Performance for 64 bit hardware?
- dahnielson
- Moderator
- Posts: 632
- Joined: Wed Jan 23, 2008 11:25 pm
- Location: Linköping / Tranås, Sweden
- Contact:
Re: performance - 64 bit hardware
According to the specification:
OCZ Core Series SATA II 2.5" SSD
Available in 30GB, 60GB, 120GB capacities
Read: up to 143 MB/sec
Write: up to 93 MB/sec
Seek: <0.35ms
OCZ Core Series V2 SATA II 2.5" SSD
Available in 30GB, 60GB, 120GB, 250GB capacities
Read: up to 170 MB/sec
Write: up to 98 MB/sec
Seek: <0.2-0.3ms
It would obviously be interesting to see actual benchmarks. But the spec looks promising compared to the spec of my current sample drive:
Seagate SATA II Barracuda 7200.11 (32MB Cache)
500 GB capacity
Speedy performance at 105Mb/s sustained data rate
Average latency 4.16 msec
Random read seek time <8.5 msec
Random write seek time <9.5 msec
I just checked the prices on OCZ Core and they are not that expensive:
OCZ Core Series SATA II 2.5" SSD
30GB 1795 SEK
60GB 2495 SEK
120GB 4495 SEK
OCZ Core Series V2 SATA II 2.5" SSD
30GB 2195 SEK
60GB 2995 SEK
120GB 5295 SEK
The 60GB drive appears to be the current sweet spot. But I mean, even 30 GB can hold a pretty decent library for playback without buffering (assuming the <0.35 msec seek time is quick enough). After all my hardware sampler only got 32 MB and building a computer with 32 GB RAM is a lot more expensive than that.
What kind of seek time performance is required for unbuffered playback?
OCZ Core Series SATA II 2.5" SSD
Available in 30GB, 60GB, 120GB capacities
Read: up to 143 MB/sec
Write: up to 93 MB/sec
Seek: <0.35ms
OCZ Core Series V2 SATA II 2.5" SSD
Available in 30GB, 60GB, 120GB, 250GB capacities
Read: up to 170 MB/sec
Write: up to 98 MB/sec
Seek: <0.2-0.3ms
It would obviously be interesting to see actual benchmarks. But the spec looks promising compared to the spec of my current sample drive:
Seagate SATA II Barracuda 7200.11 (32MB Cache)
500 GB capacity
Speedy performance at 105Mb/s sustained data rate
Average latency 4.16 msec
Random read seek time <8.5 msec
Random write seek time <9.5 msec
I just checked the prices on OCZ Core and they are not that expensive:
OCZ Core Series SATA II 2.5" SSD
30GB 1795 SEK
60GB 2495 SEK
120GB 4495 SEK
OCZ Core Series V2 SATA II 2.5" SSD
30GB 2195 SEK
60GB 2995 SEK
120GB 5295 SEK
The 60GB drive appears to be the current sweet spot. But I mean, even 30 GB can hold a pretty decent library for playback without buffering (assuming the <0.35 msec seek time is quick enough). After all my hardware sampler only got 32 MB and building a computer with 32 GB RAM is a lot more expensive than that.
What kind of seek time performance is required for unbuffered playback?
Anders Dahnielson
Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux
Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux
- dahnielson
- Moderator
- Posts: 632
- Joined: Wed Jan 23, 2008 11:25 pm
- Location: Linköping / Tranås, Sweden
- Contact:
Re: Performance for 64 bit hardware?
Anders Dahnielson
Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux
Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux
Re: Performance for 64 bit hardware?
Hi Benno,
Well, to be honest: this will be my first compile under Linux...
It will take some exercise.. and time ..
But I will try, and hope to come back on it!
Cheers, Eddy
Well, to be honest: this will be my first compile under Linux...
It will take some exercise.. and time ..
But I will try, and hope to come back on it!
Cheers, Eddy
Re: Performance for 64 bit hardware?
I have 2 500GB SATA2 drives. Would it be advisable to say put my Strings and Percussions on one and Brass and Woodwinds on the other?
What are you doing Alex?
Thanks
What are you doing Alex?
Thanks
- dahnielson
- Moderator
- Posts: 632
- Joined: Wed Jan 23, 2008 11:25 pm
- Location: Linköping / Tranås, Sweden
- Contact:
Re: Performance for 64 bit hardware?
I'm not Alex, but I would say yes. Dividing instruments between hard drives is a good idea to get good performance when using a lot of instruments.ccherrett wrote:I have 2 500GB SATA2 drives. Would it be advisable to say put my Strings and Percussions on one and Brass and Woodwinds on the other?
Anders Dahnielson
Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux
Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux
Re: Performance for 64 bit hardware?
OK then I think I will divide my two drives into Strings and Percussion on one drive and Woodwinds and Brass on the other.
Thanks for the help.
P.S. So would solid state drives solve all these throughput concerns?
Thanks for the help.
P.S. So would solid state drives solve all these throughput concerns?
- dahnielson
- Moderator
- Posts: 632
- Joined: Wed Jan 23, 2008 11:25 pm
- Location: Linköping / Tranås, Sweden
- Contact:
Re: Performance for 64 bit hardware?
Yes. But it's not throughput that's the problem. It is the random seek time.ccherrett wrote:P.S. So would solid state drives solve all these throughput concerns?
A rule of thumb is that hard drives are only fast if you read and write to them sequentially, that's why video is stored in one big hunky file as a stream and not as individual frames even if no inter-frame compression is used. For random access it's better to use RAM. Solid states drives takes storage closer to the performance of RAM.
The way I understands it: This is important when streaming samples for the purpose of polyphony. Because you need to read a block of samples ("frames") here and another block of samples there to stream more than just one instrument sample. The jumping between those blocks for different instruments/samples require seeking. That is why you get better performance of having instruments that should play at the same time on different drives, and also why they should be kept on dedicated drives and not on the system disk (that the system processes will do a lot read/write/seek to).
Anders Dahnielson
Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux
Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux
Re: Performance for 64 bit hardware?
OK cool!
I have two identical 500 GB SATA2 drives that are just for music and no system activity happens on. So I will try it out tonight.
Again Thanks!
I have two identical 500 GB SATA2 drives that are just for music and no system activity happens on. So I will try it out tonight.
Again Thanks!
Re: Performance for 64 bit hardware?
Chris, i have two drives, split as follows:
Drive 1.
1st Violins
Cellos
Kontrabasses
Brass (including FH)
Male choir
Organs
Harp
Drive 2.
2nd Violins
Violas
Woodwinds
Tympani
Percussion
Piano
Female and mixed choirs.
I experimented with usage, according to my style of writing, and this provided a reasonable balance.
Worth experimenting to find out rough percentages of 'where you write the most'.
And my apologies to all for being missing in action these last few weeks. I've been hammering away trying to complete an album, and other related activities.
Alex.
Drive 1.
1st Violins
Cellos
Kontrabasses
Brass (including FH)
Male choir
Organs
Harp
Drive 2.
2nd Violins
Violas
Woodwinds
Tympani
Percussion
Piano
Female and mixed choirs.
I experimented with usage, according to my style of writing, and this provided a reasonable balance.
Worth experimenting to find out rough percentages of 'where you write the most'.
And my apologies to all for being missing in action these last few weeks. I've been hammering away trying to complete an album, and other related activities.
Alex.
- dahnielson
- Moderator
- Posts: 632
- Joined: Wed Jan 23, 2008 11:25 pm
- Location: Linköping / Tranås, Sweden
- Contact:
Re: Performance for 64 bit hardware?
That's a valid excuse.Alex wrote:And my apologies to all for being missing in action these last few weeks. I've been hammering away trying to complete an album, and other related activities.
I started to write some music. Then I grew frustrated with the programming of my strings and started reprogram them. Then I grew frustrated with the limitations of programming GIGs and started implementing SFZ...
Anders Dahnielson
Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux
Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux