Return to the CurtPalme.com main site CurtPalme.com Home Theater Forum
A forum with a sense of fun and community for Home Theater enthusiasts!
Products for Sale ] [ FAQ: Hooking it all up ] [ CRT Primer/FAQ ] [ Best/Worst CRT Projectors List ] [ Setup Tips & Manuals ] [ Advanced Procedures ] [ Newsletters ]

 
Forum FAQForum FAQ   SearchSearch   MemberlistMemberlist  Photo AlbumsPhoto Albums  RegisterRegister 
 MembershipClub Membership   ProfileProfile   Private MessagesPrivate Messages   Log inLog in 
Blu-ray disc release list and must-have titles. Buy the latest and best Blu-ray titles to show off in your home theater!

Is SATA III Faster Than SATA II?

 
Post new topic   Reply to topic   Printer-friendly view    CurtPalme.com Forum Index -> Home Theater PCs
View previous topic :: View next topic  
Author Message
Phil Smith




Joined: 08 Mar 2006
Posts: 7717



PostLink    Posted: Sun Aug 11, 2013 4:49 pm    Post subject: Is SATA III Faster Than SATA II? Reply with quote


        Register to remove this ad. It's free!
In real world use, no it's not. Makes you wonder why SATA III even exist:

http://www.tomshardware.com/reviews/ssd-upgrade-sata-3gbps,3469.html
Back to top
ecrabb
Forum Moderator



Joined: 13 Mar 2006
Posts: 15909
Location: Utah

TV/Projector: JVC RS40, Epson 5010


PostLink    Posted: Sun Aug 11, 2013 8:25 pm    Post subject: Reply with quote

Why does it exist? Because some people need more performance than a single drive offers (SSD or no). If you put multiple drives in a RAID SATA enclosure, you can easily saturate the 3Gbps port. The 6Gbps port will give you double the bandwidth.

SC
Back to top
View user's photo album (10 photos)
Phil Smith




Joined: 08 Mar 2006
Posts: 7717



PostLink    Posted: Sun Aug 11, 2013 8:47 pm    Post subject: Reply with quote

Google SSD and RAID. Apparently SSDs in RAID *test* really well, but just like the comparison I linked to, in the real world there's very little if any performance gain.

I actually discovered the lack of performance difference between SATA II and SATA III on my own, while testing the performance of a PCIe SATA III expansion card.

I have an interest in composing film score type symphonic music as a hobby. Streaming huge sample libraries is necessary. I have 3 250gb SDDs. My motherboard's two Intel SATA III ports are suppose to be the best. The two Marvell SATA III ports are considerably slower. I bought a PCIe 2.o SATA III card to see if it would be faster, because of PCIe 2.0's bandwidth. I'll post a copy of a post I made on music forum:

As you can see from the test, the card is still not up to Intel SATA 3 speeds, but it's a huge improvement over the Marvell SATA 3 ports on the motherboard. Sequential read speeds are higher than even Intel's!

I also tested a 7200rpm 1TB hard drive on different ports and, as expected, there was little to no difference in performance.


840 Pro (intel SATA 3 port)


sequential read: 533MB/s

sequential write: 506MB/s

random read (IOPS): 98,952

random write (IOPS): 87,932


840 Pro (PCIe SATA 3 port)


sequential read: 553MB/s

sequential write: 530MB/s

random read (IOPS): 85,353

random write (IOPS): 85,726


840 Pro (marvell SATA 3 port)


sequential read: 391MB/s

sequential write: 176MB/s

random read (IOPS): 51,525

random write (IOPS): 29,925


840 Pro (SATA 2 port)


sequential read: 279MB/s

sequential write: 265MB/s

random read (IOPS): 50,558

random write (IOPS): 45,512




840 (non pro) (intel SATA 3 port)

sequential read: 520MB/s

sequential write: 239MB/s

random read: 89,527

random write: 27,967


840 (non pro) (PCIe SATA 3 port)

sequential read: 539MB/s

sequential write: 258MB/s

random read: 75,646

random write: 61,239


840 (non pro) (marvell SATA 3 port)

sequential read: 365MB/s

sequential write: 162MB/s

random read: 47,788

random write: 21,936


840 (non pro) (SATA 2 port)

sequential read: 276MB/s

sequential write: 240MB/s

random read: 50,823

random write: 29,887




7200 1TB drive (intel SATA 3 port)


sequential read: 183

sequential write: 180

random read: 555

ram=ndom write: 635


7200 1TB drive (marvell SATA 3 port)


sequential read: 187

sequential write: 158

random read: 559

ram=ndom write: 618


7200 1TB drive (SATA 2 port)


sequential read: 173

sequential write: 171

random read: 559

ram=ndom write: 616
Back to top
Phil Smith




Joined: 08 Mar 2006
Posts: 7717



PostLink    Posted: Sun Aug 11, 2013 8:52 pm    Post subject: Reply with quote

That all looks good for SATA III, right? The problem was I couldn't tell any real world difference. So last night I timed different programs loading. This a copy of a post I made this morning:

In testing the performance of a PCIe SATA III expansion card I came up with some head scratching results. I compared loads times of the same drive hooked up to the expansion card, on board SATA III and SATA II ports.

1) I loaded a Hollywood String Gold patch and timed how long it took to load. On all ports it took 14 seconds.

2) Thinking that ilok might be slowing things down because my test patch was the first patch I loaded, and ilok was verifying my license, I loaded a different patch first and then loaded the test patch. This turned out to be true, but it didn't make any difference in the performance between the drives. On all ports it now took 10 seconds to load.

3) I loaded up several instances of Play, each loaded with a different HS patch, and looped 4 bars. It was intentionally more than the drive could stream. I deleted tracks until the loop played without any dropouts. I had to delete 6 tracks regardless of which port I had the drive on.

4) I timed my computer booting up, from the time I hit the on button until my desktop appeared, once with my boot drive on SATA III and once on SATA II. The boot times were identical.

At this point I'm convinced there's something wrong with my motherboard. So I Google my motherboard and slow SATA III. I don't find anything.

So I search for SATA III vs SATA II and find a Tom's Hardware test. The results are the same as mine. There was very little difference in real world speeds.
Back to top
Phil Smith




Joined: 08 Mar 2006
Posts: 7717



PostLink    Posted: Sun Aug 11, 2013 8:56 pm    Post subject: Reply with quote

That post had png pics from tom's hardware that I can't post here, so here's the comparison: http://www.tomshardware.com/reviews/ssd-upgrade-sata-3gbps,3469.html The real world tests start on page 13.

This is all very surprising to me. I wish I would have known this before I wasted money on the expansion card.


Last edited by Phil Smith on Sun Aug 11, 2013 9:01 pm; edited 1 time in total
Back to top
ecrabb
Forum Moderator



Joined: 13 Mar 2006
Posts: 15909
Location: Utah

TV/Projector: JVC RS40, Epson 5010


PostLink    Posted: Sun Aug 11, 2013 8:59 pm    Post subject: Reply with quote

Phil, I'm not clear from your music forum post exactly what you tested. In each test, it was a single mechanism on a single port, correct?

SC
Back to top
View user's photo album (10 photos)
Phil Smith




Joined: 08 Mar 2006
Posts: 7717



PostLink    Posted: Sun Aug 11, 2013 9:11 pm    Post subject: Reply with quote

Steve, I tested a 250gb Samsung 840 Pro and a 840 non-Pro (lower model with slower write speeds but good read speeds, which is all I need to stream samples). I tested each on the motherboard's Intel SATA III, Marvell SATA III, and SATA II ports. I also tested the drives on the PCIe expansion card. I used Samsung's benchmark software. For what I'm using them for, the "random read" speeds are the most important. The test below is the results of the Samsung 840 Pro drive connected to the on board Intel SATA III port. There are four speed values. Sequential is non-fragmented files. Random is fragmented.

840 Pro (intel SATA 3 port)


sequential read: 533MB/s

sequential write: 506MB/s

random read (IOPS): 98,952

random write (IOPS): 87,932
Back to top
ecrabb
Forum Moderator



Joined: 13 Mar 2006
Posts: 15909
Location: Utah

TV/Projector: JVC RS40, Epson 5010


PostLink    Posted: Sun Aug 11, 2013 9:30 pm    Post subject: Reply with quote

OK, great... My point is that regardless of a particular user's application, with a single mechanism, SATA II is adequate. You've got what, nearly 300 MBps worth of real-world bandwidth, which is far, far beyond what even demanding users need. Streaming music samples? Christ, you could probably stream two uncompressed 1080p video steams with that kind of bandwidth. I would think it's massive overkill for your application.

Where SATA III would start showing some real gain would be if you want to put multiple mechanisms together in a single hardware RAID enclosure. There, with multiple mechanisms, you'd start seeing the difference between the two open up with huge combined read/write throughput. Then, and only then would you start realizing the benefits of SATA III bandwidth.

If you want to see some wicked transport bandwidth, check this out:
http://www.youtube.com/watch?v=gk69pCcVSSQ

If you need huge storage space (2TB+), AND huge throughput, you need both magnetic drives AND RAID... That's when something like Thunderbolt or SATA III starts making a huge difference.

SC
Back to top
View user's photo album (10 photos)
Phil Smith




Joined: 08 Mar 2006
Posts: 7717



PostLink    Posted: Mon Aug 12, 2013 2:08 am    Post subject: Reply with quote

You are soooo wrong about the requirements of orchestral VST streaming. I certainly understand you're not familiar enough with it to know the requirements. What I don't understand is why you would assume you know.

SATA III has 600MB throughput. My 480 Pro has 540MB. That's a single mechanism pretty much using all of SATA III's bandwidth. You would have to have 3 7200RPM drives in RAID to get the same performance. And that's WAY more than the 300MB band width of SATA II.

But if you would have read the link I supplied you would know that this has nothing to do with the lack of real world differences between SATA II and SATA III ports.
Back to top
ecrabb
Forum Moderator



Joined: 13 Mar 2006
Posts: 15909
Location: Utah

TV/Projector: JVC RS40, Epson 5010


PostLink    Posted: Mon Aug 12, 2013 2:57 am    Post subject: Reply with quote

Phil Smith wrote:
You are soooo wrong about the requirements of orchestral VST streaming. I certainly understand you're not familiar enough with it to know the requirements. What I don't understand is why you would assume you know.

Not only do I not know WTF "orchestral VST streaming" is, I've never even heard the term. Ever. I still read some of my production magazines occasionally, and I still haven't seen the term anywhere. All I know is what I've learned in the nearly 20 years since I started learning audio- and video production on the Mac back in the early-90's when there were no serious PC-based tools.

Now, I know what a VST instrument is, so I assume "orchestral VST streaming" is just layering a shitload of samples in the form of VST instruments together. Am I in the ballpark?

What I don't understand is even if the orchestral samples are 24/192 uncompressed, they should only be in the 1MB/s ballpark... You should be able to layer like 100 of them together on a computer that's nothing special. Besides, doesn't the software do any read-ahead or caching so live performance isn't so dependent on raw throughput?

Phil Smith wrote:
SATA III has 600MB throughput. My 480 Pro has 540MB. That's a single mechanism pretty much using all of SATA III's bandwidth. You would have to have 3 7200RPM drives in RAID to get the same performance. And that's WAY more than the 300MB band width of SATA II.

That 600MBps is theoretical. As you're finding out, the real world performance isn't up there, is it? Wasn't that the point of the Tom's Hardware article and your thread? That, in practice, SATA III really isn't twice as fast as SATA II?

Oh, and no doubt SSD is massively faster than magnetic... It's just wickedly expensive if you need multiple TB's.

Phil Smith wrote:
But if you would have read the link I supplied you would know that this has nothing to do with the lack of real world differences between SATA II and SATA III ports.

The Tom's Hardware link? I did read it. Well, I skimmed over it - I didn't feel like readying 20 pages of data. The wrap-up seems to be that for benchmarks, there's a significant difference, but unless you're talking server or workstation stuff, it doesn't make much difference.

I guess I'm confused. On on hand, you started a thread that's basically, "Why does SATA III even exist?", implying that in the real world, there isn't much difference. On the other hand, you blast me for understating the requirements for your application. I would think that if your application is as throughput-intensive as you say it is, then that should be right in the throughput-intensive "server/workstation" strike-zone... In other words, why wouldn't you benefit from the extra bandwidth?

Even more confusing is reading your test... Almost double the read/write throughput on the PCIe SATA III interface compared to the SATA II port. Isn't that what you're after?

OK, I guess I have no idea what the hell point you're trying to make.

SC
Back to top
View user's photo album (10 photos)
Phil Smith




Joined: 08 Mar 2006
Posts: 7717



PostLink    Posted: Mon Aug 12, 2013 3:48 am    Post subject: Reply with quote

I'm saying there's virtually *NO* difference. Hence my comment questioning the existence of SATA III. You're the one that seems to be saying there is.

There is the Garage Band thing, the EMD, trance, dub, etc. thing, which is geared toward non-musicians. And then there's another entire world for people like me that have played since they were 12 years old. And then you jump up to another level with people doing classical, cinema, TV and advertising music, which are guys with music degrees and a LOT of knowledge about music. Symphonic music is admittedly a bit over my head, but not completely. I did take some music theory classes in my younger days, so I at least partially know what I'm doing.

My point is there's a lot about MIDI sequencing that you don't know about. Film score guys don't network multiple computers to stream sample libraries, and they don't make special software JUST for the purpose of networking film score setups, because it's not very demanding. It's why I have 3 smaller drives instead of one large drive (more bandwidth).

There's a lot more to this than just Garage Band, Steve.
Back to top
ecrabb
Forum Moderator



Joined: 13 Mar 2006
Posts: 15909
Location: Utah

TV/Projector: JVC RS40, Epson 5010


PostLink    Posted: Mon Aug 12, 2013 4:28 am    Post subject: Reply with quote

OK, I'm still confused. You posted this:
Quote:
840 Pro (PCIe SATA 3 port)
sequential read: 553MB/s
sequential write: 530MB/s
random read (IOPS): 85,353
random write (IOPS): 85,726

840 Pro (SATA 2 port)
sequential read: 279MB/s
sequential write: 265MB/s
random read (IOPS): 50,558
random write (IOPS): 45,512


Those look like very significant differences. 553 vs. 279. 530 vs. 265. 85k vs. 50k. How is that "virtually *NO* difference"? Is it that those are benchmarks, and that it makes no difference to the real performance with the software you're using? Again, not clear what you're getting at.

Phil Smith wrote:
My point is there's a lot about MIDI sequencing that you don't know about. Film score guys don't network multiple computers to stream sample libraries, and they don't make special software JUST for the purpose of networking film score setups, because it's not very demanding. It's why I have 3 smaller drives instead of one large drive (more bandwidth).

Gee, thanks Phil - I didn't know there was a lot of MIDI sequencing I didn't know about. I never claimed to be an expert. 90% of the audio work I've done was for video- or multimedia production, and was mostly wave editing, multi-channel editing and mixing, and sounds effects. I've done very, very little music, and what I did was years ago. So, yeah - you could fill a pickup truck with what I don't know about music production. I've never said any different. I'm asking questions, and instead of answering them you're just telling me I don't know what I'm talking about.

Phil Smith wrote:
There's a lot more to this than just Garage Band, Steve.

Condescending and unnecessary, Phil.

SC
Back to top
View user's photo album (10 photos)
Phil Smith




Joined: 08 Mar 2006
Posts: 7717



PostLink    Posted: Mon Aug 12, 2013 2:16 pm    Post subject: Reply with quote

ecrabb wrote:
Is it that those are benchmarks, and that it makes no difference to the real performance with the software you're using? Again, not clear what you're getting at.

That's exactly what I'm saying. Benchmark numbers don't correlate with real world results. Read the tom's test starting at page 13. That's where the real world test begin.

Quote:
Condescending and unnecessary, Phil.

Sorry, Steve. You seemed to be doubting that orchestra library streaming is demanding of the system. It's very demanding. For any other kind of music it's not near as big of an issue, although some of the drum kit sample libraries can be.

Part of the reason film score composers needs so much bandwidth is they stack several different brands, of say string samples, to get a big, fat sound. They also have templates that load up all their different brands and each instrument (because of different articulations that could 5-10 instances for each instrument). The pros may have as many as 200-250 instruments in their template. They won't use them all, but they want them already loaded and ready to go so they don't have a break in the creative process.

They also spend a shitload on libraries. Here's one that most of the pros would have: http://www.vsl.co.at/en/211/442/1797/1883/1881/305.htm# And they would have several other brands as well.

I, not being anywhere near their league and doing it as a hobby, have much smaller libraries. For one I don't have near as many libraries, and I also have gone with 16bit when possible. The only difference between 16 and 24bit is 45db of headroom, and I have plenty of headroom anyway.

So what I have runs quite easily on my i7 quad 3.6ghz, 16gb. I may eventually need one more drive and 16gb more RAM, but other than that I shouldn't need anything else in the foreseeable future.
Back to top
ecrabb
Forum Moderator



Joined: 13 Mar 2006
Posts: 15909
Location: Utah

TV/Projector: JVC RS40, Epson 5010


PostLink    Posted: Mon Aug 12, 2013 5:07 pm    Post subject: Reply with quote

Cool. I knew composers had some way to hear their work as they composed, and I figured it was probably MIDI in the old days and samples now, but wow... I checked out that Vienna library and listened to the Star Wars track a bit. Pretty amazing. You can tell something "isn't quite right", but it certainly isn't obvious that it isn't a real orchestra unless you really focus and listen... There are little bits here and there that just sound, for lack of a better term, "funky".

I'm amazed that you'd pick composing orchestral scores as a hobby... Seems similar to being interested in painting so trying to paint huge chapel ceilings as a hobby, or being interested in rocketry so trying to build a rocket to fly to the moon... As a hobby. Wink

Cool, though.

SC
Back to top
View user's photo album (10 photos)
Phil Smith




Joined: 08 Mar 2006
Posts: 7717



PostLink    Posted: Mon Aug 12, 2013 7:14 pm    Post subject: Reply with quote

If you didn't know in advance that it wasn't real it would never cross you mind that it wasn't.

I'm not sure about this but I think: Only mid to big budget projects use real orchestras these days. Many only use a few real musicians for critical parts (usually strings). As I understand it, almost all TV and advertizing music are samples.

As far as harmony goes, cinematic orchestra music is actually not that complicated (mostly triadic). It's orchestration and key modulation that I need to learn. Overall it's not that far out of my realm of knowledge.

My biggest challenge is finishing a song. Any kind of song. I get tired of them before I get very far along. Laughing
Back to top
CasetheCorvetteman




Joined: 09 Nov 2008
Posts: 6319
Location: Australia


PostLink    Posted: Fri Oct 11, 2013 1:41 am    Post subject: Reply with quote

Ive been running RAID0 SSDs for about 6 years now, and it makes a big difference over a single SSD, i built a new machine very recently and tested both single and RAID0 on the Intel 6Gbps ports, and it is easily alot faster. Windows loads faster, games load faster, everything is faster by almost double. Thats just by RAID0 of two SSDs, the change from SATA2 to SATA3 ports was also a noticable change.

Back to top
the big E




Joined: 20 Apr 2013
Posts: 1928
Location: speedwell Tn.


PostLink    Posted: Mon Feb 16, 2015 7:31 am    Post subject: Reply with quote

I agree with case as I have machines that use both and the sata 3 is way faster using a ssd drive
_________________
crt king of black
Back to top
Display posts from previous:   
Post new topic   Reply to topic   Printer-friendly view    CurtPalme.com Forum Index -> Home Theater PCs All times are GMT
Page 1 of 1
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum