A new twist to IntelliPower?

[fzabkar]

For years WD has been obfuscating the true spin speed of their "IntelliPower" models. Misleading statements from a variety of trusted sources, including WD's own support staff, claimed (incorrectly) that these models were capable of varying their spin speed from 5400 RPM to 7200 RPM according to work load. Now it appears that 7200 RPM drives are being passed off as "5400 RPM class".

For example, see the HD Tune benchmark graphs for the WD80EFZX-68UW8N0 and WD80EZZX-11CSGA0. These drives are advertised as "5400 RPM class".

http://www.hddoracle.com/viewtopic.php?f=21&t=2549&start=140

AFAICS, the width of the access time graph would suggest that these drives have a maximum rotational latency of 8ms. This corresponds to the time for one complete revolution at 7200 RPM (8.33ms). The corresponding latency for a 5400 RPM model would be 11ms.

[fzabkar]

Here is a 6TB drive (WD60EFRX-68MYMN0) which is advertised as "5400 RPM class":

Note the "NASware 3.0" on the label. The WWN (50014EE ...) identifies the model as a native WD.

Here is a HD Tune read benchmark for a WD60EFRX-68MYMN1:

Note that spread in the access time data points is about 10ms. This corresponds to one complete revolution of a 6000 RPM drive.

Here is a CrystalDiskInfo SMART report for a WD60EFRX-68LOBN1:

Note that the RPM is 5700. This is the speed reported by the Identify Device ATA command. Earlier "IntelliPower" models would usually report 0 for this word.

So far so good, all the information appears consistent.

Now let's examine the WD80EFZX-68UW8N0. This time CrystalDiskInfo reports a spin speed of 5400 RPM:

The drive's label once again proclaims "NASware 3.0", but this time the WWN identifies the model as a native Hitachi (5000CCA ...). So if the firmware is Hitachi's, then where does WD's NASware 3.0 fit in?

Moreover, HD Tune benchmarks show a spread in the access time data points of about 8ms. This corresponds to one complete revolution of a 7200 RPM drive.

So what is going on? Why the (apparent ?) inconsistency between the observed and reported spin speeds?

Attachments

Western_Digital-3805661149-eng_spec_data_sheet_2879-800002.pdf

(701.15 KiB) Downloaded 945 times

[data-medics]

I suppose the only way to know for sure is to open one, put a black stripe on the top plater (while disabling that head in SA), and run the drive with a digital laser tachometer.

[fzabkar]

One can analyse the noise spectrum to determine RPM:

http://www.silentpcreview.com/article786-page2.html

[fzabkar]

Something for the lawyers ...

http://www.thefreedictionary.com/nominal

nom·i·nal
adj.

1. in name only; theoretical

1. in name only, not in reality

2.
a. Existing in name only; not real

Does this mean that the ATA standard allows a manufacturer to misrepresent the actual rotation rate?

Identify_Device_word_217.gif (6.21 KiB) Viewed 31377 times

[fzabkar]

The WD80EZZX-11CSGA0 helium drive in this thread ...

viewtopic.php?f=1&t=38392

... reports a rotation rate of 5400 RPM in Identify Device word 217 (0x1518 = 5400):

download/file.php?id=18581

Code:

-- IDENTIFY_DEVICE ---------------------------------------------------------
        0    1    2    3    4    5    6    7    8    9
000: 045A 3FFF C837 0010 0000 0000 003F 0000 0000 0000
010: 564B 4A34 444D 3358 2020 2020 2020 2020 2020 2020
020: 0003 0000 0038 3833 2E48 3041 3033 5744 4320 5744
030: 3830 455A 5A58 2D31 3143 5347 4130 2020 2020 2020
040: 2020 2020 2020 2020 2020 2020 2020 8010 4000 2F00
050: 4000 0200 0200 0007 3FFF 0010 003F FC10 00FB 7100
060: FFFF 0FFF 0000 0407 0003 0078 0078 0078 0078 0C18
070: 0000 0000 0000 0000 0000 001F 970E 0034 0CDE 0040
080: 03FC 0029 746B 7D69 4773 7469 BC41 4773 007F 81EB
090: 8001 00FE BBBB 0000 0000 0008 00CA 00F9 2710 0000
100: 2AB0 A381 0003 0000 00CA 0000 6003 5A87 5000 CCA2
110: 54DE 2FDA 0000 0000 0000 0000 0000 0000 0000 40DC
120: 409C 0000 0000 0000 0000 0000 0000 0000 0021 000B
130: 0000 0000 0000 0000 0000 0000 0000 0800 0000 0000
140: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
150: 0000 0007 5757 0000 0000 0000 0000 0000 0000 0000
160: 0000 0000 0000 0000 0000 0000 0000 0000 0002 0000
170: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
180: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
190: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
200: 0000 0000 0000 0000 0000 0000 003D 0000 0000 4000
210: 0000 0000 0000 0000 0000 0000 0000 1518 0000 0000
                                        ^^^^ 0x1518 = 5400
220: 0000 0000 107F 0021 0000 0000 0000 0000 0000 0000
230: 2AB0 A381 0003 0000 0008 0000 0000 0000 0000 0000
240: 0000 0000 0000 4000 0000 0000 0000 0000 0000 0000
250: 0000 0000 0000 0000 0000 DDA5

The audio sample in that thread ...

https://forum.hddguru.com/download/file.php?id=18559

... was analysed by Devonavar of Silent PC Review.

HDD off

HDD on

There is a peak at 120Hz (7200 RPM) but none at 90Hz (5400 RPM).

Thanks for the recording (and thanks especially for including the start-up, which makes it very clear what the difference was between off and on). There's a lot of background noise in the recording, but, because you (@michaelchiklis) were able to get the microphone so close, the evidence is quite clear. This drive is definitely spinning at 7,200 RPM.

I've attached two screenshots which you can use as documentation; one that shows the audio spectrum with the HDD off, and the other with the HDD on. When the drive is turned on, there's very clearly a spike at exactly 120Hz (See the "Peak" box in the screenshot). This corresponds to 7,200 RPM. 7,200 Revolutions per minute / 60 seconds per minute = 120 revolutions per second, or 120Hz. The 120 Hz tone is clearly audible in the recording, and you can hear it increase in pitch as the drive powers up. There's no question in my mind that this tone is produced by the disk rotation.

If the drive was spinning at 5,400, we would expect the spike to occur at 90Hz (5,400RPM / 60 = 90Hz). Similarly, a rotation speed of 5,700 = 95Hz and 6,000 = 100Hz, so the other estimates I saw in the HDDguru thread are also incorrect.

Feel free to quote this and attribute it to me (Devonavar of SPCR) in the thread if you like. Hopefully this can resolve the question for you.

[fzabkar]

Another WD 7200 RPM drive pretending to be 5400 RPM …

https://goughlui.com/2019/06/02/quick-review-western-digital-my-book-10tb-external-usb-3-0-hard-drive/

https://goughlui.com/wp-content/uploads/2019/06/wd10tv-cdismartend.png
https://goughlui.com/static/wd-10tb-audiospec.png

[maximus]

The results so far indicate that some drives spin faster than what they are rated. As long as they spin faster and not slower, I think the manufactures are not on the hook for anything. Sell a drive rated for 5400 and it spins at 7200, that is a bonus because it is faster. When they sell one rated at 7200 and it spins at 5400, then there is a problem.

[fzabkar]

It may be that some people would prefer a genuine 5400 RPM drive with the expectation of lower power consumption, less noise, and higher reliability. In any case there's no excuse for deceptive specifications.

[Keatah]

With all the various and changing models of drives available today, IS there even any need to be deceptive with specifications in the first place?

[fzabkar]

I wonder if there are any models which spin at 5400 RPM but purport to be "7200 RPM class".

[fzabkar]

Here is a weird HD Tune benchmark:

Western Digital Red Pro WD141KFGX 14TB Review (Page 7 of 11):
https://aphnetworks.com/reviews/western-digital-red-pro-wd141kfgx-14tb/7
https://aphnetworks.com/review/western-digital-red-pro-wd141kfgx-14tb/hdtune.png

The access time graph is about 15msec wide. This corresponds to the rotational latency of a 4000 RPM drive.

This is WD's specification page:
https://shop.westerndigital.com/products/internal-drives/wd-red-pro-sata-hdd#WD141KFGX

Recording Technology

CMR
WD2002FFSX (2TB)
WD4003FFBX (4TB)
WD6003FFBX (6TB)
WD8003FFBX (8TB)
WD101KFBX (10TB)
WD102KFBX (10TB)
WD121KFBX (12TB)
WD141KFGX (14TB)

The difference between the WD141KFGX and the WD121KFBX is 512MB cache versus 256MB, and a maximum transfer rate of 255MB/s versus 240MB/s. Both models are listed with a disc speed of 7200 RPM.

https://shop.westerndigital.com/products/internal-drives/wd-red-pro-sata-hdd#WD121KFBX

However, look at the HD Tune graph for the WD121KFBX:

https://www.hardwareluxx.ru/index.php/artikel/hardware/storage/48656-test-i-obzor-western-digital-red-12-tb-wd120efax-i-red-pro-12-tb-wd121kfbx.html?start=2
https://www.hardwareluxx.ru/images/cdn01/E04B1FFBC94C4758ABD4FE3874E9B430/img/3AB1C3E39600420B9C99488F74BA201F/4_WD_RED_PRO_12TB_HD-Tune_R_3AB1C3E39600420B9C99488F74BA201F.jpg

The width of the access time graph is about 8msec, which is what we would expect from a 7200 RPM drive.

However, there is no question that the read performance of the WD141KFGX 14TB model is genuinely "7200 RPM class", so what is going on inside this drive?

Perhaps the clue is in the PCB:
https://aphnetworks.com/reviews/western-digital-red-pro-wd141kfgx-14tb/2

Notice that there are 3 differential pairs running between the MCU and the HDA contacts. Two of these are read channels, the third is the write channel. So it would appear that this drive incorporates TDR technology (Two Dimensional Recording), ie two read elements per head.

The original TDR implementations used these two elements to follow and read a single track rather than two separate tracks.

https://en.wikipedia.org/wiki/Two-dimensional_magnetic_recording

Could it be that this model reads two adjacent tracks concurrently, thereby compensating for the halved RPM?

[fzabkar]

I was wrong about the WD141KFGX. An audio spectrum analysis proves that it does in fact spin at 7200 RPM.

https://aphnetworks.com/forums/topic/7859-5400-rpm-class-hdd-spins-at-7200-rpm/