2048 offset vs 63

I am aware that TI on all its variations changes the offset sector from 2048 to 63, in order to avoid a boot repair, what i do not understand is how competing products avoid this, Ghost 11 (Ghost Solution Suite) Ghost 12/14, ShadowProtect and Easy Image - all keep the 2048 offset.

Is there any way to keep the offset of 2048?
Is there any significance to changing it from the vista default of 2048 to 63?

 

This is not the reason. In fact, changing it is what causes the repair to be needed. The change is because that's the way it's been up until this point. Vista is introducing a new standard to support huge drives in the future.

Because they make their programs to keep it and Acronis didn't. It's as simple as that.

Yes. Create an Entire Disk Image. When you restore the Entire Disk Image, the 2048 offset will be kept.

At present, not really. When 10TB drives become common it will make a bigger difference.

A lot of partitioning programs (including XP's and Vista's Disk Management) have problems when partitions are created using a "mix" of the offsets. For the present time, it may be better to stick with all "Vista" or all "63" on a single drive.

 

Thank you for your reply,
Are their any performance issues between 63 & 2048, in the long disk id thread, there is a mention of the reasoning of MS making this change as performance, or tis this only relevance to future multi terabyte drives?

 

any input?

 

It's mostly for future drives with large sectors. I've never run any speed comparison tests, but I doubt you'd see much difference (if any) on today's drives.

Excerpt from this link:

 

It has to do with both reasons (size and performance):

http://support.microsoft.com/kb/923332

 

Heres some interesting findings, on my drive - a WD3200AAKS i ran some benchmarks and the results showed a decrease of 4-8% and some peaks of 10% from converting a 2048 offset primary partition to 63 offset.

 

Very interesting!

I know that the RAID enthusiasts like 2048-sector offset (1 MB boundaries) and large (64k) cluster sizes for performance reasons. One poster reports (see posts 60 and 63) getting 300 MB/s read performance from his array.

 

In my case its a good old partitioned HD with max of 79.3MB/s read.