X25m 80GB READ is slow

[idata]

Hi,

I´ve a problem with my new SSD - the drive is up to date with the new 02HD Firmware, but my Read is too slow. Some hints for me.

Here is my system http://h10025.www1.hp.com/ewfrf/wc/document?docname=c01635688&cc=us&lc=en&dlc=en

thanks in advance

AS SSD Benchmark 1.4.3645.3568

------------------------------Name: INTEL SS DSA2M080G2GC SATA Disk DeviceFirmware: 2CV1Controller: amdsataOffset: 1024 K - OKGröße: 74,53 GBDatum: 28.01.2010 21:08:57------------------------------Sequentiell:------------------------------Lesen: 154,30 MB/sSchreiben: 78,52 MB/s------------------------------4K:------------------------------Lesen: 6,80 MB/sSchreiben: 8,21 MB/s------------------------------4K-64Threads:------------------------------Lesen: 78,31 MB/sSchreiben: 48,47 MB/s------------------------------Zugriffszeiten:------------------------------Lesen: 0,289 msSchreiben: 0,443 ms------------------------------Score:------------------------------Lesen: 101Schreiben: 65Gesamt: 215------------------------------

[mmokk]

parsec wrote:

It also seems to me that darkgreen is not the type to complain about something being broken while sitting on his hands. Which is why I've given credence to his story.

nobody was disupting that he had a problem. a solution was offered, but he was unable to make it work.

Your points regarding usage are good of course, although it might be argued that an HDD would not react in the same manner given the same usage history.

an ssd is not a hard drive.

[mmokk]

parsec wrote:

Certainly later the G1's ought to have been updated to use TRIM, but alas, that has not happened. Why that is remains a mystery.

probably because the g1 controller is incapable of supporting that functionality.

but an intel conspiracy to force people to purchase g2's to get trim makes for better headlines.

[idata]

No use crying over spilt milk, but TRIM was part of a development process long before the public release of Win 7 and it was built in from the very first beta version.

The G2's also came out before the official release of Win 7, yet TRIM support was provided sometime afterwards.

Indilinx SSD's that came out well before Win 7 managed to provide TRIM support with a firmware update.

If it was a hardware limitation, Intel could have said so from the start to squash conspiracy theories, but they didn't.

http://www.anandtech.com/show/2806 http://www.anandtech.com/show/2806

"The disappointing part of the announcement is that there's no TRIM support for the first gen 50nm drives. As far as I can tell, this isn't a technical limitation of the drives, but rather something Intel is choosing to enable only on the 34nm products".

Then there is the Kinston TRIM issue, which if you can believe the following would suggest that Intel placed a monetary value on TRIM.

http://www.anandtech.com/show/2968/intel-s-x25-v-kingston-s-30gb-ssdnow-v-series-battle-of-the-125-s... http://www.anandtech.com/show/2968/intel-s-x25-v-kingston-s-30gb-ssdnow-v-series-battle-of-the-125-s...

"The story goes like this: Intel let Kingston build a value version of its X25-M G2 drives. This became the 40GB Kingston V Series Boot Drive. When Intel added TRIM support to the G2 drives, Kingston was told to wait. Kingston would get TRIM support when Intel launched its own 40GB version of the G2. Here's where things get hairy. Intel and Kingston couldn't work out terms for the TRIM enabled firmware on the V Series Boot Drives. Intel wanted too much money and Kingston wanted to keep the drive price below $100. The outcome? Kingston V Series Boot Drive owners never got official TRIM support and the product was dropped altogether."

[idata]

"an ssd is not a hard drive."

It is impossible to respond without sounding sarcastic...

[idata]

Thanks for reminding me about the Flash business. Sounds like the SSD business got a little down and dirty.

But how could any manufacture be the OEM for a product and have their customer sell the identical product for less, particularly when the OEM is public knowledge? I'm leaving this one alone.

I would tend to believe AnandTech. And he's been accused of being an Intel shill.

How the G1 controller (does that mean just the chip, or chip and software? In this case, to me it means the chip alone) cannot be programmed to process TRIM does not make sense, but I suppose it's possible.

The SSD story has always been weird, the need for TRIM was recognized long ago, and the TRIM spec was written into the ATA specifications. There is much going on behind the scenes here, maybe we'll hear about it some day.

IMO, TRIM is just the tip of the iceberg of OS optimization for SSDs. We may be lucky that the NAND peg fits in the NTFS hole. Although an SSD's NAND layout can be viewed logically as sectors and cylinders, that is likely not optimal.