Micron 5100 ECO 1920GB SATA Enterprise SSD Review

Testing with Different OP Levels

I chose to use a standard pre-conditioning phase (pre-fill followed
by 4K Random Writes, run for 10,800 seconds/3 hours) to test the impact of
using different OP levels. 

I had two objectives in mind. Firstly, could a 1920GB ECO be
made to perform as a 960GB MAX? Secondly, to test if changing the OP level at
the firmware level is more effective than assigning the same level of OP at the
OS level (by, for example, limiting the size of partitions).

The Micron 5100 MAX 960GB has a default OP of 576GB.

The Micron 5100 ECO 1920GB has a default OP of 192GB.

Let’s remind ourselves what the results of this test look
like when run on the Micron 5100 MAX 960GB –

Micron 5100 MAX –
User capacity 960GB, OP 576GB

Micron 5100 ECO 1920GB SATA Enterprise SSD Review

You can see that the 5100 MAX achieves a remarkable result
where there is little or no change between the ‘fresh out of the box’
performance and a steady state of around 68,000 IOPS.

So in a hunt for the same performance, in the first OP test,
I gave the ECO the same amount of OP as the MAX by making a change at the
firmware level using the CLI (Command Line Interface) version of the Micron
Storage Executive software. 

Micron 5100 ECO –
User Capacity 1536GB, OP 576GB set at Firmware Level
Micron 5100 ECO 1920GB SATA Enterprise SSD Review

You can see
that the ECO is not performing at the same level as the MAX when it has the
same level of OP as the MAX.

So next I thought that I should try giving the ECO the same
proportion of OP as the MAX has, i.e. where the OP:User Capacity ratio is the
same ratio as 960:1536. I tested this with the revised User Capacity being set
at the firmware level and by limiting the IO range in the OakGate test platform
(with the default User Capacity applied in the firmware).

Micron 5100 ECO –
User Capacity 1320GB, OP 792 set at Firmware Level

Micron 5100 ECO 1920GB SATA Enterprise SSD Review

You can see
that with the same ratio of OP as the MAX the ECO is not yet performing at the
same level as the MAX (so the same ratio logic was flawed).

Micron 5100 ECO –
User Capacity 1320GB, OP 792 set at OS level (through restricting the IO Range)

Micron 5100 ECO 1920GB SATA Enterprise SSD Review

By comparing
this result to the previous result you can see that making a change to the User
Capacity at the firmware level is significantly more effective than at the OS
level.

Now I didn’t
have any logic left for setting an OP level that would achieve the same
performance as a MAX, so I chose to go the whole way and set the User Capacity
to 960GB.

 

Micron 5100 ECO – User Capacity 960GB, OP 1152 set at
OS level (through restricting the IO Range)

Micron 5100 ECO 1920GB SATA Enterprise SSD Review

You can see that throwing this huge level of OP at the ECO
(even though not applied at the firmware level) has achieved the goal of
enabling the 5100 ECO to perform as a MAX.  What level of OP is actually
required to achieve this goal was not identified in my testing, but it lies
somewhere between an OP Level of 792GB and 1152GB.

Of course increasing the level of OP will also increase the
endurance expectation for a drive but that is not practical for me to test.


Now let’s head to the next page, to look at Power
Consumption and Data Reliability…..