Finally, the new Intel SSD connected to the D815EEA2 motherboard via a SATA to IDE adapter:

I'd say the upgrade was worth it!
Thanks to Phil's Computer Lab for making the benchmarking tools available!

And the million dollar question is... is the partition aligned?

konc wrote:

And the million dollar question is... is the partition aligned?

Unfortunately if he is using FAT16/32 'aligned partition' does not mean anything in terms of speed...
CF vs SD storage
https://msfn.org/board/topic/151798-does-fat3 … n-its-clusters/

Falcosoft wrote:

konc wrote:

And the million dollar question is... is the partition aligned?

Unfortunately if he is using FAT16/32 'aligned partition' does not mean anything in terms of speed...
CF vs SD storage

Thank you very much for this information! 😀 This is very interesting, especially since CF cards and other flash media traditionally
used FAT32 as their primary file system. It would make the old "never re-format that flash card you bought" mantra plausible.
I read the links and the forum thread, but is there more information about this matter ?
I'd really like to know if someone wrote a DOS utility that checks the the actual location of the FAT16 root directory
and does the calculations required to find out alignment issues (FAT32 doesn't have a root directory anymore, from what I read).

This problem also affects 512e Advanced format HDD's. I have not found any official information either. Everyone (HDD manufacturers, Tech sites etc.) pretends 'partition alignment' can solve all problems and that's it.
But you can find many forums where users experienced slowdowns with aligned partitions but FAT formatted media. I have referenced another link above:
https://msfn.org/board/topic/151798-doe ... -clusters/
Unfortunately I do not know any DOS software that can do this. But you can use RMPrepUSB on Windows to get the necessary info. Here's an example of my Advanced Format 1GB Western Digital with 'perfectly 4K aligned' 98GB FAT32 volume shown by RMPrepUSB:

FAT32
000B Bytes Per Sector = 512 (0200h)
000D Sectors Per Cluster = 64 (40h)
000E Reserved Sectors before first FAT = 32 (0020h)
0010 Number of FATs = 2 (02h)
0011 Root Entries = 0 (0000h)
0013 Total Log Sectors (small) = 0 (0000h)
0015 Media Descriptor = 248 (F8h) HDD
0016 Sectors per FAT table = 0 (0000h)
0018 Sectors per Track = 63 (003Fh)
001A Number of Heads per Cylinder = 255 (00FFh)
001C Hidden Sectors preceding Partition = 2048 (00000800h)
0020 Total Log. Sectors (big) = 204797952 (0C34F800h)
0024 Log. Sectors per FAT = 25000 (000061A8h)
0028 Mirroring Flags = 0 (0000h)
002A Version No. = 0 (0000h)
002C Cluster No. of Root Dir Start = 2 (00000002h)
0030 Log. Sector No. of FS Info Sector = 1 (0001h)
0032 First logical sector number of a copy of the three FAT32 boot sectors, typically 6 = 6 (0006h)
0040 Physical Drive Number = 128 (80h) First Fixed Disk
0042 Extended Boot Signature = 41 (29h)
0043 Volume ID Serial No. = 3256109152 (C2144860h)
0047 Volume Label = WDC_FAT
0052 FileSystem Type = FAT32

First FAT begins at LBA 2080
Second FAT begins at LBA 27080
Root Directory begins at LBA 52080
First file data (cluster 0) begins at LBA 52144
Drive 0 WDC WD10EZEX-00BN5A0 F/W Rev.=01.0 Serial No.= [ bytes = 00 00 00 00 00 00 00 00 ]
Reported size 1,000,200,658,432 bytes (931.5094GiB) Last LBA 1,953,516,910
RMPrepUSB Max 1,000,194,048,000 bytes (931.5032GiB) Last LBA 1,953,503,999

@Edit:
In case of Advanced Format 4K sector HDD's it's enough if you can get FAT/Clusters start at an LBA address divisible by 8 (8 * 512 = 4096). But for flash media more strict alignment can be necessary.

I guess you learn something new every day. I originally partitioned and formatted the compact flash card using gparted, paying no attention to alignment, under Tiny Core Linux. The card was split to about 6gb ext2 for Linux and 2gb fat32 for Windows. I used Clonezilla to transfer my setup from the card over to the SSD. I then used Windows fdisk and format to make an extended fat32 partition with the remaining space on the SSD. Of course, the benchmarking was done under Windows. I don't think I'll be doing any re-partitioning anytime soon, but I'll have to read more about partition alignment for future reference.

Of course the OCD in me couldn't let this go...
So I present to you benchmarks for the freshly aligned Intel X25E SSD:

It is clear to me something a little screwy was going on during the original round of testing. Despite that, I think it is pretty obvious there was some improvement even with the Fat32 partitioning.

My assumption is that the SSD needs trimming to keep the SSD aligned.

That's more like it. The original SSD setup was slower than the old platter drive in linear reads.

bjwil1991 wrote:

My assumption is that the SSD needs trimming to keep the SSD aligned.

Partition and data block alignment cannot change after creating the partition and file system. So you do not have to 'keep the SSD aligned'. Trim is another concept to improve speed by informing the SSD controller about space not used by the file system anymore. So alignment and trimming are 2 totally different things that can affect SSD speed.

Ah. Good to know. When I cloned my Windows partition from the 1TB HDD (150GB) to my 480GB, I used a software to allow merging to another drive while making the SSD bootable and aligned the partitions easily without hassle, and the performance was a lot better (takes 30 seconds total from POST to desktop instead of 5 minutes with the HDD). I still have my 1TB HDD for storage and games.