Jump to content
Corsair Community

Performance comparison, X256 vs P256


Cadencia

Recommended Posts

Just for interest's sake, here are benchmarks done on my wife's computer and my computer. The P256, on the left in the uploaded pdf, is in a 3.2GHz quad core Intel, 8G RAM at 1066. The X256, on the right in the pdf, is in a 3.2GHz quad core AMD, 8G RAM at 1066. Both systems are running Win7 ultimate 64, the SSD is the system disk, no tweaks to influence the benchmarks.

 

I did these measurements today, and they are almost identical to the measurements I took just after installation of the drives a couple of weeks ago.

 

Both systems are subjectively excellent - no noticeable difference in responsiveness. The odd microsecond here or there is hard to detect by eye!

Comparison.pdf

Link to comment
Share on other sites

However the benchmarks in your tests show exactly what all ratings and tests on the market say, that the X series is faster by a semi large margin than the P series... all ratings say that the Indilinx is superior to the Samsung, not to say that the Samsung is bad (hey I like Samsung, I have one of their monitors).... Thanks for your consistant excellent testing by the way... I find it fasinating...
Link to comment
Share on other sites

  • Corsair Employees
The P-Series seems to be a bit lower than it should, I would suggest using Parted magic to wipe the drive (Under Tools Select Erase Disk) and then Select the drive from the list and select Secure erase then format the drive using the Quick format with 32K Allocation option and test the drive with Atto and you can use Accronis to image it to another drive.
Link to comment
Share on other sites

Yes, I hear you. But the performance has not changed significantly since new. The sequence of events with this drive is:

 

1. Install latest firmware

2. Install Windows 7 Ultimate 64 bit

3. install other software

4. Benchmark

5. Use for a couple of weeks.

6. Benchmark - the one you see here.

 

I'm not worried by the apparently slightly slow performance - it works well enough. I suspect, in fact, that the TRIM might actually cause a bit of a slowdown during the benchmarks, as the continual write and re-write gets handled by the firmware - that's conjecture, of course, entirely unsupported by testing.

 

In any case, I uploaded the benchmarks out of interest, and so that folk could see one user's experience, not as a complaint. So, thanks anyway, but I'll leave well enough alone.

Link to comment
Share on other sites

The X25 makes for pretty looking benchmarks but in real world performance my X32s as single drives respond faster. I like the 40GB cut down Intel drive but it may not suit everyone.

 

If using drives for a heavy use server etc where a lot of concurrent IO happens and high IOPS are needed then the Intel would make the better choice. However for an all round combined boot application and file creation drive, you can't beat the X series good balanced read and write speeds.

 

Example of a real world test:

 

Created a script to launch Photoshop CS4, Adobe Acrobat, Winamp, Spyware Blaster and Open Office Writer. Executed the script as soon as the desktop became visible and the Intel drive completed loading the applications in 12 seconds. The single Corsair X32 completed consistantly faster at ~8sec. Remember that the applications are being launched while the OS is still loading meaning that the disk is being accessed a lot.

 

The ||X32 also beats the X25M on load times too. I was actually shocked when I first created my RAID array because the animated Windows 7 boot logo didn't even have time to draw the glowing Windows logo before the desktop appeared. Slight usage of the array combined with installing device drivers and start up programs have of course increased boot time. It may sound unfair comparing a single Intel drive to RAID0 X32's but bear in mind that in theory, if AS SSD is 100% believable then the Intel drive pulls better performance figures as a single drive than the pair of X32's in RAID. This just goes to show that a synthetic benchmark doesn't really tell you much about the actual feel and snappiness of a SSD, just RAW numbers alone don't give the full picture.

Link to comment
Share on other sites

Yes, Psycho101, nice to hear from someone with experience of both makes. I've only touched Corsair drives (4 successfully, plus one X256 received yesterday DOA - bummer). Certainly on my own computer (X256) it takes longer for the POST than the actual boot-up of Windows. And, even better, hibernation works properly in Win 7 so it just pops up.

 

But a point from another post of yours: I have been banging on in various posts about alignment of the partition being on a 4K boundary. But I notice you like a 512K boundary. Clearly an offset of 1024K does both, so i'ts an academic point. I had thought 4096 was good because, although the OS can write a 512 byte sector as the minimum amount of data, the SSD can't write less than 4096, as I understand it. So, using the default NTFS allocation of 4K would result quite frequently in data spanning two SSD blocks unnecessarily. I think your point is that the SSD has to erase chunks of blocks 512K at a time. My thought is that it is the firmware that keeps a map of which 4K blocks belong in which 512K chunk, so erasing and wear levelling would be done out of sight of the operating system, thus independant of alignment, even independant of file system. Or is that too simple a scheme, and does the OS get involved after all?

 

I wish there were properly published specifications of how this stuff hangs together.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...