Western Digital Blue disc test – CMR classic versus SMR shingles – appendix first

This article builds on a recent hard drive comparison test in which we built the latest WD Blue 4TB SMR (WD40EZAZ) with shingled magnetic data logging against the conventional WD Blue 3TB CMR model (WD30EZRZ), offering a third higher capacity at a comparable price. We are now supplementing this test with several disciplines, which aim to thoroughly examine the function of the SMR disk in a very high load and find its possible limits.

We published a Western Digital Blue hard drive comparison test in early September 2021. Its purpose was to compare a pair of drives (3TB CMD WD Blue (WD30EZRZ) and 4TB SMR drive WD Blue (WD40EZAZ)) with a comparable purchase price but with different capacity. It was not even a matter of comparing the numbers in the results of individual disciplines, but rather a thorough comparison of the behavior of both disks with regard to their different way of recording data and finding possible problem situations that may surprise us in practice.

When I was going through the price lists of domestic retailers in mid-October to check the sales prices of the tested drives, I found out that the WR Blue CMR models with a capacity of 2 TB and above are hopelessly sold out with an unknown stock. Some vendors even talked about ending their production. However, CMR models are still featured directly on the manufacturer’s website, so we should not lose the choice between CMR and SMR in the future. In the following picture, we will only recall the situation on the market from August 2021.

At the beginning of the September article, I announced that this is my first practical opportunity to really touch a disk with a shingled magnetic data record. Until then, I preferred to avoid SMR disks and, moreover, not enough CMR models on the market pushed me into any change.

For the purposes of the test, I then chose the same default methodology that I used in the previous tests of CMR disks, so that I could compare the results with those previously measured. From synthetic applications, I re-selected HD Tune Pro 5.5, AS SSD Benchmark and PCMark 8. For practical tests, I tried to simulate the tasks that will be most often used on a data disk, and which depend on the disk’s performance. So there were several ways to copy a file folder and load images into Adobe Photoshop. We also compared the consumption of disks and their temperatures. Due to the disk with a shingle record, the discipline of random access using the Iometer 1.1.0 application has been added.

After considering the arguments from the constructive discussion under the September article, I came to the conclusion that the chosen way of using the Iometer application in the test did not have to reveal the real limits of the disk with SMR. Usually I try to create the least favorable situation that the tested disk has to face. Therefore, I chose the goal of the Iometer application as a small partition O with a size of approximately 17 GB, which I created on the last percentage of data space at the inner edge of the plates, where the recording speed should be the lowest due to the lowest peripheral speed of the plates.

In this case, Iometer created a special test file before the start of the test, corresponding to the full size of partition O (ie about 17 GB) and the subsequent three-hour writing of data with random access was directed to this file. I gradually completed six such three-hour disciplines (with different data block sizes), while a total of 979 GB of data (50% write and 50% read) passed through the 17GB section O in one go. But the catch was that the SMR disk probably directed all the write access to the CMR zone, where it constantly overwrited the data in the volume corresponding to the 17GB test file, so the actual write to the SMR zone of the disk probably did not occur at all. That’s probably why I didn’t see any significant decrease in write speeds throughout the test.

As part of the continuation of this comparison test (appendix), this time with the help of the Iometer application I will first try to determine the size of the CMR zone of a shingled disk, so that I can then subject it to several practical disciplines of writing data in adequate volume. Details on the testing methodology can be found in the relevant chapters devoted to individual disciplines.


Source: PCTuning – Všechen obsah by pctuning.cz.

*The article has been translated based on the content of PCTuning – Všechen obsah by pctuning.cz. If there is any problem regarding the content, copyright, please leave a report below the article. We will try to process as quickly as possible to protect the rights of the author. Thank you very much!

*We just want readers to access information more quickly and easily with other multilingual content, instead of information only available in a certain language.

*We always respect the copyright of the content of the author and always include the original link of the source article.If the author disagrees, just leave the report below the article, the article will be edited or deleted at the request of the author. Thanks very much! Best regards!