top of page

FCM FlashCore Modules

What are FlashCore Modules (FCM)?

FCM's are NVMe drives that compress data on the drive itself, not at the controller while maintaining micro latency performance. This means there is no performance loss when using FCM for compressed data storage.

What makes IBM FCM's different to competitors offerings? 

FCMs compress and encrypt, whereas the competitors only encrypt. Noteencryption is available with the purchase of licensing

What sizes do FCM's come in?

4.8TB/21.99TB at 4.5:1 compression ratio (Physical/Effective) 

9.6TB/21.99TB at 2.3:1 compression ratio (Physical/Effective)

19.2TB/38.4TB at 2.3:1 compression ratio (Physical/Effective)

38.4TB/87.96TB at 2.3:1 compression ratio (Physical/Effective)

What does Physical and Effective size mean?

Each FCM drive type has two maximum data storage limits, being the physical and effective limit. The physical limit is the maximum uncompressed data that can be stored. The effective limit is the maximum compressed data stored, providing the optimal compression ratio has been reached.  The optimal compression ratio is described with each drive type above, and the 4.8TB is the odd one out with 4.5:1 , whereas the rest of the FCM family are almost half that at 2.3:1 (see below section for explanation).

Suppose we take the 4.8TB FCM as an example. It can store 4.8TB of uncompressed data, maybe an image library repository of jpegs. It has an optimal compression ratio of 4.5:1. If your data is compressable at this optimal level, you will achieve an effective storage capacity of 21.99TB. 

Data with a higher compression ratio fills the metadata space, and the physical portion becomes underutilized, so the maximum effective capacity will not be achieved. The opposite happens when lower compressible data is stored, the physical capacity fills before the metadata capacity, and once again, the effective capacity will not be reached. You will need to know more about your data and its compressibility before choosing the correct FCM module. 

Why does the 4.8TB FCM have a higher compression ratio?

It has the same amount of metadata space as the 9.6TB. Think of metadata space as tickets for compression and data writes, both drives have a fixed amount of tickets to offer to compress and write data.  When there are no more tickets (metadata full) no more compression or data can be stored. 

4.8TB

VS

9.6TB

4.8TB curved Text.png
9.6TB curved Text.png

Although these two drives have the same amount of Metadata space and at first thought, you would ask why would I buy the 9.6TB module when it has the same effective capacity as the 4.8TB? The 9.6TB has double the physical size enabling the drive to hold a larger amount of uncompressed, or low compression ratio data.

If your data can achieve 4.5:1 compression, the 4.8TB  module would be your first choice. The 9.6TB is more effective for data with a compression ratio of 2.3:1 or lower.

What happens if I write a highly compressible workload to an FCM?

Even if you write a 10:1 compressible workload to an FCM, it will still be full when it reaches

the maximum effective capacity, in other words, filling up the Metadata space. Any spare data space remaining on the physical capacity, at this point will be used to improve the performance of the module and extend the wear.

Metadata fills

to 100%

9.6TB with compressed Data.png

Empty space is

used to improve Performance

What capacity is shown for FCMs in the GUI and on the CLI?

The GUI and CLI will show you an estimate of the free physical space based on the data that’s been written and the data reduction achieved with complression.

Can I fill up FCMs past 85%?

If you fill the system more than 85% full then this will mean that the system has more work to do to manage the free space which may impact performance.

bottom of page