First OCZ Core Series SSD results!

[KoHaN69]

Not to ruin the mood, but... are any IDE SSDs planned in production?

[XS Janus]

^That was a real Debbie Downer post.

That said I'm not to impressed with real life scores and behavior of these drives.
Funny things being posted and talked around the web.
Good thing they have a low price, so their main benefit will be to convince other vendors to release better cheaper products soon.
I hope at least 1 or 2 of them release something better till Christmas.

Things like these:
http://www.alternativerecursion.info/?p=106
http://hardforum.com/showthread.php?t=1326446

[RealTelstar]

I would buy a $100 32GB SSD.

I'll buy 12 of those

[flopper]

seems there is an issue with ahci enabled with the drives and services that writes a lot to the disk.
errors seems more to the chipset/OS conflicts than the drives.

If the engineers can run the disks, and have them working (unless they are all lying) on computers it seems more of an user error and some behaviours from chipset/drivers.

[LuckyNV]

^That was a real Debbie Downer post.

That said I'm not to impressed with real life scores and behavior of these drives.
Funny things being posted and talked around the web.
Good thing they have a low price, so their main benefit will be to convince other vendors to release better cheaper products soon.
I hope at least 1 or 2 of them release something better till Christmas.

Things like these:
http://www.alternativerecursion.info/?p=106
http://hardforum.com/showthread.php?t=1326446

seems there is an issue with ahci enabled with the drives and services that writes a lot to the disk.
errors seems more to the chipset/OS conflicts than the drives.

If the engineers can run the disks, and have them working (unless they are all lying) on computers it seems more of an user error and some behaviours from chipset/drivers.

Hmmm, I guess they are a bit of a gamble for the cost, I was tempted for a 60GB version, Janus' links look a bit shocking to me...

[Praz]

I've had a 64GB Core for about a week now. As of yet have seen no install or corruption issues some users are seeing. Unable to do write benches. The trial version of HD Tach doesn't allow it and HDTune needs the drive to be set up as a single partition. Can't do this as I have XP and Vista installed on it. The downward spikes, really evident in WB99, I believe is the testing algorithm falling across block boundaries. If the tests were configured for SSDs in the same way they are for conventional hard drives the results would most likely be different.

Edit: Ran in XP.

[One_Hertz]

Could someone from OCZ comment on whether the serious issues with the writes are permenant? I just ordered 2 64gb ones...

edit: Thats it I am 1337. Never posting again.

[alfaunits]

Praz, you can do random write tests with IoMeter (it does not require a single partition or anything(.

[RyderOCZ]

Could someone from OCZ comment on whether the serious issues with the writes are permenant?

What serious issues with writes?

You guys are all putting your salt in a bunch of benchmarks that were written before an SSD was someone's *** dream.

As this technology continues to unfold, benchmarks are not going to show you what you are used to seeing, nor are the numbers presenting indicative of real performance.

[MnM]

What serious issues with writes?

You guys are all putting your salt in a bunch of benchmarks that were written before an SSD was someone's *** dream.

As this technology continues to unfold, benchmarks are not going to show you what you are used to seeing, nor are the numbers presenting indicative of real performance.

The "serious" issues mentioned here are not benchmarks:

"I just tried copying a 1.5GB file from one directory on the SSD to another, and it was only copying at 16.2MB/s max. As it was copying, it started at about 14MB/s then gradually went up to the 16.2MB/s.

While installing Crysis, which took quite a while (didn't time it, i'd say it was at least 8 min.) I couldn't really do much else on the computer, as if the drive was maxed out and didn't want to process anything else. I only had the Crysis install going and tried to open up My Computer. It didn't like it at all. Took a while to come up."

"Well, I tried copying that same 1.5GB file again, and it actually copied at upto 45MB/s- Finished copying in about 30 seconds this time.

I tried it again a few moments after that just to see if it would do it again, and it slowed down to about 23-25MB/s"

"Well, things ain't looking good at all. There are 2 major issues that I, and some others are having with their OCZ Core SSD Drives.

1. File Corruption in Windows XP and Vista
2. RAID 0 does not work- atleast I have not found anyone who has it working. It fails creating the array." - later on he mentioned that the only wa to create the array wa sin Windows. he was never able to create one in bios.

http://hardforum.com/showthread.php?t=1326446

I think this is what the user wanted some comments on.

[RyderOCZ]

I have posted the same thing over at the OCZ forum.

OCZ cannot reproduce the data corruption, the failing install, or any of this. That leads me to a faulty unit, that needs to be replaced under warranty.. simple as that guys.

When 3000+ units go out the door all on the same day, we are going to see people having issues.

Yes you need to disable AHCI.

These units are MLC, just like you have read... comments about how an Mtron or XXXX unit work (which are SLC based) just fine do not concern me at this time, they are using a different flash technology and subsequently may should not be compared.

[Jazz_Data]

Please keep in mind that one of Ryder's duties is to post here on these boards to "protect" the company's image.

However, ssd's being a new technology and driver / firmware not totally worked out for all possible chipsets / setups, problems will occur.

People will jump on the ssd bandwagon, OCZ will sell their core units like hot cakes, and some will be dissapointed.

As for me, i'll still wait a couple of months (or a year) before everything is solid. Those samsung 256 gigers look promising...

[Buckeye]

What serious issues with writes?

You guys are all putting your salt in a bunch of benchmarks that were written before an SSD was someone's *** dream.

As this technology continues to unfold, benchmarks are not going to show you what you are used to seeing, nor are the numbers presenting indicative of real performance.

I have to agree with Ryder on this. With my Raid different programs report different things and they don't match up often.

What does concern me tho is the graph lines have so much noise in them, on board Raids are simply not the best to test these with imo. That or the testers have simply left on to many back ground apps thats messing things up with those graphs. I generally do not get that much noise, however when I went to Raid 5 I started to see a bit.

http://www.xtremesystems.org/forums/...0&postcount=29

I do believe that any of the benching programs simply do not show the real deal behind these SSD's. That and the Raid card makes a ton of difference.

[Jazz_Data]

I do believe that any of the benching programs simply do not show the real deal behind these SSD's. That and the Raid card makes a ton of difference.

Yes, most of the HDD benchmarking software are written and results based on disk drives with high seek time / caching, etc. They do not "expect" an SSD. Atto does a good job but better benchmarking softs will be written for SSDs...

[RyderOCZ]

Also... regarding Windows installations.

Default drivers on your install disk are not the way to go. Get the drivers slipstreamed onto the install media, you need the latest drivers for your chipset/controller during install.

[flopper]

Also... regarding Windows installations.

Default drivers on your install disk are not the way to go. Get the drivers slipstreamed onto the install media, you need the latest drivers for your chipset/controller during install.

That are not common I guess for many to have all files slipstream onto their install dvd.

And with new technology, which windows most often have been ready for before it was out there are due to be technical issues in the upstart.

Since I read a lot of forums, the user often havent done things properly in the begining and ofc, most just like to install and lift off

[Zytek_Fan]

Isn't this drive made by Super Talent?

[dicecca112]

Other testing apps do not show the dramatic up and down jitter.

1 drive sample on 1 controller, OCZ is on it

I'm curious as to what program OCZ uses to benchmark their drives or which programs are recommended

[bowman]

Isn't this drive made by Super Talent?

No, it'd be more correct to say that both the OCZ and ST drives are made by Samsung. If not made then at least a reference design, I think.

It seems they are branching out and enhancing the controllers themselves now, though. Which is nice. No point in having loads of rebranded identical drives..

Not to ruin the mood, but... are any IDE SSDs planned in production?

Yep, around the same time as the AGP 4870X2.

[Karolis]

Guru3D did a review about the 64 GB version - http://www.guru3d.com/article/ocz-co...-drive-review/

[Praz]

Haven't tried RAID so can't comment on the issues some are seeing. However, the file corruption thing is a complete mystery to me. I installed my drive in a system to test some ram with. The ram is Cell Shock PC2 8000 and the board is the DFI UT X48. As good as this ram is it has not been possible to get a good overclock (over 1060MHz or so) with it on any DFI board since the release of the P35. Oskar has done a new BIOS with unlocked settings not normally available. After countless failures the board is now completely stable at 8x500, 1200 memory. Duel booting XP and Vista64, I used both operating systems for different tests while tuning the board. Even with all the failed overclocks I have yet to see a single corruption issue. Not saying the corruption issue doesn't exist but I would think if it's a common problem I have gone out of my way to try to induce it.

[dicecca112]

http://forums.legitreviews.com/post116997.html#p116997

Some benches by me as well.

[stevecs]

and yet still no random I/O (read & write) benchmarks on these things. Still hoping that someone can run something like a XDD (and/or iozone for a complete gamut of tests)?

[dicecca112]

post a link to a program you want, and I'll run it this weekend.

[Xenogias]

Has anyone done any power consumption measurements? For those of you who put it in a notebook, did you noticed increased/decreased battery life?
Barring that, does OCZ have any consumption numbers they can share with us?