SSD Vs Velociraptor

[Denithor]

Originally posted by: tjcinnamon
Otherwise wait for the 80GB ioXtreme (if it exists) 500MB/s to 700MB/s read and 600MB write. Just under $1000 release Q1.

Got any benchmarks to back up those numbers?

And what kind of system connection are those going to use, SATA3 is limited to 375MB/s transfers...

 

[IntelUser2000]

Got any benchmarks to back up those numbers? And what kind of system connection are those going to use, SATA3 is limited to 375MB/s transfers...

Huh?? SATA2-300 can do 300MB/s. SATA3-600 will do 600MB/s. It won't matter anyway because the FusionIO drive runs on the PCI Express x4 interface.

And in the SATA world, SATA 3Gb/s means 300MB/s.

 

[Golgatha]

Originally posted by: gizbug
So, does the Velociraptor really beef up your system when used as a boot drive with windows?

My system seems snappier since I changed from RAID 0 750GB drives to a single Velociraptor. Sustained transfer went from 114 MB/sec on the RAID 0 to 120 MB/sec with the single Velociraptor 300GB drive. Backup and recovery is significantly easier and more reliable as well.

 

[QuantumPion]

Originally posted by: taltamir

Originally posted by: Blain
MTBF is a tool used for estimating failure rate.
* It doesn't mean that every drive in the product line will fail sooner than the estimate.
* It doesn't mean that every drive in the product line will last longer than the estimate.
It's simply a measurement that all HD manufacturers spend time and resources on calculating.

actually it takes REPAIR into account... item fails, is repair, and fails again, and repaired again, the mean of the "time" it stays "working" between each failure is called the MTBF (mean time between failures).

The issue is that it is on guesses and numbers they are literally pulling out of their ass. They have obviously not tested a hard drive for thousands of years, to get 153 years of operation on average between each repair.

What it DOES mean that EVERY SINGLE DRIVE from that line WILL fail LONG BEFORE the allotted time. Because the allotted time is pure bullshit.

You are totally wrong and don't know what you are talking about.

MTBF is a statistical measure of failure rate. a MTBF of 1.5e6 hours means that in a sample size of 1000 drives, on average 6 will fail per year (1 / 1.5e6 hours = 0.006 year^-1).

 

[Denithor]

Originally posted by: IntelUser2000

Got any benchmarks to back up those numbers? And what kind of system connection are those going to use, SATA3 is limited to 375MB/s transfers...

Huh?? SATA2-300 can do 300MB/s. SATA3-600 will do 600MB/s. It won't matter anyway because the FusionIO drive runs on the PCI Express x4 interface.

And in the SATA world, SATA 3Gb/s means 300MB/s.

Wrong.

Second gen SATA
3Gbps = (3,000,000,000 bit/sec)/(8 bit/byte) = (375,000,000 byte/sec)/(1,000,000 byte/MB) = 375 MB/s

First gen SATA
1.5Gbps = 187.5 MB/s

So I was incorrect in the generation number (called it SATA3 vs SATA2). But at least my math is right. And the next generation will do 6Gbps or 750 MB/s.

 

[Golgatha]

Originally posted by: Denithor

Originally posted by: IntelUser2000

Got any benchmarks to back up those numbers? And what kind of system connection are those going to use, SATA3 is limited to 375MB/s transfers...

Huh?? SATA2-300 can do 300MB/s. SATA3-600 will do 600MB/s. It won't matter anyway because the FusionIO drive runs on the PCI Express x4 interface.

And in the SATA world, SATA 3Gb/s means 300MB/s.

Wrong.

Second gen SATA
3Gbps = (3,000,000,000 bit/sec)/(8 bit/byte) = (375,000,000 byte/sec)/(1,000,000 byte/MB) = 375 MB/s

First gen SATA
1.5Gbps = 187.5 MB/s

So I was incorrect in the generation number (called it SATA3 vs SATA2). But at least my math is right. And the next generation will do 6Gbps or 750 MB/s.

Actually you're wrong.

2 bits are used for error correction and each byte is 8 bits, so you're using 10 bits for every byte of transferred data. Dividing by a factor of 10 bits per byte of actual data gives the 300 MB/sec figure for the theoretical maximum data throughput of SATA2.

http://en.wikipedia.org/wiki/8...gies_that_use_8b.2F10b

 

[Shmee]

Originally posted by: Golgatha

Originally posted by: Denithor

Originally posted by: IntelUser2000

Got any benchmarks to back up those numbers? And what kind of system connection are those going to use, SATA3 is limited to 375MB/s transfers...

Huh?? SATA2-300 can do 300MB/s. SATA3-600 will do 600MB/s. It won't matter anyway because the FusionIO drive runs on the PCI Express x4 interface.

And in the SATA world, SATA 3Gb/s means 300MB/s.

Wrong.

Second gen SATA
3Gbps = (3,000,000,000 bit/sec)/(8 bit/byte) = (375,000,000 byte/sec)/(1,000,000 byte/MB) = 375 MB/s

First gen SATA
1.5Gbps = 187.5 MB/s

So I was incorrect in the generation number (called it SATA3 vs SATA2). But at least my math is right. And the next generation will do 6Gbps or 750 MB/s.

Actually you're wrong.

2 bits are used for error correction and each byte is 8 bits, so you're using 10 bits for every byte of transferred data. Dividing by a factor of 10 bits per byte of actual data gives the 300 MB/sec figure for the theoretical maximum data throughput of SATA2.

http://en.wikipedia.org/wiki/8...gies_that_use_8b.2F10b

I do believe you(golgatha) are right. That is why Sata II is sometimes advertised as 3.0 Gb/s, sometimes 300 MB/s, but never 375 MB/s

 

[IntelUser2000]

Wrong. Second gen SATA 3Gbps = (3,000,000,000 bit/sec)/(8 bit/byte) = (375,000,000 byte/sec)/(1,000,000 byte/MB) = 375 MB/s First gen SATA 1.5Gbps = 187.5 MB/s So I was incorrect in the generation number (called it SATA3 vs SATA2). But at least my math is right. And the next generation will do 6Gbps or 750 MB/s

One would expect that since I phrased "SATA 3Gb/s=SATA 300MB/s" I would know what's going on. But rather than actually looking up to see what's up you just say I'm wrong.

6Gbps/10=600MB/s I think my math is actually correct...

 

[VulcanX]

Gees u dudes all maths majors or summin? wow im confused right now bout the speeds u calculated... and if u want MB/s speeds just take ur 1000Mb/s for instance and divide it by 8 and that will be ur speed in MB/s (seeing how theres 8 bits in a byte), and i dunno anything bout the error correction (thats completely new to me). *shock*

 

[taltamir]

Originally posted by: QuantumPion

Originally posted by: taltamir

Originally posted by: Blain
MTBF is a tool used for estimating failure rate.
* It doesn't mean that every drive in the product line will fail sooner than the estimate.
* It doesn't mean that every drive in the product line will last longer than the estimate.
It's simply a measurement that all HD manufacturers spend time and resources on calculating.

actually it takes REPAIR into account... item fails, is repair, and fails again, and repaired again, the mean of the "time" it stays "working" between each failure is called the MTBF (mean time between failures).

The issue is that it is on guesses and numbers they are literally pulling out of their ass. They have obviously not tested a hard drive for thousands of years, to get 153 years of operation on average between each repair.

What it DOES mean that EVERY SINGLE DRIVE from that line WILL fail LONG BEFORE the allotted time. Because the allotted time is pure bullshit.

You are totally wrong and don't know what you are talking about.

MTBF is a statistical measure of failure rate. a MTBF of 1.5e6 hours means that in a sample size of 1000 drives, on average 6 will fail per year (1 / 1.5e6 hours = 0.006 year^-1).

1. if you don't take repair into account, then it is named differently.
2. And using 1000 drives for a year is still pulling numbers out of their ass.
3. Notice that almost all drives advertise the exact same MTBF, that is impossible, reliability differs between models, but when you are making up numbers they match.
http://en.wikipedia.org/wiki/MTBF

 

[Golgatha]

Originally posted by: VulcanX
Gees u dudes all maths majors or summin? wow im confused right now bout the speeds u calculated... and if u want MB/s speeds just take ur 1000Mb/s for instance and divide it by 8 and that will be ur speed in MB/s (seeing how theres 8 bits in a byte), and i dunno anything bout the error correction (thats completely new to me). *shock*

Yeah, it's different than figuring speeds from your ISP, USB 2.0, Ethernet etc. due to 10bits transfer needed for 8bits of data throughput on a SATA bus. Confused the hell out of me until I researched it further because I saw every HDD manufacturer stated the speed as 300 MB/sec transfer rate for SATA2.

3.0 Gbps = 3,000,000 bits per second (bps) = 3000 Mbps

3000 Mbps / (10 Mb/1 MB of information transferred) = 300 MB/sec of actual data transferred per second.

It's somewhat a matter of semantics, because you actually are transferring 375 MB/sec worth of 1s and 0s over the SATA2 bus. But the actual information throughput (i.e. useful data you'll actually record onto the HDD) has a theoretical 300 MB/sec transfer speed due to the 2bits "wasted" for each transfer to ensure data integrity.

 

[taltamir]

I wouldn't call dividing by 8 or dividing by 10 being a math major. This is preschool algebra.