Subject: Storage | August 4, 2009 - 12:36 PM | Allyn Malventano
Fusion-io passed along a press release with some of their internal testing results of the ioXtreme. In the below release, they claim a 6x performance gain under Maya as compared to a high end 4-drive RAID. Details are scarce at present and we have yet to see hard benchmark figures from the ioXtreme, but here is the tease they provided on the published spec sheet:
Subject: Storage | August 3, 2009 - 02:21 PM | Jeremy Hellstrom
You may think of SSDs as a very new technology for mainstream users; something that is only true compared to platter technology. Intel has already hit their second generation and one of the biggest improvements was the price; not to take anything away from its performance numbers. Now it is OCZ's turn, with the third generation of their Vertex series. The newest is a 120GB Vertex Turbo drive which
Subject: Storage | July 30, 2009 - 03:32 PM | Jeremy Hellstrom
PhotoFast may not sound like a big name in storage, but believe it or not, their G-Monster V5 256GB is one of the fastest SSDs on the planet, though also one of the most expensive, at $1000. Tweaktown tested it out on an 8 core Opteron system and found that the system its self was a bottleneck for the drives performance. Even so, the G-Monster V5 has hitting almost 250 MB/s read speeds and 210MB/s write speeds.
Subject: Storage | July 28, 2009 - 12:32 PM | Jeremy Hellstrom
As Allyn has proved, the new technology used in the second generation has almost completely removed the fragmentation over time problem. You can see that confirmed by The Tech Report's review of the drive, where they tested the response times of drives before and after heavy testing. They still anxiously await Win7 and the TRIM commands that will help Indilinx and Samsung controlled drives performance; the question is can it help them cl
Subject: Storage | July 27, 2009 - 12:11 PM | Allyn Malventano
Much of the below press release is industry jargon, but what you can take home from this is Fusion-io paired an ioDrive with a Dell PowerEdge and MS SQL Server to produce a solid, high performing product that puts out some impressive stats. This should make companies want to adopt the tech, which leads to volume sales and ultimately drives down the cost of consumer-grade products (in this case PCI-e SSDs). The potential end result? Cheaper ioXtreme SSDs for the rest of us.
Subject: Storage | July 24, 2009 - 06:36 PM | Allyn Malventano
***Public Service Announcement***
Intel has informed me of a bug in the X25-M G2 shipping firmware (02G2). This
bug is specific to setting a hard drive password in your BIOS. If you
have set a BIOS HDD password, then subsequently change or disable that
password, the SSD may become inoperable. Intel is correcting the issue
in firmware and will be posting a fix shortly. In summary:
Subject: Storage | July 23, 2009 - 01:14 PM | Jeremy Hellstrom
If you watched the live Podcast last night you got a sneak preview of the results of Allyn's testing of the 2nd generation Intel X25-M. Today the hard numbers have arrived in his full review. Write speeds remain the drives Achilles Heel, the Indilinx controlled drives still end up on top though this new drive from Intel now has less than a 0.1ms Random Access time. It is better than its predecessor across the board, and cheaper to boot!
Subject: Storage | July 22, 2009 - 09:15 PM | Allyn Malventano
You may remember us talking about the upcoming OCZ Colossus SSD - the first standard 3.5-in form factor solid state drive first shown at Computex 2009. As it happens, some preliminary performance results happened to swing by our email this morning.
We received a review sample of Intel's next generation Solid State Drive just this afternoon. Allyn promptly put the drive through its paces and has posted a review for your viewing pleasure. I think you'll find the speed increases, though modest, coupled with the price decreases, will make this drive more than worth your time.
Subject: Storage | July 22, 2009 - 04:11 PM | Allyn Malventano
Our Intel X25-M G2 sample just arrived. I promptly threw it on the storage testbed and got some initial data. I'll let the pics do the talking: