Sign in with
Sign up | Sign in
Your question
Solved

Difference between WD5003AZEX and WD5002AALX (WD Black 500GB)

Last response: in Storage
Share
October 9, 2012 2:29:51 AM

Hello,

I'm in the market for a WD Black 500GB, I found two variants: WD5003AZEX and WD5002AALX. Apart from the 64MB vs 32MB Cache, what is the difference between the two models? I looked at the WD website and the data sheet (PDF) showed no notable differences. Also which one is better [if any]?

Thanks!
October 9, 2012 10:55:47 AM

Thanks! Sadly though, the document doesn't list something we already don't know about the features. Looks like the only difference is the cache.
m
0
l
Related resources

Best solution

October 10, 2012 1:25:28 AM

The WD5002aalx is the older version. It has a maximum internal transfer speed of 126MB/s sustained. It pulls approximately 6.0w under a load and 0.9w at idle. It uses a single 500GB capacity platter.

The WD5003azex is the newer version. It has a maximum internal transfer speed of 150MB/s sustained. It pulls approximately 6.8w under a load and 0.8w at idle. It uses a single 1TB (theoretical) capacity platter.

Neither one uses advanced format, so compatibility should be good for either one. For my money, I would go with the WD5003azex.

* I just realized that the "Z" should indicate that this is an Advanced Format drive, however, all the other data I can find on the WD5003AZEX points to the contrary. I just received one for my system and there is no indication on the label that it is an Advanced Format drive. I guess I will find out for sure when I try to clone my old drive over to it without using the WD Align utility.
Share
October 10, 2012 7:37:35 PM

Best answer selected by boeder.
m
0
l
October 10, 2012 8:41:15 PM

Update:

I have installed my WD5003AZEX and cloned my old WD2500JD over to it with no hiccups. In my application the WD5003AZEX is benching 126.5 MB/s avg. read (148.69 MB/s max.) and 80.62 MB/s avg. write (105.78 MB/s max.).

My interface is SATA 300 on old hardware. The WD5003AZEX is NOT an AF drive (Yay!). Glad I could be of help!
m
0
l
a c 316 G Storage
October 10, 2012 9:10:00 PM

This document confirms that the WD5003AZEX is not an Advanced Format model:
http://www.wdc.com/wdproducts/library/OVR/ENG/2178-0010...

Therefore ISTM that WD's documentation can't be relied upon.

Thanks for the feedback.

BTW, AFAICT, the WD5003AZEX is benchmarking like a drive that has a data density of 700GB per platter, whereas the WD5002AALX has the transfer rate that is to be expected of a 500GB per platter drive.

Would you mind showing us a HD Tune read benchmark graph? That will tell us if the WD5003AZEX has been shortstroked.

Edit:

I just found this benchmark graph:
http://upic.me/i/c7/21-8-25551-11-29.png

It shows a WD5003AZEX-00K1G with a maximum sustained data transfer rate of 180 MB/s. The minimum is 92 MB/s, ie about half the maximum. This suggests that the drive in the graph is fully stroked.

Its transfer rate is what is to be expected of a non-AF drive with a data density of 1TB per platter, ie it has 1 head whereas your drive benchmarks as if it has 2 heads.

(180 / 126)^2 x 500GB = 1 020GB

ISTM that WD may be playing silly buggers again. Perhaps the DCM or the model number suffix can provide a clue as to the differences. For example, certain characters in the DCM reflect the head type, and are used for matching the electromechanical components.

Can you tell us your drive's DCM (on the label)?


m
0
l
October 11, 2012 3:13:18 AM

fzabkar said:
This document confirms that the WD5003AZEX is not an Advanced Format model:
http://www.wdc.com/wdproducts/library/OVR/ENG/2178-0010...

Therefore ISTM that WD's documentation can't be relied upon.

Thanks for the feedback.

BTW, AFAICT, the WD5003AZEX is benchmarking like a drive that has a data density of 700GB per platter, whereas the WD5002AALX has the transfer rate that is to be expected of a 500GB per platter drive.

Would you mind showing us a HD Tune read benchmark graph? That will tell us if the WD5003AZEX has been shortstroked.

Edit:

I just found this benchmark graph:
http://upic.me/i/c7/21-8-25551-11-29.png

It shows a WD5003AZEX-00K1G with a maximum sustained data transfer rate of 180 MB/s. The minimum is 92 MB/s, ie about half the maximum. This suggests that the drive in the graph is fully stroked.

Its transfer rate is what is to be expected of a non-AF drive with a data density of 1TB per platter, ie it has 1 head whereas your drive has 2 heads.

(180 / 126)^2 x 500GB = 1 020GB

ISTM that WD is playing silly buggers again. Perhaps the DCM or the model number suffix can provide a clue as to the differences. For example, certain characters in the DCM reflect the head type, and are used for matching the electromechanical components.

Can you tell us your drive's DCM (on the label)?


I suspect my drive could do better than it is. I did a little more playing around and it looks like it is operating in PATA mode (at SATA-150 speeds). It's still a lot faster than my old drive, but I'm not realizing it's potential because it isn't in AHCI mode. It's an old rig with a Pentium D 945 and 945 chipset. If I get it running full speed, I'll let you know. For the fun of it; DCM: EHRNKV2CE, MDL: WD5003AZEX-00K1GA0.
m
0
l
a c 316 G Storage
October 11, 2012 4:01:38 AM

If the maximum transfer rate is being throttled by the SATA interface, then you should see a long flat 150 MB/s plateau at the outer zones. The minimum transfer rate would still be 90 MB/s, though. The burst rate should correspond to the interface limit.

This thread discusses the DCM (Drive Configuration Matrix), at least for older WD models:
http://forum.hddguru.com/western-digital-what-dcm-t6488...

It seems that the 5th and 6th characters identify the Media type and Headstack, respectively.
m
0
l
October 11, 2012 6:03:48 AM

fzabkar said:
If the maximum transfer rate is being throttled by the SATA interface, then you should see a long flat 150 MB/s plateau at the outer zones. The minimum transfer rate would still be 90 MB/s, though. The burst rate should correspond to the interface limit.

This thread discusses the DCM (Drive Configuration Matrix), at least for older WD models:
http://forum.hddguru.com/western-digital-what-dcm-t6488...

It seems that the 5th and 6th characters identify the Media type and Headstack, respectively.


Interesting results...

SiSoftware Sandra

Benchmark Results
Drive Index : 124.27MB/s
Results Interpretation : Higher index values are better.
Random Access Time : 16ms
Results Interpretation : Lower index values are better.

Performance vs. Speed
Drive Index : 17.67kB/s/rpm
Random Access Time : 0.002ms/rpm
Results Interpretation : Higher index values are better.

Detailed Benchmark Results
Speed at position 0% : 72.69MB/s (49%)
Speed at position 3% : 148.73MB/s (99%)
Speed at position 7% : 147.01MB/s (98%)
Speed at position 10% : 108.09MB/s (72%)
Speed at position 13% : 144.06MB/s (96%)
Speed at position 17% : 146.89MB/s (98%)
Speed at position 20% : 142.68MB/s (95%)
Speed at position 23% : 141.21MB/s (94%)
Speed at position 27% : 140.96MB/s (94%)
Speed at position 30% : 106.96MB/s (71%)
Speed at position 33% : 133.55MB/s (89%)
Speed at position 37% : 138.03MB/s (92%)
Speed at position 40% : 133.41MB/s (89%)
Speed at position 43% : 135.23MB/s (90%)
Speed at position 47% : 136.59MB/s (91%)
Speed at position 50% : 100.07MB/s (67%)
Speed at position 53% : 133.40MB/s (89%)
Speed at position 57% : 128.73MB/s (86%)
Speed at position 60% : 127.93MB/s (85%)
Speed at position 63% : 138.31MB/s (92%)
Speed at position 67% : 123.68MB/s (83%)
Speed at position 70% : 89.98MB/s (60%)
Speed at position 73% : 125.25MB/s (84%)
Speed at position 77% : 116.56MB/s (78%)
Speed at position 80% : 114.88MB/s (77%)
Speed at position 83% : 119.39MB/s (80%)
Speed at position 87% : 112.10MB/s (75%)
Speed at position 90% : 112.53MB/s (75%)
Speed at position 93% : 115.77MB/s (77%)
Speed at position 97% : 112.74MB/s (75%)
Speed at position 100% : 104.96MB/s (70%)
Random Access Time : 16ms
Full Stroke Access Time : 8ms

Performance Test Status
Run ID : WDC WD5003AZEX-00K1GA0 500GB (SATA600, 3.5", NCQ)
Platform Compliance : x86
System Timer : 3.40GHz
Use Overlapped I/O : Yes
I/O Queue Depth : 4 request(s)
Block Size : 1MB

Volume Information
Capacity : 465.76GB

Physical Disk
Model : WDC WD5003AZEX-00K1GA0
Version : 80.00A80
Interface : SATA
Removable Drive : No
Queueing On : Yes

Not 100% sure what this benchmark can tell us but I'm happy to hear your theory.
m
0
l
a c 316 G Storage
October 11, 2012 8:06:10 AM

Your data are a little difficult to visualise. A graph would have been better.

That said, there are dips at 0%, 30%, 50%, 70%, and 100%. I don't know why this should be so, but we need to ignore them and try to visualise a smooth curve through these points.

The first thing to notice is that the transfer rates decline all the way from 0% to 100%. There is no plateau, so your drive's performance is not being throttled by the SATA interface. If you examine the HD Tune curve for the faster (180 MB/s) version of your model, you will notice that the transfer rate of 150 MB/s occurs at the 50% point (250GB). Therefore, if the interface had been a bottleneck, we would have seen a flat line from 0% to 50%. Instead we are seeing a gradual decline from 149 MB/s to 134 MB/s (if we correct for the dip).

Furthermore, the minimum transfer rate for the faster drive is 92 MB/s whereas your drive's minimum is about 109 MB/s. This suggests that your drive has been shortstroked. The reduction in performance at the 100% mark is ...

(109 / 149) = 73%

The 73% mark for the faster drive occurs when its transfer rate drops to ...

(109 / 149) x 180 = 132 MB/s

This occurs at about the 350GB point on the HD Tune graph, ie about 70% of total capacity.

Therefore, this suggests that the full capacity of your drive's platters is ...

500GB / 0.70 = 714 GB

An alternative way to infer the platter capacity is by comparing the transfer rates between your drive and a model that is known to have 500GB platters, ie ...

(149 / 126)^2 x 500GB = 699 GB

In short, it appears that your drive has a single shortstroked 700GB platter and 2 heads, whereas the drive in the HD Tune screenshot appears to have a single full stroked 1TB platter and a single head.
m
0
l
October 12, 2012 3:44:29 AM

That makes sense in terms of the data available... I find it odd that WD would sell 2 drives that have the same part numbers and supposedly the same performance characteristics with such different internal architectures and such different real-world performance figures. I was curious if the logic of the drive were smart enough to throttle back the drive so it wasn't bouncing off the 150MB/s bandwidth limiter. Sort of like if you were driving a stick-shift and 65mph in 3rd gear was right at redline, you would shift into 4th? Just curious. This drive is over twice as fast as my old one and I got a good deal on it, so it's ok either way but I still feel kind of gypped if my drive is a lesser version of the same drive that someone else is getting 25MB/s better overall performance from.
m
0
l
a c 316 G Storage
October 12, 2012 4:34:36 AM

WD does this kind of thing quite often. In fact I've seen numerous similar threads, in various forums.

Seagate has also started doing it. Their old model numbering system indicated the number of platters in the second last numeric digit (eg the "2" in ST3500620AS), but the new "simplified" model numbers obscure all these physical details.

For example, see http://www.users.on.net/~fzabkar/HDD/GRCC4CD9.TXT

Notice that the ST2000DM001 comes in three flavours, with 4, 5, or 6 heads.

Here is a thread which demonstrates the performance variations:
http://forums.seagate.com/t5/Barracuda-XT-Barracuda-Bar...
m
0
l
October 12, 2012 5:11:55 AM

Thanks for the info. I'm still going to try to get the thing running at SATA-300 speeds with AHCI and NCQ and see if it improves at all, but it does look unlikely that I will get the numbers achieved by that other guy.

I would expect any manufacturer to want to simplify their lines so that all "X" (hard drives in this case) of type "Y" would be the same to keep down production costs and to ensure uniform performance within a line to avoid customer complaints. I guess the average Joe isn't spending a lot of time benchmarking his hard drives and determining the internal architecture based on the results, but still...
m
0
l
a c 316 G Storage
October 12, 2012 5:21:03 AM

Ironically, I have seen one thread in this forum where WD "simplified" their product line by taking a 750GB drive with two 500GB platters and then shortstroking it to reduce it to 500GB. One would think that it would have been less expensive to sell the same drive with a single full-stroked 500GB platter, but perhaps they had a shortage of the 500GB models and needed to fill an order in a hurry.
m
0
l
a c 316 G Storage
October 12, 2012 7:21:24 PM

Just to complicate matters further, the WD5003AZEX has a 700GB platter whereas its 2TB and 1.5TB siblings appear to have 600GB platters:

http://www.wdc.com/wdproducts/library/SpecSheet/ENG/287...

Comparing the transfer rates ...

(138 / 126)^2 x 500GB = 600GB

This begs the question, why didn't WD use 3 x 700GB platters instead of 4 x 600GB platters for the WD2002FAEX?

Are there now 700GB per platter (3 disc) and 1TB per platter (2 disc) versions of the WD2002FAEX?
m
0
l
!