Raptor150 and Laptop

[musiclover]

I have a new Dell Inspiron 9400 laptop with T2500 (2x2000Ghz) core duo which I want to use for rather serious music recording: ideally up to 16 simultaneous channels of 24bit/96khz audio + effects.

I want to "replace" the exiting 2.5" 5400 fujitsu with a raptor 150, by using the existing SATA1 (I wish there was a laptop with an external SATA port!) and connecting SATA cable to the Raptor in an external SATA enclosure.

Anybody has any thoughts whether this would work? I mean, would you expect any bottelnecks that would hinder the "normal" transfer rate you'd expect from a raptor connected to a reqular mobo? Also, would an external sata/RAID0 with two raptors connected to a single SATA port speed up the write transfer rate?

Any enlightening thoughts appreciated.

musiclover.

 

[musiclover]

Sorry, not possible.
The best 2.5" SATA drive availiable are the 7200RPM ones from Hitachi (160GB I think).

Raptor(any version) are 3.5"

Thanks, wusy, but note, I am not trying to squeeeeze the 3.5" raptor into the 2.5" hd compartment. the raptor will be in an external enclosure. I will remove the existing 2.5" just to get access to the SATA port and connect the SATA cable of external enclosure to it.

 

[moparman390]

Sounds possible, just not very portable. If you have the SATA cable run right from the Raptor into the SATA port for the standard laptop HDD, it should work as long as you can power it (I'm not quite sure how that will work out coming of the laptop, but if the laptop uses molex (4 pin) or SATA power connectors, that should work as long as the laptop power supply can handle it). You may need an extension power cable, or to power it with an external enclousre (for power only). I'm not sure, but I would think this is possible with a DIRECT CONNECTION via SATA from the laptop to the HDD.

My two cents: I think it's more trouble then what it's worth.

 

[musiclover]

Thanks, everyone. I think I will go for it but I'm not sure about the RAID0. I wil post the results when I'm done.

I'm pretty sure it will work. I'm not sure only about any laptop-related bottlenecks. Specs say it is SATA but Dell is insultingly short on details http://www1.ca.dell.com/content/products/productdetails.aspx/inspn_9400?c=ca&cs=CADHS1&l=en&s=dhs&~section=specs#tabtop

I pulled out the 100G Fujitsu HD the laptop came with, http://www.fujitsu.com/us/fcpa/news/pr/20050510-02.html
and there is a standard SATA port. I will use external enclosure with its own power so there is not going to be any issue with Raptor(s) being underpowered.

Portability is not an issue either. I'm not going to get rid of the Fujitsu andl reconnect it when I'm not doing any recordings.

I'm still wondering, would RAID0 with 2hd's connected to a single SATA increase performance over a single HD?

Any "portable" RAID external cases for only 2HD's? I'd prefer it's made of aluminum because of heat issue.

m.

 

[nobly]

So you're trying to put a RAID 0 through your laptop SATA port (internal) that will go outside to a enclosure.

Interesting. Not mobile really, but I think it should work. I mean its only that your HDD is external now. Kind of like a swappable HDD. eSATA would be perfect for you, but as wusy said, I'm not sure if laptops have them yet.

But doing a RAID 0, you need a RAID controller. Not sure if your laptop has one or not. RAID 0 would increase performance over a single SATA port. In this case, your laptop probably only has 1 channel of SATA. But since the SATA 1 spec calls for 150MB/s of theoretical bandwidth, you'd be fine with 2 HDDs in RAID 0. This is commonly done in motherboard chipsets as well. I think NVRAID has 2 SATA channels, but 4 ports, so in essence 2 ports share 1 channel of SATA.

 

[musiclover]

Thanks nobly. That's exactly what I'm trying to do. I don't really care about portability that much. You right, its lika a swappable HDD. I think there is a laptop with RAID capability and 2 2.5' internal HD's i saw somewhere, but I would prefer an outboard RAID controller as I want to preserve the horspower of the CPU's for processing at the same time.

For recordings I have to look at real sustained write transfer rate. Tomshardware test of a single Raptor 150 suggests you can count on 60Mb+ sustained transfer rate. So I'm trying to find out whether I can really get a meaningful write increase using 2Raptors RAID0 over one SATA channel given SATA 150Mb/sec theoretical transfer rate.

BTW, does the NCRAID with 4 ports over 2 SATA channels compromise performance over 4 ports over 4 SATA channels? In practical, sustaned write transfer rate, I mean.

 

[nobly]

BTW, does the NCRAID with 4 ports over 2 SATA channels compromise performance over 4 ports over 4 SATA channels? In practical, sustaned write transfer rate, I mean.

Sorry, I can only answer you theoretically on this one.
No, I don't believe so. Basically what NVRAID does there is it has 2 port multipliers (so that 2 ports can share 1 channel).
So theoretically, HDD's only spit out 60-80MB/s sustained. That 80MB/s comes from the raptors, not sure how valid that is, but I think its a max that some people have managed to hit.

So yeah, for SATA 1, 150MB/s, its sharing 2 drives which can do 120-160MB/s combined for RAID 0. So it shouldn't be a problem. That 160MB/s is over the limit, but really, at that point I think you'd rarely hit that 160MB/s.

So yeah, I don't think 4 ports 2 channels is any different than 4 ports 4 channels. Although you should beware of other background bottlenecks - say if the controller's a PCI card, its limited to 133MB/s by the PCI bus.

On the flip side, if you are talking 4 ports 1 channel, you'd start hitting some issues with 3-4 SATA drives in RAID 0.

Side note: that last part is probably why SATA2 with 300MB/s theoretical is here.

 

[musiclover]

Thanks nobly,

I've just found a site which answers some of my questions:
http://www.gamepc.com/labs/print_content.asp?id=raptor150raid&cookie%5Ftest=1

It looks like RAID0 would make a difference: recording (=sustained transfer rate) would benefit from two Raptors RAID0. According to the last two graphs, the difference between a single Raptor and two Raptors in RAID0 is >50%, while difference between 2 and 3 Raptors is less than 9%. I think the 9% difference is meaningless considering the additional risk of failure when 3rd disk is added, before even considerin the cost of adding the 3rd Raptor.

If recording many channels simultaneously is 'multithreaded' transfer, than it looks I must have a RAID0, according to the data on the website, and should not worry about the 150Mb/sec ceiling in single SATA port transfer rate. Any thoughts about relying on Gamepc tests?

I hope the rest of the guts of my Dell Inspiron 9400 will handle the recording load.

Any recommendation as to a decent RAID external box with a controller you know about?

 

[musiclover]

I just checked the Nodesoft’s DiskBench benchmark, and it appears that it measures disk performance in a right way. It claims it is not a synthetic benchmark, and that it tests

“How fast are my disks really. In a real life situation. Not in a synthetic benchmarking program that will display figures that I will never be able to meet. The program is using the current filesystem to save a file. If it is fragmented, the performance will be degraded, as it is in real life.” http://nodesoft.com/DiskBench/

The guys at http://www.gamepc.com/labs/print_content.asp?id=raptor150raid&cookie%5Ftest=1 say that they run Nodesoft disk bench:
“we ran a set of disk write speed tests on these drives, testing how quickly the drives could write 1 GB and 2 GB file sizes to the disk in 100 MB chunks”. This is pretty much my ‘real life’ situation.