88 Comments

• Digg
• Bury

In a single drive? No.

• Reply

• Digg
• Bury

931.322575 gigabytes, actually.

• Reply

• Digg
• Bury

@phantomAI clueless windoze d3fr4gr

• Reply

• Digg
• Bury

Hitachi did the same thing, and it's covered by the same site, just fyi:<a class="user" href="http://www.dailytech.com/article.aspx?newsid=5590&quot;&gt;http://www.dailytech.com/article.aspx?newsid=5590&lt;/a&gt;

• Reply

• Digg
• Bury

@ adml_shakewhat the hell is a bonner?

• Reply

• Digg
• Bury

I should also mention that not all new Maxtors suck. I'm very happy with my MaxLine IIIs.

• Reply

• Digg
• Bury

To be fair:10K RPM Drive vs. 4-7200 RPM drives (Raid0 / Stripe)The possible write speed is 4x the possible write speed of a single drive with COMPARE ABLE controller cards.Generally speaking, the poor performance (yes, I said poor) of the RAID0 can be attributed to a software raid0. You should be seeing about 1200MB/s out of the 4 drives (300MB/s x 4). On a software RAID0, the computer can only send a read or write command to 1 drive at a time, thus the 300MB/s peek. The drives are able to keep up even with the slower seek times.The entire purpose of RAID0 is for additional read/write bandwidth. This assumes you actually have enough data to write a single block of data greater than the largest block available with SATA or SAS (generally not necessary for home users). Games and Databases are the largest data hogs right now, followed closely by graphics applications with very large raw images.Raid5 vs Raid0To be true to the performance measures, on a good controller card the only difference should be from 1 of three problems:(1) Raid5 is compared against the same drive number as Raid0. - This is an incorrect assumption because you need the same data write width on both setups to make a fair comparison - Raid5 with 5 750s ~ same performance as Raid0 with 4 750s(2) Raid5 with a poor controller (e.g., On-board Raid5) - The parity is generally calculated by your CPU, causing the performance problem. - Most Mother Board do not have the CACHE required to send a read or write command to 2 drives at once, let alone 4 drives. I'm not even sure if the SATA command set (eh, low level programming language) includes a way to do this. It is generally a function reserved for SCSI and SAS systems.(3) Raid5 with more writes than reads. - This gets into hardware theory a bit - The closest comparison I can think of is when you fill the cache of the controller (512MB is not unusual) with write transactions and then you want to read from the drives. - This is about the same as trying to send a big file through a modem and requesting another web page... you've filled the transmit buffer to capacity and your request for another web page must wait for available buffer memory before it can be sent.I've also noticed a randomly poor performance from Raid5 in SATA (7200 RPM) and SAS (10K RPM) environments, but the problem is random and doesn't occur very often. I attribute this to the "age" of the Serial technology, it's simply not as mature as SCSI.

• Reply

• Digg
• Bury

I have encountered with it. It is really good one. And, one thing I would say that I have bought only in £94.00. If someone is interested then visit this below mentioned link:<a class="user" href="http://www.dclstore.co.uk/products.asp?prod=1300&quot;&gt;http://www.dclstore.co.uk/products.asp?prod=1300&lt;/a&gt;

• Reply