Is it Power Plan set to High Performance?
Here are the some additional aspects regarding the performance results you obtained from your system:
- The advertised performance of the Intel® SSD 750 Series is measured using IOmeter* with specific conditions to take advantage of the drive's capabilities, some tools such as AS SSD allow only "fixed" tests, and in most cases the results will be different from the advertised.
For example, for sequential writes, we use 128K Transfer request size, and 1 worker with Q-depth 32, this is not noted in the AS SSD test. For a more detailed description of the testing procedures, please review the Intel® SSD 750 Series Evaluation Guide, page 13 contains Sequential Writes information.
- Make sure you are using the Intel® Solid-State Drive Data Center Family for NVMe Drivers, version 220.127.116.117. The latest Intel driver has a lot better performance than the Windows* in-box driver.
- Confirm that the SSD is connected to a PCIe Gen3 x4 connector. The drive is backwards compatible with lower speeds but will not achieve its best performance with slower connections.
- It is highly recommended to contact the Computer Manufacturer Support and make sure you are using the most recent BIOS for your motherboard.
Yes - PowerPlan is set to High
The Intel Driver is believed to be latest - see the Toolbox image.
The 750 SSD id definately Gen 3 slot
I am using the latest ASUS Bios
Would the AS SSD Test really result in a 50% difference (I would expect some differences but suprised that it would be 50%) ?
Some more information
The image below is from Intel 750 PCIe SSD Review (400GB) | The SSD Review where a Intel 750 SSD was tested using AS SSD (same version as I used)
The results between my use of AS SSD on my Intel 750 are roughly similar to those below except for sequential writes
Sorry - pushed [enter] too fast.
I understand that my testing using AS SSD would be somewhat different from the "official" test results by Intel as per the link within your response.
However, I am puzzled why my results are reasonably similar to the testing done by SSD Review Site and Intel except for Sequential Writes where the difference in results is not of the order of 10 - 20% but is around 50%.
I am after advice as to what could be possibly causing this difference
I cannot refer to the test results using AS SSD since that is not a tool we use for benchmarking, however, I advise you to test the performance using Iometer and the parameters mentioned in the document I provided before.
For further assistance, please export and save the SSD log from Intel® SSD Toolbox and we will be glad to review it.
Also, let us know the exact model of your motherboard and the BIOS version currently used.
Thanks for your response.
I have looked at page 13 of the document of the link re Intel 750 SSD Performance Testing.
I'll use IO Meter to test my SSD tonite when I get home from work, however are you please able to also answer a couple of questions as I have never used IO Meter previously:-
- The Intel 750 SSD Performance Document refers to erasing and then pre- filling the SSD. As this is my Windows 10 OS Drive, I won't be doing this (assume this is OK?)
- As a consequence of point 1 above, are there any changes required to any of the settings from Page 13 of the Document (for example, to ensure that the 8GB writes / reads use empty SSD Space and do not overwrite any existing SSD OS or Data)
Apologies if my questions are basic but want to be certain
Since this is your Boot drive and it is already in use, you do not need to prefill the SSD, however, you might want to perform a shorter test, not the full 20 minutes test.
Thanks - I was also after advice as to whether any of the instructions from p13 re IO Meter needed to change (eg. does the starting disk sector = 0 on a SSD that has not been pre-filled with data but is an existing OS & data SSD ?):-
1. Click the Disk Target tab.
2. Click on Worker
1. 3. Click on the target drive (in this case, the NVMe Intel drive).
4. Set Maximum Disk Size to 16777216 sectors.
5. Leave Starting Disk Sector at 0.
6. Set # of Outstanding I/Os to 32 per target
You can use the same size and sector values recommended in the document, it shouldn't interfere with your existing data.
Please let us know once you are able to complete your testing.
Tried to run IO Meter but got some Unexpected Kernel Trap Error....
So I ran ATTO Hard Disk analysis instead & got these results (128K Sequential Writes is around 1000 MB per Sec)
Which compares favourably with equivalent from SSD Review
So I think that the Intel 750 SSD I have is OK - must just be something with the AS SSD Tool
We are glad to know you were able to obtain performance results closer to the ones mentioned in the Intel® SSD 750 Series Product Specification using a different tool. As you know from the Evaluation Guide, if the test parameters are not set according to the SSD specification, the results may not reflect the real performance you would get under normal operation.
The reason why Iometer did not work in your case, or why AS SSD did not show accurate results escapes the scope of this forum, however, it is good to know that other tools can be used for this purpose.