2 different versions of wd blue 3D 2.5" ssd drives. One is significantly slower. Beware.

Kdawg

[H]ard|Gawd
Joined
Aug 12, 2017
Messages
1,116
There's a gimped version of the wd blue 3d ssd being sold at bestbuy and walmart with shit random 4k speeds.


On the outside, they look identical. Both advertised as 3D nand.

2 model numbers: wds500g2b0a and wdbnce5000pnc

The latter is sold at bestbuy and walmart, and you should avoid this shit at all costs.

Bestbuy also sells other WDBxxxxxxPNC model sizes which you should probably avoid as well.
WDBNCE2500PNC-WRSN
WDBNCE0010PNC-WRSN
(edit: this 1tb performs well, made in malaysia fw 401120RL)



I haven't torn down the g2b0a, but I imagine the circuit board and components are the same design as the PNC I tore down previously in another post.

The popular version is made in malaysia. 5000pnc is made in china.


I dunno what's going on, but the bestbuy/walmart version might be using rejected chips? who knows.
But the performance difference is huge.

oddly enough, the firmware is different, however, thousands of user tests have shown that the g2b0a with x61190wd fw was just as fast as the 411040wd fw in my sample.

In this comparison, both drives were made in feb and march 2020....

wdblues.jpg
 
Last edited:
I have one of those in my old machine, bought from Walmart.
Mobo is only Sata 2, will the random 4k be lower due to the interface?
 
My box is the one with the White on top.

View attachment 244752
yep. see, you got the gimped version with the shitty rnd4k q1t1 speeds.... less than 30/60 r/w

normal 4k speeds should be ~40/100 MB/s r/w for modern computers within the last 5 years.


my ancient athlon 64 sata 2 really gimps an ssd's potential.

cdmwdblue.png
 
Last edited:
Do they honestly think enthusiasts are that stupid or is it more aimed at beige box OEMs etc... the fact this crap is on a shelf says they do think we are so stupid. I have never been a fan of WD after past issues others I knew had and this will just keep me away from them even longer.
 
Still several times better than that 5400rpm 320GB Toshiba that was in granny's laptop. Not an issue for most.
 
I have the "whitebox" WD500G2B0A which I purchased from Newegg a few years ago, and I still have the box too. This is my CrystalDiskMark 7 results. I have a 9900K, no MCE and overclocking of any sorts, 2666 RAM, all P and C states enabled, and mitigation is enabled. The controller is using the latest 17.8.0.1065 drivers. Also, only had a 100GB left on the drive when performing the test, and recently a couple of weeks ago I did TRIM the drive. I was recently thinking about getting a 1TB because of cost and performance is decent. Although, I never buy much from brick and mortar I will pay more attention to the product code if I do procure one.

Western Digital SSD 500GB Desktop.png
 
Last edited by a moderator:
I have the "whitebox" WD500G2B0A which I purchased from Newegg a few years ago, and I still have the box too. This is my CrystalDiskMark 7 results. I have a 9900K, no MCE and overclocking of any sorts, 2666 RAM, all P and C states enabled, and mitigation is enabled. The controller is using the latest 17.8.0.1065 drivers. Also, only had a 100GB left on the drive when performing the test, and recently a couple of weeks ago I did TRIM the drive. I was recently thinking about getting a 1TB because of cost and performance is decent. Although, I never buy much from brick and mortar I will pay more attention to the product code if I do procure one.

View attachment 245136

just curious, what does CDM random 4k Q1T1 look like for your samsung sata drives currently filled as is. ?

can you run just the Q1T1 random test and post results?
 
I've seen the original model number come with the firmware revision you have on the Best Buy drive, and it performs normally. Firmware revision is tied to controller + flash. It is possible for two drives with the same firmware version to be optimized differently, however. For example I have an EX920 and SX8200 (Non-Pro) with the same firmware revision - they both use the SM2262 with B16A flash - but they perform differently. The EX920 has a much smaller SLC cache and is faster at low queue depth 4K.

You have to keep in mind that synthetic benchmarks are testing within the SLC cache. This means, for example, that the program time (and thus speed) won't be taking the base flash into account. It's in single-bit mode. Yes, 4K LQD will be lower on DRAM-less drives, different controllers perform differently, etc., but the same controller (w/DRAM) and same type of flash should not be drastically different. This appears to be a firmware issue of some sort.

I appreciate you posting this information and I will be investigating this as I'm already investigating if they've diminished the DRAM at higher capacities. I just don't want to jump to conclusions. FYI your one benchmark is with iaStorA and the other iaStorAC.
 
Last edited:
just curious, what does CDM random 4k Q1T1 look like for your samsung sata drives currently filled as is. ?

can you run just the Q1T1 random test and post results?

Sure, I was curious and did a full bench. I have not done any benching in a long time with any of my drives. The drive has some mileage and wanted to see how it is doing. I do not use RAPID.
Samsung 850 EVO.png
 
I've seen the original model number come with the firmware revision you have on the Best Buy drive, and it performs normally. Firmware revision is tied to controller + flash. It is possible for two drives with the same firmware version to be optimized differently, however. For example I have an EX920 and SX8200 (Non-Pro) with the same firmware revision - they both use the SM2262 with B16A flash - but they perform differently. The EX920 has a much smaller SLC cache and is faster at low queue depth 4K.

You have to keep in mind that synthetic benchmarks are testing within the SLC cache. This means, for example, that the program time (and thus speed) won't be taking the base flash into account. It's in single-bit mode. Yes, 4K LQD will be lower on DRAM-less drives, different controllers perform differently, etc., but the same controller (w/DRAM) and same type of flash should not be drastically different. This appears to be a firmware issue of some sort.

I appreciate you posting this information and I will be investigating this as I'm already investigating if they've diminished the DRAM at higher capacities. I just don't want to jump to conclusions. FYI your one benchmark is with iaStorA and the other iaStorAC.


I actually went through 3 different drives of the 5000pnc variety.
2 from bestbuy and 1 from walmart, and they all perform the same in 4k lqd
granted they are all made within weeks of one another.

and yeah, I had to downgrade my iastorAC driver because it kept crashing during benchmarks.

But I did test the 5000pnc previously with iastorA, and the results were no different....

WDC WDBNCE5000PNC_500GB_1GB-20200428-1703.png




even though you say benchmarks are testing slc cache, it still does make a tangible difference, particularly during boot times.

the 860 evo was almost 4 seconds faster to boot up than my gimped WD 5000pnc.
 
I assumed wrong.

Looks like the bestbuy 1tb version of the wdbnce drive is just fine and performs well, like the best blue ssds.

It has firmware 401120RL, which even some 5000pnc and 2500pnc models have, according to google.

so I guess the x61190wd 5000pnc made in china is the gimped version?

There might be some 5000pnc made in malaysia I'm guessing.

This 1tb 0010pnc is made in malaysia, just like the amazon and newegg versions.


as-ssd-bench WDC  WDBNCE0010P 5.21.2020 2-38-56 AM.png


WDC  WDBNCE0010PNC_1000GB_1GB-20200521-0245.png


wdblue1tb.png


21-May-2020_02-20.png
(4mb blocks)



Inside, the nands are made in malaysia, the board design is also the same as the current 500gb version.
Both 500gb and 1tb have the same amount of dram.

IMG_20200521_033615.jpg
IMG_20200521_032355.jpg
IMG_20200521_032306.jpg
 
Last edited:
I must admit I never bother with firmware on SSDs anymore. Back in the day it was the first thing I did but now so used to them and so common I can't be bothered.
 
firmware issues seem it's only a problem now on intel Enterprise ssds (0MB bug even on recent intel Enterprise SSDs) or the older samsung 840 SSDs or cheap china made micro recycled NAND SSDs
 
I hate that there's two different types of M.2 cards one is SATA and the other is NVMe. When looking at SSDs in general SATA or M.2 I always look up the model number. Sadly WD likes to play tricks with their drives, I try to avoid them. BUT again always look up the model numbers of any drive.
 
There's a gimped version of the wd blue 3d ssd being sold at bestbuy and walmart with shit random 4k speeds.


On the outside, they look identical. Both advertised as 3D nand.

2 model numbers: wds500g2b0a and wdbnce5000pnc

The latter is sold at bestbuy and walmart, and you should avoid this shit at all costs.

Bestbuy also sells other WDBxxxxxxPNC model sizes which you should probably avoid as well.
WDBNCE2500PNC-WRSN
WDBNCE0010PNC-WRSN
(edit: this 1tb performs well, made in malaysia fw 401120RL)



I haven't torn down the g2b0a, but I imagine the circuit board and components are the same design as the PNC I tore down previously in another post.

The popular version is made in malaysia. 5000pnc is made in china.


I dunno what's going on, but the bestbuy/walmart version might be using rejected chips? who knows.
But the performance difference is huge.

oddly enough, the firmware is different, however, thousands of user tests have shown that the g2b0a with x61190wd fw was just as fast as the 411040wd fw in my sample.

In this comparison, both drives were made in feb and march 2020....

View attachment 244732
So... You're right to be angry but, the drives you thought were bad, are actually good. I saw the latest posts and agree.
However, there seems to be a newer issue. I have 2 1tb sata ssd 2.5 WD Blue, dated April 20 2020 - FW 401120RL. They are GREAT!
So nice that I ordered 2 more, both dated June 2 2020 - FW 415000RL. These show as Sandisk G5 Bics4. NOT WD. They are much slower and are several degrees Celsius warmer.
WD SSD Dashboard as well as the Sandisk SSD Dashboards see them and say the firmware are "Up to date". I spent a week with Sandisk but mostly WD support. They simply won't address this. Just some more info for the group here. Thanks! Back to Samsung for me.
 
WD seems to be pulling a kingston with there SSDs with Mixing different type of NAND chips (as there are ones that have more or less bandwidth per NAND chip or the NAND itself is faster or slower)

the slower nand chips are cheaper but WD is not making a New WD colour type SKU for them (well they are but only the crypt part numbers tell you the difference) i had bad experiences with Sandisk controllers so i tend to avoid them (and WD is using sandisk controler and with this mostly Dramless design with built in 10-16MB of SRAM inside the controller to try and mitigate lack of dram)

never had a kingston fail though they mess around with different NAND chips
 
I hate that there's two different types of M.2 cards one is SATA and the other is NVMe. When looking at SSDs in general SATA or M.2 I always look up the model number. Sadly WD likes to play tricks with their drives, I try to avoid them. BUT again always look up the model numbers of any drive.
SATA drives will always have the B-key, you can tell them at a glance. Almost all m.2 SATA drives will be B + M Key these days because most m.2 slots are M-Key only, but work for NVMe and SATA.

M2_Edge_Connector.png
 
Last edited:
I wonder if they're will be a class action lawsuit over these type of shenanigans as well like there is over the SMR fiasco that WD pulled.
 
WD seems to be pulling a kingston with there SSDs with Mixing different type of NAND chips (as there are ones that have more or less bandwidth per NAND chip or the NAND itself is faster or slower)

the slower nand chips are cheaper but WD is not making a New WD colour type SKU for them (well they are but only the crypt part numbers tell you the difference) i had bad experiences with Sandisk controllers so i tend to avoid them (and WD is using sandisk controler and with this mostly Dramless design with built in 10-16MB of SRAM inside the controller to try and mitigate lack of dram)

never had a kingston fail though they mess around with different NAND chips

I cover the technology in their flash on my sub-reddit, including BiCS4, as well as their controllers. For SATA they generally stick to Marvell's 88SS1074. BiCS4 is actually slightly faster than BiCS3 but has other advantages (incl. coming in 512Gb/die). The NVMe controller is a custom tri-core design with 32MB of SRAM or so (on the SN550, the SN750 has DRAM), I also cover that technology to a large extent in my postings. I cover aspects like chip/bus speeds (MT/s) in my SSD Basics as well - 4K performance, however, is a much more complicated subject.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
I'm on the verge of buying this 2TB WD SSD

https://www.amazon.com/M-2-2280-Unidad-estado-sólido/dp/B073SBRHH6/ref=sr_1_8?dchild=1&keywords=ssd+2tb+internal&qid=1597087526&refinements=p_n_feature_three_browse-bin:6797522011&rnid=6797515011&s=pc&sr=1-8

I was going to buy a Samsung 860 EVO same size, but Samsung increased their prices, and its 2TB model is $100 more than the WD. Considerable difference.

Now how high is the risk I get any of these problematic WDs mentioned in this thread? How can I check I'm buying a good unit?
Check the Manufacturer date. June 2, 2020 or newer will be the Sandisk SSD G5 BICS4 JUNK.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
Back
Top