read rate (which obviously you can't), the difference between copying that data at 250MB/s (roughly the max read rate of a Vertex - X25-M is a little lower) and 150MB/s (max theoretical for SATA I) is 23 seconds. The real world difference will be far lower (and in fact speeds will be under the SATA I threshold, so again moot). How many people here honestly do those sorts of data transfers regularly enough that 20 seconds makes a big difference?
- Finally, in no way does this impact the major selling point of SSDs - the random reads/writes and access times. Both are in no way constrained by the SATA bus.
Does it suck that they've decreased the spec in one specific case in a new generation of MBP? Sure. But it doesn't make any performance difference outside of benchmarks. And if you just want a machine that puts up pretty benchmarks, build a PC with desktop parts, and then go get a life.
Interesting and not scientific at all - but I was moving some files via class 6 SD card to my new 13 inch MBP with Intel 160GB SSD and also moving them from the same SD card to my PC desktop which has a 300GB Velicoraptor 10K RPM drive (SATA 3GB Interface).
The specific files were ISO files of Office 2007 Ultimate (545KB), Windows 7 RC1 (2,471,656KB) and Vista Ultimate w/SP2 (3,167,396KB). So altogether about 5.9GB.
To try to make this as equal as possible I put the card into a a USB Sandisk MicroMate reader so that each computer was using the USB 2.0 interface.
PC moving 5.9GB from SD card to 300GB 10000RPM Velociraptor HD - 6 min 40 seconds
New MBP 13inch moving 5.9GB from SD card to 160GB Intel X25 SSD drive - 7 min 18 seconds.
Not sure what this tells us and these are large file - so for what it is worth.
Oh - this was done with Vista with SP2 on the PC and Windows 7 RC1 on the Macbook via bootcamp.
I posted this in a related thread, but I think it needs to be reiterated here.
And in reply to aleksandra, it's important to understand how the OS and programs perform disk access. When you load a program, it's not reading one big chunk of disk space, it's lots of small pieces, not necessarily in order. That's why SSDs are so much faster than HDDs in the first place - they excel in the small random access tasks. Latency (how fast a particular piece of data can be accessed) isn't limited by the SATA bus, and the small random reads and writes (that SSDs are best at) only reach around 50 MB/s for reads (i.e. 1/3 of SATA 1 speeds, and 1/6 of SATA 2), and even less for writes on the best SSDs.
The sequential speeds really only matter when operating on large files all at once, doing things like copying disk images, etc. And as explained above, there are other limiting factors for those operations.
And in reply to aleksandra, it's important to understand how the OS and programs perform disk access. When you load a program, it's not reading one big chunk of disk space, it's lots of small pieces, not necessarily in order. That's why SSDs are so much faster than HDDs in the first place - they excel in the small random access tasks. Latency (how fast a particular piece of data can be accessed) isn't limited by the SATA bus, and the small random reads and writes (that SSDs are best at) only reach around 50 MB/s for reads (i.e. 1/3 of SATA 1 speeds, and 1/6 of SATA 2), and even less for writes on the best SSDs.
The sequential speeds really only matter when operating on large files all at once, doing things like copying disk images, etc. And as explained above, there are other limiting factors for those operations.
22 seconds is better![]()