falcon review

alcoholbob

Diamond Member
May 24, 2005
6,365
433
126
They reviewed the Vertex a month ago and just used the same numbers. The Vertex was probably run on the 1275 firmware and the Falcon on 1370. I never bothered to go above and beyond standard benchmarking so I can't verify this, but the OCZ boards have some guys who compared 1275 and 1370, and it seems like 1370 has more consistent performance across a wide range of sector offsets.

Anyway, those ATTO write numbers are impossible. There's no way a Vertex can hit 260MB/s on writes (and writes are never higher than reads on the Vertex). The Vertex EX and the Intel X25-E max out at 210MB/s writes. At best with the Indilinix MLC controller you might see ~180MB/s on very large file blocks, and typically < 160 MB/s at 512KB and below.

The ATTO read numbers are believable since it's about what I've seen myself and on OCZ forums.
 

taltamir

Lifer
Mar 21, 2004
13,576
6
76
yes, having only ONE review and it being made in such an unprofessional manner is not good enough. Thanks for pointing out some specifics as to why they could have gotten the numbers they did.
 

Idontcare

Elite Member
Oct 10, 1999
21,110
59
91
Originally posted by: Astrallite
They reviewed the Vertex a month ago and just used the same numbers. The Vertex was probably run on the 1275 firmware and the Falcon on 1370. I never bothered to go above and beyond standard benchmarking so I can't verify this, but the OCZ boards have some guys who compared 1275 and 1370, and it seems like 1370 has more consistent performance across a wide range of sector offsets.

Anyway, those ATTO write numbers are impossible. There's no way a Vertex can hit 260MB/s on writes (and writes are never higher than reads on the Vertex). The Vertex EX and the Intel X25-E max out at 210MB/s writes. At best with the Indilinix MLC controller you might see ~180MB/s on very large file blocks, and typically < 160 MB/s at 512KB and below.

The ATTO read numbers are believable since it's about what I've seen myself and on OCZ forums.

When writes get borked I always suspect there is some active caching involved in the loop...not bothering to read the review myself yet but by any chance are they using a sata controller that has cache or does the falcon itself have more buffer cache than the vertex?
 

geoffry

Senior member
Sep 3, 2007
599
0
76
Originally posted by: Idontcare
Originally posted by: Astrallite
They reviewed the Vertex a month ago and just used the same numbers. The Vertex was probably run on the 1275 firmware and the Falcon on 1370. I never bothered to go above and beyond standard benchmarking so I can't verify this, but the OCZ boards have some guys who compared 1275 and 1370, and it seems like 1370 has more consistent performance across a wide range of sector offsets.

Anyway, those ATTO write numbers are impossible. There's no way a Vertex can hit 260MB/s on writes (and writes are never higher than reads on the Vertex). The Vertex EX and the Intel X25-E max out at 210MB/s writes. At best with the Indilinix MLC controller you might see ~180MB/s on very large file blocks, and typically < 160 MB/s at 512KB and below.

The ATTO read numbers are believable since it's about what I've seen myself and on OCZ forums.

When writes get borked I always suspect there is some active caching involved in the loop...not bothering to read the review myself yet but by any chance are they using a sata controller that has cache or does the falcon itself have more buffer cache than the vertex?

From the specs I've seen both drives have EXACTLY the same cache, 64 MB.

I've been looking at the falcons the past week or so. It looks like an exact duplicate of the Vertex but with a smidge more storage room. (perhaps that will make "used" drives slower as the vertex keeps more flash forwrite balancing or whatever its called)
 

taltamir

Lifer
Mar 21, 2004
13,576
6
76
Originally posted by: geoffry

You would need to frequently wipe it using software from the manufacturer.

Although from what I've read the TRIM command, which will be in windows 7, will slow the performance degradation as an SSD is used.

not slow, completely eliminate. there is no "performance degradation", there is internal fragmentation that can happen up to a point. It can't continue dropping forever.

The real solution would have been to design flash chips that can be erased in blocks smaller than 512kb... ideall in 4kb blocks (same as what it can write)