Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Gasp! The all-new and ridiculously fast G.Skill "Falcon" SSD

chowmein said:
http://www.newegg.com/Product/Product.aspx?Item=N82E16820231256

Interesting SSD I just found. Claims a whopping max 230MB/s read and 190MB/s write!!! :eek: Also has a 64MB DRAM cache. Intel better watch out! This SSD's read and write speeds are closing in fast on the X-25E SSDs and it looks like it'll totally smoke the regular X-25's write speeds!
Click to expand...

Again, for most boot disk applications we should be more concerned with small random reads/writes rather than large sequential throughput, which is the most commonly cited spec. My guess is that this drive won't be faster than the Intel drives in real world tests.
 
Thiol said:
Again, for most boot disk applications we should be more concerned with small random reads/writes rather than large sequential throughput, which is the most commonly cited spec. My guess is that this drive won't be faster than the Intel drives in real world tests.
Click to expand...

You never know! No one has it yet, anyways. And, the max write speed is 20MB/s faster than the X-25E.
 
chowmein said:
You never know! No one has it yet, anyways. And, the max write speed is 20MB/s faster than the X-25E.
Click to expand...

Yeah it would be really nice to do some reviews that compare the SSD's to the 500GB WD or Seagate drives.....

Also, do SSD's slow down as they get full just like normal drives? That's something that is often overlooked in real world tests IMHO. People look at fresh drives but there's bound to be a performance drop on a (standard) 320GB drive when it's got 300gb of crap on it, even if it was faster than a 500GB drive initially. Does this same phenomena apply to the SSD's?
 
chowmein said:
You never know! No one has it yet, anyways. And, the max write speed is 20MB/s faster than the X-25E.
Click to expand...

Yes we "do" know. The G.Skill uses the same Indilinx Barefoot controller that OCZ uses in the Vertex. It's NOT going to come close to beating the Intel SSD in random read/write performance.
 
I don't think so, because each memory cell in the SSD can be accessed directly, rather than a write/read head on a normal HDD that would otherwise have to move into position...
 
Mac4Obama said:
Yeah it would be really nice to do some reviews that compare the SSD's to the 500GB WD or Seagate drives.....

Also, do SSD's slow down as they get full just like normal drives? That's something that is often overlooked in real world tests IMHO. People look at fresh drives but there's bound to be a performance drop on a (standard) 320GB drive when it's got 300gb of crap on it, even if it was faster than a 500GB drive initially. Does this same phenomena apply to the SSD's?
Click to expand...

There is some slowdown as you fill the drive up because they all have the same issue "where do I put this data on a drive that's almost filled up" though the issue doesn't appear to be too significant on SSD because you're not dealing with rotational latency that changes depending on if you're writing on the outside edges versus inside.
 
nuckinfutz said:
Yes we "do" know. The G.Skill uses the same Indilinx Barefoot controller that OCZ uses in the Vertex. It's NOT going to come close to beating the Intel SSD in random read/write performance.
Click to expand...

Really? This SSD is new. Now do you know for sure? :rolleyes:
 
chowmein said:
Really? This SSD is new. Now do you know for sure? :rolleyes:
Click to expand...

GSKill don't make their own stuff, they often rebadge other same controllers and memory chips. Just look at their history of of SSDs, they always follow others. It is highly luckily they either have the Indilinx or Samsung controller, in either case we already know the random IOPS, they won't beat intel in that area.

OCZ's upcoming Vertex EX (SLC version of Vertex) will have 1.5-2 times more random IOPS than the regular Vertex according to latest result in their forum.

230/190 MBps doesn't mean anything, Vertex right now can do that kind of speed or more than that. It's the random IOPS that count and right now Intel is the beast in that area but most people don't need that kind of high random IOPS that Intel is pulling.
 
Mac4Obama said:
Yeah it would be really nice to do some reviews that compare the SSD's to the 500GB WD or Seagate drives.....

Also, do SSD's slow down as they get full just like normal drives? That's something that is often overlooked in real world tests IMHO. People look at fresh drives but there's bound to be a performance drop on a (standard) 320GB drive when it's got 300gb of crap on it, even if it was faster than a 500GB drive initially. Does this same phenomena apply to the SSD's?
Click to expand...

There is no HD that can beat those second generation SSDs such as Vertex or Intel, not even super fast 15K RPM drives you see in enterprise. The latency is always going to be less than .05ms for SSDs. HD can't even do less than 8ms. Latency is where the snappiness come from and SSD will always win in this area.

As for slowing down, in READ speed, it will NEVER slow down. Write speed, it will slow down a bit after the whole drive is used, we'll talking like 190MBps>140MBps which is still faster than any hard drive can do. But latency will always remain .01ms for most SSDs.
 
MikhailT said:
GSKill don't make their own stuff, they often rebadge other same controllers and memory chips. Just look at their history of of SSDs, they always follow others. It is highly luckily they either have the Indilinx or Samsung controller, in either case we already know the random IOPS, they won't beat intel in that area.

OCZ's upcoming Vertex EX (SLC version of Vertex) will have 1.5-2 times more random IOPS than the regular Vertex according to latest result in their forum.

230/190 MBps doesn't mean anything, Vertex right now can do that kind of speed or more than that. It's the random IOPS that count and right now Intel is the beast in that area but most people don't need that kind of high random IOPS that Intel is pulling.
Click to expand...

Alright, point taken. But we should at least wait for benchmarks before saying that this is yet another SSD that gets smashed to pieces by the X-25s.
 
Mac4Obama said:
Also, do SSD's slow down as they get full just like normal drives? That's something that is often overlooked in real world tests IMHO. People look at fresh drives but there's bound to be a performance drop on a (standard) 320GB drive when it's got 300gb of crap on it, even if it was faster than a 500GB drive initially. Does this same phenomena apply to the SSD's?
Click to expand...

They do indeed slow down. Well-studied problem since discovery. Not that bad in reality and varies with each SSD. Some benchmarks:

http://www.tomshardware.com/reviews/ssd-performance-power,2279.html
 
chowmein said:
Alright, point taken. But we should at least wait for benchmarks before saying that this is yet another SSD that gets smashed to pieces by the X-25s.
Click to expand...

chowmein you are not incorrect. The G.Skill Falcon or the OCZ Vertex are faster than the Intel SSD. If a person needs fast sequential reads or writes the Indilinx SSD are the way to go.

Intel has optimized the heck out of their SSD for basic computer tasks but it doesn't do as well against standard hard drives in sequential stuff.

So it really depends on how the SSD is going to be used and I'm not ruling out the G.Skill stuff getting faster in the random read/write areas with further optimizations.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back