SSD Write Endurance 25nm Vs 34nm

[Christopher]

Anvil,

That's a good of a place to start as any.

How should we enter the information?

[Anvil]

Just PM (or e-mail) me the info and I'll enter the data and make it part of post #1.
(This part does not have to be updated on a frequent basis)

I'll post the other spreadsheet a little later, still playing with the template.

[johnw]

From the reviews I read it appears that the 830 gives up random performance over the 470

I haven't seen that. From the reviews I have seen on the 830, the performance looks similar to the 470 for writes. The 830 is probably faster for writes in the 128GB model (and larger), but probably about the same in the 64GB model.

[Christopher]

johnw,

You might have your opportunity to pick up a Cherryville 60 sooner than we had previously believed.

http://thessdreview.com/latest-buzz/...aid-all-along/

If this is to be believed, it's release is imminent, sometime in the next 3 weeks probably. (I'll believe it when I see it though.)

Intel, you know I'm going to buy one -- even if it's just an over-priced SF2281 with 25nm IMFT. You know, just like the ones that have been around since May.

----------------------------------------------------------------------------

Concerning the 470 vs. 830, here is what Anand Lal Shrimpi wrote in his 512GB 830 review

Just as we saw in Crucial's move to the m4 it looks like random read speed actually dropped a bit between the SSD 470 and 830. And just as we mentioned before, most desktop workloads don't demand super high 4KB random read performance so if this move was done to improve the 830's behavior elsewhere then it is a worthwhile tradeoff. Even with the drop however the 830 delivers the fastest 4KB random read performance of any high-end drive we've reviewed.

I have no frame of reference, since I don't own a 470. The comparison becomes more clouded since the 830 review samples were all 256/512.

[Anvil]

I don't believe the 520 will be seen this side of New Year, latest I've seen was Q1 2012.

If it was to be released shortly there would have been some references to "it" when searching for part# and there is next to nothing to find on those part#.

I did find it interesting that there is a 180GB edition, the gap between 120 and 240 always was a bit on the large side.

[johnw]

You might have your opportunity to pick up a Cherryville 60 sooner than we had previously believed.

No, Intel 520 is delayed until 2012Q1. thessdreview is a source that consistently demonstrates terrible journalism. It was several days ago that an Intel slide was leaked showing the 520 series coming out in 2012Q1. Several sites picked that up. But still that guy makes the comment about it coming out by New Year's. Never take anything you see on thessdreview without a huge grain of salt.

Ao1 even posted it here a couple days ago.

http://www.xtremesystems.org/forums/...=1#post5006751

You do know there are other threads on xtremesystems than just this one, right?

[johnw]

Concerning the 470 vs. 830, here is what Anand Lal Shrimpi wrote in his 512GB 830 review

That is random READ. I was talking about writes. And anyway, all the 2Xnm flash has a slightly lower 4KiB random read than the 3Xnm flash.

[Anvil]

I've posted a preliminary summary in post #1, it is still in the works and I still need some input from all contestants.
(Most info can be found but it would be quicker if you all PM'd me the "missing" info)

[Anvil]

Kingston SSDNow 40GB (X25-V)

523.83TB Host writes
Reallocated sectors : 05 14
Available Reserved Space : E8 99

MD5 OK

35.03MiB/s on avg (~13 hours)

--

Corsair Force 3 120GB

01 75/50 (Raw read error rate)
05 2 (Retired Block count)
B1 53 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 468976 (Raw writes) ->458TiB
F1 624203 (Host writes) ->610TiB

MD5 OK

106.34MiB/s on avg (~13 hours)

power on hours : 1793

[Computurd]

If this is to be believed, it's release is imminent, sometime in the next 3 weeks probably. (I'll believe it when I see it though.)

I dont see any specific dates, just guesses. If you are reading that literally, then you are reading too much into it.

much outside of three weeks, and you are in Q1 2012 anyway. with specs and Product Information Sheets coming out, it isnt long.

Many also decried the very same source when he claimed that the new intel was going to be SF. Quite the show in these very forums of bashing the same source for contending that it would be an SF many many months ago, when no one else had a clue. How many yelled "no way! BS!"

How hard it was not to just let it fly! Of course it turned out to be true, its called 'insider info'.

there's even more interesting things in the future, believe it or not. MUCH more interesting.

....but there is a difference between what people know, and what they can say (at least directly)

and fyi anyone who buys into roadmaps when they arent released directly from Intel, be it Xbit or anyone else who publishes them, needs to have their head examined. Yeah, they are sometimes right, and make good conversation. LOL how many fake roadmaps have you seen? if you've been following tech for a few years, probably only a few thousand.

you need to look at this line in that article..

Intel did not comment on the news-story as the plans are not made public.

are we going to decry Xbit for shoddy journalism now, as well?

[B.A.T]

Todays update:
Kingston V+100
279.2182 TiB
1238 hours
Avg speed 26.04 MiB/s
AD still 1.
168= 1 (SATA PHY Error Count)
P/E?
MD5 OK.
Reallocated sectors : 00


Intel X25-M G1 80GB
92,6313 TiB
19357 hours
Reallocated sectors : 00
MWI=79 to 71
MD5 =OK
42.10 MiB/s on avg


Sadly my local tech dealer didn't have sata cables in stock so I have to take look elsewhere tomorrow.

[johnw]

Many also decried the very same source when he claimed that the new intel was going to be SF. Quite the show in these very forums of bashing the same source for contending that it would be an SF many many months ago, when no one else had a clue. How many yelled "no way! BS!"

How hard it was not to just let it fly! Of course it turned out to be true, its called 'insider info'.

(1) You did not disclose that you are a biased source on the subject of thessdreview, as a responsible journalist would

(2) My main complaint with the original article, as I stated here previously, was that the author apparently had a confidential source for the information, but rather than stating in the article that "this information comes from a confidential source who we vetted and have reason to believe is genuine" as a responsible journalist would, the author made no statement of the kind but rather listed a couple very flimsy pieces of evidence.

Internet tech journalism has never been very good, but thessdreview stands out as the worst of the worst in this regard.

[bluestang]

^^ Sometimes I think you just like to argue with people. It gets old sometimes!

[johnw]

^^ Sometimes I think you just like to argue with people. It gets old sometimes!

Are you trying to start an argument?

[bluestang]

Nope, but you just proved my point. So I will move on back to testing now.

[johnw]

I've posted a preliminary summary in post #1, it is still in the works and I still need some input from all contestants.
(Most info can be found but it would be quicker if you all PM'd me the "missing" info)

Here is some information from my notes on the Samsung 470 that should help you fill in some blanks for that column.


On 2011-08-20, the 64GB Samsung 470 SSD died with the error:

NexusDB: nxEnduranceTbl: INTERNAL: Error when reading from file
[E:\DATA\Endurance.nx1, error 21, The device is not ready]
[$2401/9217].

Also, the ASU error log was full of entries like:

08/20/11 3:01:15PM - Write error 5

Screenshots at xtremesystems.org post #1385:

http://www.xtremesystems.org/forums/...=1#post4932126

It wrote a total of 525.607 TB before it died.

I tested it on a Core i5-2400 machine running Win7 x64

No overprovisioning

42GB static data (41992617078B)

2011-06-29 Started testing
2011-07-05 MWI exhausted at 59.35TiB, sa177 reached normalized 1 - raw 5017
2011-08-20 ended testing at 478.04TiB

113MiB/s average write speed during test

I guess the flash is rated for 5000 P/E cycles since sa177 (~MWI) raw value was 5017 when the normalized value reached 1

FINAL REPORT:
478.037 TiB, 1300 hours, sa177: 1/1/39078, sa178: 2/2/996, sa235: 99/99/2, 3.59 MB/s

Write errors continuously logged, ASU mostly unresponsive (waiting for writes to complete that rarely do?)


storagereview.com has some pictures of the circuit board of a 256GB Samsung 470 SSD. The controller is

Samsung, S3C29MAX01-Y340, NZG0C1, S1017 ARM

and the 256GB model has 16 16GiB flash modules labeled:

Samsung 022, K9HDG08U5M, HCK0, FPA026D2

[johnw]

Nope, but you just proved my point. So I will move on back to testing now.

No, I did not prove your point. Clearly you were trying to start an argument, since such a post could serve no other purpose. Pot, kettle, black.

(Now I proved your point, but of course I also proved my point that you were just trying to start an argument)

[devsk]

No, I did not prove your point. Clearly you were trying to start an argument, since such a post could serve no other purpose. Pot, kettle, black.

(Now I proved your point, but of course I also proved my point that you were just trying to start an argument)

you really like to be the one who has the last word, don't you? Good for you! It will help you in many ways, as long as you control and balance it well!

[Christopher]

Hey BAT,

I'd volunteer to overnight you some sata cables, but apparently overnighting packages to Norway isn't exactly "cost effective" from America...

[Anvil]

I'm pretty sure it would be about as expensive as over-nighting a package within Norway, overnight prices here are insane.

@johnw
Thanks!, I've found most of your notes in the thread, will find out about the NAND but I'm pretty sure it's 32nm DDR NAND

[Christopher]

If BAT can scrounge up a SATA cable by tomorrow afternoon, we can start testing the new M4 and the 830 at the same time. That's an interesting head-to-head match-up.

[B.A.T]

Hey BAT,

I'd volunteer to overnight you some sata cables, but apparently overnighting packages to Norway isn't exactly "cost effective" from America...

Mail in Norway are insanely expensive, so I'll fix it and save you a fortune

If BAT can scrounge up a SATA cable by tomorrow afternoon, we can start testing the new M4 and the 830 at the same time. That's an interesting head-to-head match-up.

I'm going to Lillehammer tomorrow and I'll pass several computershops so you are on

[Christopher]

In this corner, weighing in at .14 pounds, it's the Samsung 830 64GB!

[B.A.T]

And in this corner, weighing, in at 75 gram it's the blue menace, Crucial M4 64GB

[Christopher]

Now that is a fight worth watching!

The Dark Avenger should be here in 18hrs from now. It took a day to actually ship out, but only takes one business day to deliver. I'll be in the lobby waiting for the UPS man to brings us the precious.

I'll run Anvil, CDM, and then maybe some other random bench. Then it goes in the torture chamber.

I'll try to synchronize the start with the M4.