I'd love to know this as well. Wouldn't the system profiler show the exact model number? There are different versions of the 128GB Samsung drives... some do 90/70 while others do 220/200. That's a big difference.
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.
SSD Comparison Chart
- Thread starter emt1
- Start date
- Sort by reaction score
I'd love to know this as well. Wouldn't the system profiler show the exact model number? There are different versions of the 128GB Samsung drives... some do 90/70 while others do 220/200. That's a big difference.
Mods love everyone, they just don't know how to show it
.Apple uses Samsung drives (either P128 or P256), there have been known Toshiba drives. The drives Apple puts in are the 220/200 MB/s.
Thanks for the info. According to the review, OCZ Summit is also firmware updateable by end users. Not sure how it works with the latest SATA firmware 1.7 released by Apple on mid-09 uMBP. People reported some SSD models are not working well after firmware 1.7 update.
I have the 120GB Vertex SSD from OCZ and have no problems so far. It's quick and fast thought boot up takes about 28 seconds now. I was hoping for under 20 seconds but I guess not.
Meh. My 7200rpm takes around the same time to boot up. Do you mean boot to the login prompt or to the desktop?
You're right the chart has both columns, but except for the Intel and Vertex drives no one has filled in any of the random speeds there.
The max speeds are almost worthless, but the random are much more important. That is why I simply asked that people focus their data collection on the random figures.
Great idea, but without random speed data, the chart isn't very helpful.
Mods love everyone, they just don't know how to show it
Apple uses Samsung drives (either P128 or P256), there have been known Toshiba drives. The drives Apple puts in are the 220/200 MB/s.
Max speeds are essentially marketing info, but useless in real applications. Random speeds are the most valuable for comparative purposes.
Looks like in a few weeks we'll see bigger and faster Intel 34nm SSD on the chart.
http://www.theinquirer.net/inquirer/news/1406326/intel-34nm-nand-ssds-launch-weeks
http://vr-zone.com/articles/intel-ssd-roadmap--bigger--faster/6508.html?doc=6508
http://www.theinquirer.net/inquirer/news/1406326/intel-34nm-nand-ssds-launch-weeks
http://vr-zone.com/articles/intel-ssd-roadmap--bigger--faster/6508.html?doc=6508
emt1, thanks for all your efforts on this. Great work.
As additional info PC Perspective put out their own chart that is totally exhaustive with respect to SSD comparisons.
http://www.pcper.com/article.php?aid=736
As additional info PC Perspective put out their own chart that is totally exhaustive with respect to SSD comparisons.
http://www.pcper.com/article.php?aid=736
It isn't "totally exhaustive" it doesn't show any random speeds - the only ones that really matter and really allow for a proper comparison.
Hate to sound like a broken record, but without the random speeds, the data is mostly useless.
Go away. Seriously. Lots of people like the info, and I specifically stated in the first post that I want more information. You have some cash to drop? Buy some SSDs and test the random 4K speeds.
Well, can you provide some data? Seeing as you are an avid proponent for random read data, please enlighten us.
I like the chart as it is. Can it use more info? Yes it could. But, that requires the OP to devote time and resources into testing them out (and I think he has the first but not the latter). I know what little information we have is useful as many people have based their decision on it and has thus far, been, impressed.
Also, this chart is made by the people here in MacRumors (emt1 just collects the data), not some fancy reviewer. So, we know that what we get in info is truth, not mashed up marketing mojo.
I basically agree that random speeds are pretty key - as with the GB/$ ratio, most ppl are using these as boot/app drives and not for transfering 4.5GB MKVs back and forth. I think if the OP adds the columns for random and you can even copy random read/write data from Anandtech's 'anthology' for the several drives he tested you can make the graph/thread considerably more useful.
I appreciate the data and I will add what I can when my 2 different SSDs arrive in the next day or two.
Anyone know how to check the 2MB random speeds? XBench only tests 4kb and 128kb irrc.
Pure max speeds seem to be an out-dated way to evaluate drive performance, probably more marketing driven than singularly useful to the user. Yeah I know my Intel has slow write speeds, but what this graph doesn't show is how dominant it is in random read and writes, which is why I chose it over the drives that, on this graph, appear superior and certainly better value.
Please consider adding the columns and letting people fill in the data - no one person needs to drop piles of cash to get info on all these drives, the community will fill in the blanks.
I appreciate the data and I will add what I can when my 2 different SSDs arrive in the next day or two.
Anyone know how to check the 2MB random speeds? XBench only tests 4kb and 128kb irrc.
Pure max speeds seem to be an out-dated way to evaluate drive performance, probably more marketing driven than singularly useful to the user. Yeah I know my Intel has slow write speeds, but what this graph doesn't show is how dominant it is in random read and writes, which is why I chose it over the drives that, on this graph, appear superior and certainly better value.
Please consider adding the columns and letting people fill in the data - no one person needs to drop piles of cash to get info on all these drives, the community will fill in the blanks.
I'm sorry I didn't mean to have my post taken the wrong way. I do very much appreciate the hard work put into providing the information, but I was merely asking people to focus on relevent comparative data.
Max speeds can be manipulated and don't reveal true operating performance. If anything, max speeds are just "mashed up marketing mojo."
Many independent commentators discuss the random speeds as by far the best means for comparison, which is in part why many manufacturers do not reveal them.
Perhaps the best source if you want to read more about this is:
http://www.anandtech.com/storage/showdoc.aspx?i=3531&p=1
He discusses performance downgrade, how SSDs work, why random is superior, and how he has been a thorn in manufacturers sides to focus random performance as opposed to max speeds.
There are many other sources out there, but this is the best one. It is long but if you read it you will get a good idea why random speeds are so much more important.
As such, I was merely requesting that we pool the collective resources of those on this board and try and fill out the random speed performance on the drives so that people can better compare. If people who buy the drives try and pass this information along, focusing on random speeds, we can amass a good wealth of information here.
Hal1980 is right. The Vertex drive has better sequential speeds 130 MB/sec vs 71 MB/sec of the intel. But Intel has optimised their SSDs for better random read/write performance and you can see that in the charts from anandtech 23 MB/sec vs 6.5 MB/sec from the vertex. Intel knows most operations in OS are small reads and writes, this is where the X25 is almost 4 times better that the vertex. As well as providing over 3.5 times the IOPS.
So yes the vertex can be as fast as 130 MB/sec but will be as slow as 6.5 MB/sec, that's a wide range. And the more you use it, the slower it will be. Whereas Intel has a range of 71 MB/sec to 23 MB/sec much less dramatic and 4 times better minimum speed. In normal and intensive use Intel wins, in specialised cases eg. transferring large files to a clean drive Vertex wins, but that scenario is likely to happen less than 5% of the time. So overall Intel is the best, which we already knew.
So yes the vertex can be as fast as 130 MB/sec but will be as slow as 6.5 MB/sec, that's a wide range. And the more you use it, the slower it will be. Whereas Intel has a range of 71 MB/sec to 23 MB/sec much less dramatic and 4 times better minimum speed. In normal and intensive use Intel wins, in specialised cases eg. transferring large files to a clean drive Vertex wins, but that scenario is likely to happen less than 5% of the time. So overall Intel is the best, which we already knew.
I did read the article after it came out. I understand that randoms are important, but you must understand that we are not minimizing the importance of random. Just the fact that we can't provide random speeds because we have nothing to test. Can we use help? Yes we can. Anyone who has an SSD listed here can post their results. Those help greatly as we fill in the chart.
Also, performance varies (degrades)in the write portion only, the read of any SSD is never affected as time passes.
Ok, based on Ars Technica's recent article that specualtes that SSDs are not going to price drop for two years, I just purchased two 120GB Vertexes for my two systems. Thanks for the great resource here.
Probably should have waited a week to see what intel releases
Once, when I was younger, a friend of mine showed me his new tattoo and asked me what I thought. I told him I didn't like it. Naturally, my friend's feelings were hurt, but he took it like a man. My dear old daddy was watching this, and after my friend left, he smacked me in the head and told me a very important lesson about kindness: the answer to the question, "Do you like my tattoo?" is always "Yes!"
Your dad is a smart man. I wouldn't have smacked you in the head though, I would have stolen your kidney and thrown you out a window.