SSD in old CoreDuo machine?

I have an old CoreDuo MacBook Pro, with the 2GB RAM limit. I spend a lot of time waiting on disk. (iStat Menus will show I have available RAM and minimal CPU usage, but the disk indicator is constantly churning.)

I've been thinking of upgrading the stock 100GB, 5400rpm disk to an SSD. I'm wondering if anyone has tried them in this specific model. I have it in the back of my head that the SATA controller is actually part of the bottleneck, but I have no idea where I read that. I don't necessarily need to actually achieve 200MB/sec. reads and writes, but it's a lucky day if I can exceed 5MB/sec. right now, and that's really awful even for a 5400rpm disk. I don't want to invest $300 on an SSD and find out that the motherboard/SATA chip is going to limit me.

Please note that I am specifically interested in the throughput on an old CoreDuo MBP. There are plenty of other threads about more general SSD questions, but I can't find any existing threads asking what I'm asking, hence the new thread.

Thanks in advance!

Do you have less than 15% free space on that drive? If so, this could be the single issue causing your problem.

Sure, an SSD will work fine in that model. Your disk controller only supports SATA speeds (1.5 Gb/s), but even the best MLC SSDs can only read at about 250 Mb/s, not to mention slower write speeds of 220 Mb/s.

What you have to look out for though is capacity. SSDs will slow down when the volume is at or near capacity. So, although the prices are attractive on 128MB SSDs right now, if they're filled to the brim, you can expect slower performance. Also, performance varies widely on MLC SSDs due to the predominance of crappy controllers in bargain-bin SSD drives. I would recommend one of the Intel X-25 series drives.

Also, I would HIGHLY recommend reading this article before you plunk down any of your hard-earned cash on a bad drive:

AnandTech: The SSD Anthology: Understanding SSDs and New Drives from OCZ

There's a wide variety of SSDs out there, some of which have some serious problems. It helps to study the controllers and technologies behind them before you select a particular model. The worst thing you could do is spend upwards of $400 on a drive that causes your machine to stutter.

I have an early-2008 model, but it still has the same 1.5Gb/s SATA controller as your machine. Still, I've been studying SSDs for awhile now, trying to determine if there is one that will fit my needs for less than $500. It seems like the prices are jacked up right now on most of the decent drives, since Samsung has apparently identified a problem in their controllers.

Thanks all for the replies.

Interesting... I have 35 GB on my 100GB drive free. However, I downloaded Xbench, and some of the numbers seem way off, even for what I'd expect from a 5400rpm laptop drive, which is to say, not much at all. (Granted, Xbench hasn't been released since 2006 I think, but it seems to work still on Snow Leopard.)

Some numbers look fine. Sequential write, 4K blocks: 30.92 MB/sec. (All numbers are without cache.) Sequential reads with 4K blocks are only 14.55MB/sec., but still acceptable enough for a 5400rpm laptop drive.

Random seeks, though, are far worse than I've ever run into. A random write with 256K blocks manages 18.74 MB/sec., but with 4K blocks, it only managed 0.73 MB/sec. (!) The worst case is random reads on 4K blocks, 0.49 MB/sec. (256K blocks manages to bring the number up to 15.43 MB/sec.)

When my machine is really slow, I'll watch the disk meter in iStat Menus, and I'm rarely breaking 5MB/sec., so I'm definitely coming in at the lower end of things.

Are these numbers typical for small random reads/writes? The seem outrageously bad.

XBench is so bad, it's almost silly. It doesn't stop any OS I/O operations, which will frequently cloud any objective measurements.

I would recommend Geekbench for a more realistic test.

I agree with cwa about XBench. It's horrible. I've run it on my machine with pretty decent gear, I maintain this machine and keep it running well and XBench tells me that my Mac runs worse than most other same model MBs.

I wouldn't worry too much about benchmarks. If the machine runs well and you keep it running smoothly, it doesn't matter what a benchmark says. As long as the machine doesn't slow down unnecessarily and you can do what you need without interruption, it doesn't matter if your score is low. My score is low compared to other XBench results but I would venture that I take better care of my machine than most people.