SSD Write Endurance 25nm Vs 34nm

[Christopher]

What controller and NAND?

Phison controller with some kind of Toshiba NAND.

I'm almost 10 minutes in to the first loop and its only written 4GiB. Definitely not a suitable testing drive. I don't know what it would be suitable for... Its basically a Compact Flash card in a 2.5 inch form factor.

The drive's chassis is maybe the most attractive I've seen, but this is the first SSD I've ever wished ill upon. J.Microns are better than this by a long way I think, and they're not very good.



This... is sad. I wasn't expecting much, but this is unimaginably terrible.


Shame on Patriot for ever thinking these drives were a good idea. And this is with the "good" Version 3FW.

This is a Hall of Shame drive. It's going in the closet, never to be spoken of again.

[Christopher]

Samsung 830 64GB Update, Day 3
FW:CXM01B1Q

GiB written:
[B]27270.80/B]

Avg MB/s
107.83

Per-day average:
9090.27

[B.A.T]

@Christopher
Looks like the patriot don't support ncq. There is no scaling from 4K to 4K16QD on both read and write.

[B.A.T]

Todays update:
Kingston V+100
283.3777 TiB
1288 hours
Avg speed 25.29 MiB/s
AD still 1.
168= 1 (SATA PHY Error Count)
P/E?
MD5 OK.
Reallocated sectors : 00


It has dropped out again. This is the 4th time the last week. There is someting wrong with this one. I'm home again tomorrow and I think I'll try an SE and see what happens.

Intel X25-M G1 80GB
108,5066 TiB
19458hours
Reallocated sectors : 00
MWI=64 to 56
MD5 =OK
46.66 MiB/s on avg


m4
20.1124 TiB
73 hours
Avg speed 81.55 MiB/s.
AD gone from 93 to 89.
P/E 346.
MD5 OK.
Reallocated sectors : 00

[Christopher]

@Christopher
Looks like the patriot don't support ncq. There is no scaling from 4K to 4K16QD on both read and write.

It's the worst. drive. ever.

Not supporting NCQ is the least of it's problems. I've used it a little more, and it's actually worse than the numbers suggest.

My mind is blown.

[bluestang]

Still trying to access the M225 since the "Write Error 0" crash and drop out a couple hours ago.
Nothing I have tried in the to get it back has worked this time.
I'll keep trying!

[Christopher]

Still trying to access the M225 since the "Write Error 0" crash and drop out a couple hours ago.
Nothing I have tried in the to get it back has worked this time.
I'll keep trying!

Try the D-flash!
(of course, you might want to take a screen shot of CrystalDiskInfo before you do.... you won't see that near-PB of writes on in SMART, so documentation is key).

[bluestang]

Last CrystalDiskInfo screenshot is from yesterday afternoon. I'd really hate to lose the SMART data by D-flashing. System does recognize it as YADAPONG whatever with the jumper on though. As a last resort I'll try D-flashing tomorrow maybe...MF'er!

[Christopher]

Last CrystalDiskInfo screenshot is from yesterday afternoon. I'd really hate to lose the SMART data by D-flashing. System does recognize it as YADAPONG whatever with the jumper on though. As a last resort I'll try D-flashing tomorrow maybe...MF'er!

You tell that drive who's boss!

[Hopalong X]

It's the worst. drive. ever.

Not supporting NCQ is the least of it's problems. I've used it a little more, and it's actually worse than the numbers suggest.

My mind is blown.

I have read about and have seen a U.S. national TV news story on similar drives coming out of PRC.
No links I don't remember where I saw the news stories.
They ran the gammet from cheap clones of name brands all the way to a USB thumb drive wired up inside an SSD case.
Get home and find you have a 4GB SSD- well a thumb drive NAND inside anyway.
Seems like more work than it would be worth to the scammers.

[Christopher]

I have read about and have seen a U.S. national TV news story on similar drives coming out of PRC.
No links I don't remember where I saw the news stories.
They ran the gammet from cheap clones of name brands all the way to a USB thumb drive wired up inside an SSD case.
Get home and find you have a 4GB SSD- well a thumb drive NAND inside anyway.
Seems like more work than it would be worth to the scammers.

I should point out that the product description didn't include the model... it had been de-labeled, and no mention of the controller was given in the product description. It was refurbished by Patriot, and all of their refurbished SSDs for sale have refurbished part numbers, like RF32SSD, where the RF stands for refurbished. It's a legit product, just not very fast. I was amazed that in 2009/2010 someone at Patriot thought that the Phison controller was going to be a good product.

No, the Phison controller just sucks. It's really just a Compact Flash/SD-style controller adapted for use in SSDs. The PS-100 is actually assembled in America like other Patriot products. It has Toshiba NAND (maybe the 40nm stuff) and the chassis is fantastically beautiful. There's just not much going on in the controller area. But the controller and NAND are made in Japan and the drive still has some kind of warranty.


Actually, I dated a girl for six months who had a really nice chassis too. She didn't did have much going on upstairs, either. She wasn't nearly as cheap as the refurbished Patriot though.

[Hopalong X]

No, the Phison controller just sucks. It's really just a Compact Flash/SD-style controller adapted for use in SSDs. The PS-100 is actually assembled in America like other Patriot products. It has Toshiba NAND (maybe the 40nm stuff) and the chassis is fantastically beautiful. There's just not much going on in the controller area. But the controller and NAND are made in Japan and the drive still has some kind of warranty.


Actually, I dated a girl for six months who had a really nice chassis too. She didn't did have much going on upstairs, either. She wasn't nearly as cheap as the refurbished Patriot though.

No comment on the last sentence.
Banishment would be xtreme.

That suxs the SSD is legit but almost useless.

[B.A.T]

Still trying to access the M225 since the "Write Error 0" crash and drop out a couple hours ago.
Nothing I have tried in the to get it back has worked this time.
I'll keep trying!

So close to 1PB and you have to drag that M225 over the finishing line by its ears.....
It's som typical....

It's the worst. drive. ever.

Not supporting NCQ is the least of it's problems. I've used it a little more, and it's actually worse than the numbers suggest.

My mind is blown.

Have you open it and taken a look under the hood?

[Christopher]

Have you opened it and taken a look under the hood?

I will. I'll post some pictures.

I've seen the inside of one from a review (which also pointed out that it wasn't very good).


I've voided the 30 day warranty.


The details are hard to capture, but it's a nice looking chassis.


Those are TH58NVG5D2ETA20 Toshiba chips, Phison PS3016-S3 controller, Hynix cache


I assume it was originally built in Taiwan, then refurbished in Texas.






EDIT.
So I'm still confused and bewildered by the Phison controlled Patriot. I was wondering what performance would be like with SLC NAND. I imagine it would not be much better.

So I've bought a brand-new 3M Imation S-Class 32GB Phison controlled SLC drive super cheap. I expect to be disappointed by that drive too, but at least it was cheap and has SLC. Maybe it will be a good contender for a new endurance drive (but I'm not getting my hopes up).

Has anyone ever used an Imation S-Class 32GB? It's sure as hell no X25-E, but maybe it's better than the Phailzon 100 Patriot.

I wanted something quirky and off the beaten path to test, and the Patriot let me down. How does any SSD have worse random performance than a Quantum Fireball HDD from 15 years ago? I couldn't complete a run of AS SSD on the Patriot, it would just hang on the random write section. Sequential speeds are decent, but if you encounter any kind of random access, watch out. I don't know if its just starved for cache or if the controller and firmware are just that bad.

[Anvil]

Kingston SSDNow 40GB (X25-V)

539.08TB Host writes
Reallocated sectors : 05 14
Available Reserved Space : E8 99
POH 4890
MD5 OK

34.40MiB/s on avg (~25 hours)

--

Corsair Force 3 120GB

01 79/50 (Raw read error rate)
05 2 (Retired Block count)
B1 46 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 506372 (Raw writes) ->495TiB
F1 673981 (Host writes) ->658TiB

MD5 OK

106.47MiB/s on avg (~145 hours)

power on hours : 1925

WRD decreased by 1 from 47 to 46.

[bluestang]

^ Well, hopefully the Force 3 can break 1PB

[Anvil]

The Force 3 needs to go 2PB vs the other (60-64GB) drives, it's got RAISE and so it will be interesting when things start to develop.

[bluestang]

Well, I tried D-flashing the M225 w/o success ....

Drive is accessible in "factory mode" so I thought I would give it a shot. First, I ran the NAND cleaner by OCZ for my chips and it definitely saw all those bad blocks. Then I attempt to D-flash and it didn't like that at all -- "High Erase Failure" came up a few times along with "corrupt" and "scan list not found" (don't quote me on that one). Finally ended with see good.bin and bad.bin

I think SSD makers need to have/use separate flash to store they're FW and other such pertinent drive info on. This way when someone writes 968TB to a drive they can still flash since it won't be stored on the regular drive NAND and then be able access it.

I'm out of ideas

Edit: I'm taking the next half of the day off of work...I'm too depressed!

[devsk]

Edit: I'm taking the rest half of the day of of work...I'm too depressed!

Boss: what happened? why aren't you at work?
bluestang: I am mourning the death of my beloved M225...
Boss: Get back here!
bluestang: I thought you cared...<tears>

I am working from home today to celebrate 968TB of writes by the M225...

[bulanula]

Sad to see the M225 die like this. Maybe there is another way ? Any news on the 470 that someone here tried to read the NAND directly, One_Hertz ?

[Christopher]

I just assumed that the M225 would kinda last forever. And now it's gone.

Sadness.

[Christopher]

Samsung 830 64GB Update, Day 4
FW:CXM01B1Q

GiB written:
36318.15

Avg MB/s
107.64

Per-day average:
9106.05

96 hours



The average per day is a little higher. I have an alarm that goes off at 14:27 every day so I can take screenshots. Figuring an exact daily average is going to get much more difficult.

Having good SMART data is essential, and it's the worst aspect of the 830. Why would they not provide decent SMART data? Why?

[Anvil]

Well, I tried D-flashing the M225 w/o success ....
...
I'm out of ideas

Edit: I'm taking the next half of the day off of work...I'm too depressed!

What a shame, it did put up a good fight though!

Let's see on Monday before we close the case.

[B][I]
...
Having good SMART data is essential, and it's the worst aspect of the 830. Why would they not provide decent SMART data? Why?

The 830 is a lot better than the 470, there were movements to other attributes on my drives, not just F1.
(I don't think I found what they represented)

[Christopher]

Umm... the Samsung's had some major movement... maybe?
Reallocated sectors went from 0 to 4096



Unless CDI is wrong about those being reallocated sectors, or it only increments in values of 4096.

I tend to think of that as a big deal for a 4 day old drive, but it's possible that it's normal. Maybe that's the equivalent of 1 reallocation with the Mushkin or something.

Maybe it's nothing to worry about.

EDIT

Okay, its not a big deal (probably). I had been taking smart data pics with CDI because all ofthe attributes were unnamed with SSDlife. I installed the installer version instead of the portable and attributes are all there. peep it.

[B.A.T]

Todays update:
Kingston V+100
Still disconnected. I was home today way to late to start with SE and taking a look at it. I'll do this tomorrow.

Intel X25-M G1 80GB
112,7744 TiB
19487 hours
Reallocated sectors : 00
MWI=56 to 45
MD5 =OK
44.70 MiB/s on avg


m4
28.4050 TiB
102 hours
Avg speed 81.38 MiB/s.
AD gone from 89 to 85.
P/E 490.
MD5 OK.
Reallocated sectors : 00


@Bluestang
Sorry about the M225.... It put up a hell of a fight..