Brian

WD Black 4TB Desktop Hard Drive Review (WD4003FZEX) Discussion

44 posts in this topic

Very Impressive, If/when they go to the 1000GB per patter, I can easily see the drive breaching the 200MB/Sec barrier sustained throughput

Share this post


Link to post
Share on other sites

Any extra moving parts are bad. And no I'm not an SSD freak. But I think we're at the stage when 1TB per platter must be standard. For me Seagate's approach to this makes more sense regardless of WD's efforts to clear their 800gb platter inventory ;)

Why don't you like five platters? The helium drives will have more than that.

Share this post


Link to post
Share on other sites

Any extra moving parts are bad. And no I'm not an SSD freak. But I think we're at the stage when 1TB per platter must be standard. For me Seagate's approach to this makes more sense regardless of WD's efforts to clear their 800gb platter inventory ;)

kinda get where you're coming from, imo I think WD did great. now time to grab a few and hope there are no issues with new drive. thanks for the great review Brian, compared with other 4tb and old WD 4tb was awesome.

@storage review, the sequential 128KB bench mark, it says 16 threads 16 queue, does that mean at a higher queue depth?

Edited by unityole

Share this post


Link to post
Share on other sites

I wonder how specifically they improved performance so much, say the sequential read/write speeds without more data density (data per track)? Does it have the piezoelectric "wrist" like the 2TB Caviar Black?

Edited by danwat1234

Share this post


Link to post
Share on other sites

I wonder how specifically they improved performance so much, say the sequential read/write speeds without more data density (data per track)? Does it have the piezoelectric "wrist" like the 2TB Caviar Black?  

They actually improved there performance over the previous generation by just utilizing there 64MB cache in a more optimized way. Now, the cache can be intelligently transferred to a process like reading over writing if it needs more speed. That's the only major difference from the previous generation responsible for increased speed.

Sent from my MK16i using Tapatalk 2

Share this post


Link to post
Share on other sites

But how is the cached used in long sequential operations that would allow for higher speeds without quicker throughput through the heads? Thanks

The only thing missing is a hybrid drive option with this drive.

Edited by danwat1234
1 person likes this

Share this post


Link to post
Share on other sites

I think they fixed whatever problem the previous 4 TB Black had with random access, in addition to what they did to improve caching efficiency. Just look at how good the 2 TB Black is compared to the old 4 TB one in random workloads. This even translates into real applications so far that even there the 4 TB Black is mostly slower than the age-old 2 TB model. The new model looks as superior as a WD Black should be!

What about platter density of the smaller models? It will be difficult to hit 1 TB with 800 GB platters..

MrS

Share this post


Link to post
Share on other sites

Looks good... sadly I think the days of me actually getting super-excited about this are long past, especially given the premium pricing of this and the increasing affordability of SSD's.

Still, seeing this kind of performance improvement within the same harddisk architecture generation is impressive.

Share this post


Link to post
Share on other sites

kinda get where you're coming from, imo I think WD did great. now time to grab a few and hope there are no issues with new drive. thanks for the great review Brian, compared with other 4tb and old WD 4tb was awesome.

@storage review, the sequential 128KB bench mark, it says 16 threads 16 queue, does that mean at a higher queue depth?

Effective queue depth is ThreadsxQueue... although it looks like the label on that chart is wrong. For HDD sequential its 1T/4Q.

1 person likes this

Share this post


Link to post
Share on other sites

I wonder how specifically they improved performance so much, say the sequential read/write speeds without more data density (data per track)? Does it have the piezoelectric "wrist" like the 2TB Caviar Black?

Yes, this line also has the PZT dual-stage actuator. I'm guessing the performance jump is due to code improvements in conjunction with other tweaks.

Share this post


Link to post
Share on other sites

THe previous 4 TB Black had the dual stage actuator as well. As I said, access times are vastly improved and are IMO a prime candidate to cause this speed-up (in addition to the cache tuning). What enabled them to find the tracks this much faster is a different question.. maybe "just" regular read head improvements?

MrS

Share this post


Link to post
Share on other sites

Or they specifically tuned the firmware to cheat benchmarks? ;) Who knows...

an likely explaination. however I do believe sequential performance are hard to cheat, not impossible though. I'd like to think WD is legit as it can be not some samsung phone and SSD tweaked for benchmark softwares =/

Edited by unityole

Share this post


Link to post
Share on other sites

I like to test my desktop drives on Linux, because most vendors ignore it and target AS SSD, DiskMark, HD Tune and other Windows stuff. Running a find(1) with a cold buffer cache gives a good impression about random read access times. Likewise you can mount something -o sync, extract uncompressed tar archives and delete huge directory trees to get a peek on random write performance.

If a drive looks bad there, it may even be the case, that the vendor wants it to look bad, so they can sell expensive enterprise drives with a slightly different firmware instead. -_-

The 8 GB flash caches take the issue to next level. While having almost no effect in practice, because they are to small. 8 GB is just big enough, to make drive look good in benchmarks against SSDs and impress with mostly irrelevant boot times. And that is what sells them.

I'm accusing no-one here, but I'm very skeptical and suspicious. If something looks like a engineering marvel out of nowhere, it most likely isn't true. Remember SoftRAM 95? ;)

1 person likes this

Share this post


Link to post
Share on other sites

an likely explaination. however I do believe sequential performance are hard to cheat, not impossible though. I'd like to think WD is legit as it can be not some samsung phone and SSD tweaked for benchmark softwares =/

I'd be incredibly amazed if someone specifically tried to cheat our benchmarks. Our sample range is 4TB (capacity of the drive), we use direct IO (test isn't buffered) and the testing environment is Linux. I'm not sure how you would get around random access requests skipping around such a massive area.

1 person likes this

Share this post


Link to post
Share on other sites

Is it possible that the old black was a re-branded Hitachi with minor tweaks and that this one is the real?

Probably not. HGST and WD are still separate companies, owned by the same parent WDC. I think the regulatory agencies would frown on rebranding until the firms are completely merged.

Share this post


Link to post
Share on other sites

I'd be incredibly amazed if someone specifically tried to cheat our benchmarks. Our sample range is 4TB (capacity of the drive), we use direct IO (test isn't buffered) and the testing environment is Linux. I'm not sure how you would get around random access requests skipping around such a massive area.

Never underestimate the power of the dark side! They'll analyze all the benchmarks done in reviews, in addition to anything that may make it into benchmarks in the future, and make their drive perform better in them. It's an extensive analysis and really evil business. They'll even admit to doing it.. just ask for "firmware tuning" :P

MrS

Share this post


Link to post
Share on other sites

I'd be incredibly amazed if someone specifically tried to cheat our benchmarks. Our sample range is 4TB (capacity of the drive), we use direct IO (test isn't buffered) and the testing environment is Linux. I'm not sure how you would get around random access requests skipping around such a massive area.

I don't think, they are able to specifically cheat your professional reviews. (They will do some generic optimizations for sure.)

My thoughts did go in a different direction: End users buy the drives, run the popular "measure" tools, which show nice colored bars and post the values into forums and Amazon reviews. On these amateur reviews a consumer drive has to look awesome. And if it looks a bit better than reality, it will just sell better than the competition. ;)

So I bet all vendors sit in their labs and tune the firmware for these artificial benchmarks used by consumers and end up with "optimizations", which have a negligible effect on the real world performance of a drive.

Edited by jtsn
1 person likes this

Share this post


Link to post
Share on other sites

Is it possible that the old black was a re-branded Hitachi with minor tweaks and that this one is the real?

No. They're operating independently.

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