G.Skill Titan 256GB SSD

[aka1nas]

I just pulled the trigger on the 128GB model and one of the Athena 5.25-inch backplanes.

http://www.newegg.com/Product/...x?Item=N82E16816119006

 

[Denithor]

Originally posted by: Viper GTS
I hope the sites that are testing these drives are letting them reach steady-state performance before publishing their data.

Which is why I'm waiting until AT reviews the drive.

Can we get a review around here? Pretty please?

 

[Old Hippie]

Originally posted by: aka1nas
I just pulled the trigger on the 128GB model and one of the Athena 5.25-inch backplanes.

http://www.newegg.com/Product/...x?Item=N82E16816119006

Another adventurous soul! :thumbsup:

 

[martensite]

Originally posted by: Denithor
You do want to disable automatic defrag because that will wear out the drive faster (these drives don't get fragmented & more writes = shorter lifespan).

True in a way. Defragging these SSDs often is not a good idea since it uses up write/erase cycles. So turning off auto defrag is essential.

Fragmentation is a function of the filesystem, not the disk itself, so the file system will fragment just as on any other drive. But given the blazing speeds with random and sequential reads, fragmentation will hardly matter in these areas. Same with sequential writes.

From what I've read, the weakness seems to be with random writes, and not due to file fragmentation but free space fragmentation over time. So the Diskeeper guys suggest free space consolidation as one solution to force sequential writes.
http://www.diskeeperblog.com/a...2/hyperfast_is_al.html
http://downloads.diskeeper.com/pdf/HyperFast.pdf

 

[taltamir]

everything fragments, its just that fragmentation has NO negatives on an SSD... even on a spindle drive i dont recommend defragging, since it does more harm then good in most scenarios there as well.

 

[coolVariable]

Originally posted by: Denithor

Originally posted by: Viper GTS
I hope the sites that are testing these drives are letting them reach steady-state performance before publishing their data.

Which is why I'm waiting until AT reviews the drive.

Can we get a review around here? Pretty please?

Seconded!

 

[videogames101]

Anyways, I'm glad to see that I'll finally be able to replace the last mechanical component in my PC.

Solid State Electronics FTW!

 

[aka1nas]

Originally posted by: Old Hippie

Originally posted by: aka1nas
I just pulled the trigger on the 128GB model and one of the Athena 5.25-inch backplanes.

http://www.newegg.com/Product/...x?Item=N82E16816119006

Another adventurous soul! :thumbsup:

Hehe, and with the backplane I can eventually fit all my storage into a single 5.25-inch bay as SSDs get bigger/cheaper and I replace my mechanical drives.

 

[imported_TimBob]

256GB Titan is now available on Newegg for anyone feeling extremely adventurous.

 

[bradley]

Both the 256GB and 126GB drives are in stock, 38 and 56 remaining respectively. I wonder how long before Intel drops their prices. Although there's already a scathing review on the 126GB drive involving stuttering. I'm waiting for more real-world reviews myself.

 

[imported_TimBob]

Originally posted by: bradley
Both the 256GB and 126GB drives are in stock, 38 and 56 remaining respectively. I wonder how long before Intel drops their prices. Although there's already a scathing review on the 126GB drive involving stuttering. I'm waiting for more real-world reviews myself.

Yeah, I noticed that review as well, but not putting much stock in it. Not quite sure why he bought the drive considering just a miniscule amount of research would have shown that the hated jmicron controller is used in the drive.

Mine will be here tomorrow barring any UPS screw up. Hopefully it'll get here early so I have some time to play with it while I'm supposed to be working.

 

[aka1nas]

Bit-tech did a review on the 256GB Titan already:
Link

Edit: It looks like Brazhmy over at XS has a pair of the 128s in RAID 0 and sounds pretty much happy with them(he did note that the write speed is noticeably slower than his previous 3-way VRaptor RAID 0, but I think I can live with that).

 

[Denithor]

Originally posted by: aka1nas
Bit-tech did a review on the 256GB Titan already:
Link

:laugh:

Yes - we know - that's where most of the info in the OP came from...

:beer:

 

[aka1nas]

Originally posted by: Denithor

Originally posted by: aka1nas
Bit-tech did a review on the 256GB Titan already:
Link

:laugh:

Yes - we know - that's where most of the info in the OP came from...

:beer:

*Facepalms*

 

[taltamir]

if it has a raid0 of two jmicrons inside (using a THIRD chip...) and you put it in raid0... is it raid 00?

 

[Rubycon]

Originally posted by: taltamir
if it has a raid0 of two jmicrons inside (using a THIRD chip...) and you put it in raid0... is it raid 00?

I'm shore they use two controllers for porting not to improve throughput by division.

 

[Denithor]

More like raid

 

[martensite]

Originally posted by: Viper GTS

I've been doing a lot of SSD testing in the last few weeks, & every model I've seen slow considerably once every block has been written at least once. This results in drastically lower performance, and greatly increased access times.

Do you have any hypothesis as to why this is happening?

It's upgrade time for my system, and I've already upgraded the video card. CPU is next. I was hoping to put in an SSD as the OS drive...but I don't want to pay an arm and a leg, and end up with crappy performance from the drive after a few weeks.

EDIT: Google gives this interview of an Intel engineer at hardocp.
EDIT#2: http://enthusiast.hardocp.com/...wxLCxoZW50aHVzaWFzdA==

Q. Do SSDs get slower as they get older?

A. This is a complicated issue. Multiple factors can contribute to performance degradation during the life span of a SSD. Most significant is the internal fragmentation of the data stored on the flash media. Unfortunately, there is currently no way to measure it externally. Like I said above, fragmentation is not crippling to a SSD to the extent that it is for HDDs. Benchmark programs might be able to quantify the difference between optimal and fragmented internal layouts, however the user experience should not degrade substantially (at least, not with our drives). Normal usage also should not cause worst case internal fragmentation the way a benchmark program might. There are several ways that SSD optimized file systems should be able to partially alleviate this problem (see later question on file systems).

So is it wear levelling (i.e.'internal' fragmentation) that is leading to the performance degradation over time? Doesn't sound too good :/

 

[martensite]

Originally posted by: taltamir
... even on a spindle drive i dont recommend defragging, since it does more harm then good in most scenarios there as well.

OT: I defrag normal HDDs regularly (auto actually), I find it useful especially on drives with large number of frequently changing files. Never has it done any harm. So, let's agree to disagree lol.

 

[noname98967]

Originally posted by: martensite

Originally posted by: Viper GTS

I've been doing a lot of SSD testing in the last few weeks, & every model I've seen slow considerably once every block has been written at least once. This results in drastically lower performance, and greatly increased access times.

Do you have any hypothesis as to why this is happening?

It's upgrade time for my system, and I've already upgraded the video card. CPU is next. I was hoping to put in an SSD as the OS drive...but I don't want to pay an arm and a leg, and end up with crappy performance from the drive after a few weeks.

EDIT: Google gives this interview of an Intel engineer at hardocp.

Q. Do SSDs get slower as they get older?

A. This is a complicated issue. Multiple factors can contribute to performance degradation during the life span of a SSD. Most significant is the internal fragmentation of the data stored on the flash media. Unfortunately, there is currently no way to measure it externally. Like I said above, fragmentation is not crippling to a SSD to the extent that it is for HDDs. Benchmark programs might be able to quantify the difference between optimal and fragmented internal layouts, however the user experience should not degrade substantially (at least, not with our drives). Normal usage also should not cause worst case internal fragmentation the way a benchmark program might. There are several ways that SSD optimized file systems should be able to partially alleviate this problem (see later question on file systems).

So is it wear levelling (i.e.'internal' fragmentation) that is leading to the performance degradation over time? Doesn't sound too good :/

Diskeeper has a program called Hyperfast which (I think) claims to correct this fragmentation. Can anyone verify this?

 

[aka1nas]

Originally posted by: martensite

Originally posted by: Viper GTS

I've been doing a lot of SSD testing in the last few weeks, & every model I've seen slow considerably once every block has been written at least once. This results in drastically lower performance, and greatly increased access times.

Do you have any hypothesis as to why this is happening?

It's a known limitation. The very first time you write to a block on an SSD, you can just write out the data to the block in question. Every time after that, you would have to then read the block to a cache, erase the block, and then rewrite it with the changed contents. Smarter controllers can hide the additional latency somewhat, and I would hope that more reputable manufacturers are rating their drive write speeds with the latter process in mind.

 

[taltamir]

Originally posted by: Rubycon

Originally posted by: taltamir
if it has a raid0 of two jmicrons inside (using a THIRD chip...) and you put it in raid0... is it raid 00?

I'm shore they use two controllers for porting not to improve throughput by division.

from what i have read there is a splitter that does some load balancing, but not exact striping like in raid0. it has two jmicrons connected to it. I wonder if those two jmicrons are actually the B revision...

 

[martensite]

Originally posted by: aka1nas

It's a known limitation. The very first time you write to a block on an SSD, you can just write out the data to the block in question. Every time after that, you would have to then read the block to a cache, erase the block, and then rewrite it with the changed contents. Smarter controllers can hide the additional latency somewhat, and I would hope that more reputable manufacturers are rating their drive write speeds with the latter process in mind.

Thanks for the info.

1. So it happens only to MLC SSDs?

2a. Will it plateau out after the first erase-rewrite cycle to the block? That is, the speed of writing to the same blocks for the (n+1)th time will be indentical to the (n)th time where n>1. If this is true, then the performance will degrade continuously until all the blocks undergo the write-erase cycle once, and then not deteriorate further (assuming no bad blocks etc). Filling up a new drive completely and then erasing all the filler files leaving only the OS+applications+whatever should immediately indicate the long term steady state performance of the drive then?

or

2b. Does each block(s) have some funny property whereby the (n+1)th erase-write cycle is slower than the (n)th cycle, for n>1?
Either way, it's not very comforting given the price of these drives

 

[aka1nas]

It applies to all SSDs. The performance does not continue to degrade after successive writes, it's just that every write to a block after the first time has to also do that cache+erase operation. I think that some manufacturers may also do a full write cycle prior to shipping the drive, so write performance in those drives would be static.

 

[imported_TimBob]

Why does UPS always want to make me the last stop of the day?