Adam_a

Micron 5100 ECO SSD Review Discussion

8 posts in this topic

The ECO is the read-intensive model of Micron’s new 5100 family of data-center drives. This SATA drive comes in capacities up to 8TB for the 7mm 2.5” model and up to 2TB for the M.2 model. The drive has security features such as AES-256-bit encryption and users will be able to optimize the drive using Micron’s FlexPro firmware. According to Micron, the new drives are targeted for use cases such as Big Data, content delivery, database management, virtualized environments, and open compute.

 

Micron 5100 ECO SSD Review

Share this post


Link to post
Share on other sites

Have you considered running non-OLTP tests on Read-Intensive drives? Seems like it would make more sense as this drive obviously isn't targeted toward the workloads you tested...

Share this post


Link to post
Share on other sites

That is a valid point, but a bigger problem is this isn't the first time we've seen this. They seem to slump in workloads that aren't synthetic, which as you can see the primary comparable to them in that Toshiba comparable didn't have the problem in either workload.

We do have the larger capacity Max drive coming in soon, which is more geared towards these workloads to see how it stacks up.

 

 

Share this post


Link to post
Share on other sites

Performance over the lifespan of the drive isn't a great metric to go by. If its too slow to accomplish certain tasks at the onset, knowing it will stay that slow day after day doesn't really help you out that much.

We did get the 1.92TB Max in today, which we will be comparing in the same tests.

Share this post


Link to post
Share on other sites

There is an element of performance as viewed through the lens of a synthetic loadgen tool versus the real world. This is the write speed chart in VMware seen as I'm moving two large VMs onto a datastore sitting on the 1.92TB Max.writespeed.PNG

Share this post


Link to post
Share on other sites

Have you considered using the TPC-H test in Benchmark Factory with SQL Server to test Read Intensive drives?

Also, the fact that the Toshiba RI drive did well in an OLTP test doesn't really tell us anything besides the fact that it can do well in workloads it's not designed for. Car & Driver doesn't test sports cars' ability to tow a boat. And if they did, and one sports car stood out as having good towing ability, we wouldn't think that makes it a better sports car. Right?

Try out TPC-H, should be pretty easy to add it to your process if you're already doing SQL Server and Benchmark Factory.

Share this post


Link to post
Share on other sites

But this isn't a civic pulling a yacht test. Those synthetic benchmarks are used across all drives, workload doesn't change there. This would be like comparing two cheap economy cars and realizing that one not only gets good gas mileage, but handles really well still.

This is the same type of spread you see on NVMe or SAS products. Many drives offer the same or similar level of performance to a degree, but the thing that differentiates each model is the DWPD.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now