Disk Controller Performance
With the variety of disk drive benchmarks available, we needed a means of comparing the true performance of the wide selection of controllers. The logical choice was Anand's storage benchmark first described in Q2 2004 Desktop Hard Drive Comparison: WD Raptor vs. the World. The iPeak test was designed to measure "pure" hard disk performance, and in this case, we kept the hard drive as consistent as possible while varying the hard drive controller. The idea is to measure the performance of a hard drive controller with a consistent hard drive.
We played back Anand's raw files that recorded I/O operations when running a real world benchmark - the entire Winstone 2004 suite. Intel's iPEAK utility was then used to play back the trace file of all IO operations that took place during a single run of Business Winstone 2004 and MCC Winstone 2004. To try to isolate performance differences to the controllers that we were testing, we used the Maxtor MaXLine III 7L300S0 300GB 7200 RPM SATA drive in all tests. The drive was formatted before each test run and a composite average of 5 tests on each controller interface was tabulated in order to ensure consistency in the benchmark.
iPeak gives a mean service time in milliseconds; in other words, the average time that each drive took to fulfill each IO operation. In order to make the data more understandable, we report the scores as an average number of IO operations per second so that higher scores translate into better performance. This number is meaningless as far as hard disk performance is concerned in our tests, as it is just the number of IO operations completed in a second. However, the scores are useful for comparing "pure" performance of the storage controllers in this case.
The performance patterns hold steady across both Multimedia Content IO and Business IO, with the ULi and ATI based disk controllers providing the fastest IO operations followed by the on-board NVIDIA nForce4 SATA controllers. The Silicon Image 3132 controller is able to provide performance almost equal to that of the NVIDIA nForce4 chipset.
With the variety of disk drive benchmarks available, we needed a means of comparing the true performance of the wide selection of controllers. The logical choice was Anand's storage benchmark first described in Q2 2004 Desktop Hard Drive Comparison: WD Raptor vs. the World. The iPeak test was designed to measure "pure" hard disk performance, and in this case, we kept the hard drive as consistent as possible while varying the hard drive controller. The idea is to measure the performance of a hard drive controller with a consistent hard drive.
We played back Anand's raw files that recorded I/O operations when running a real world benchmark - the entire Winstone 2004 suite. Intel's iPEAK utility was then used to play back the trace file of all IO operations that took place during a single run of Business Winstone 2004 and MCC Winstone 2004. To try to isolate performance differences to the controllers that we were testing, we used the Maxtor MaXLine III 7L300S0 300GB 7200 RPM SATA drive in all tests. The drive was formatted before each test run and a composite average of 5 tests on each controller interface was tabulated in order to ensure consistency in the benchmark.
iPeak gives a mean service time in milliseconds; in other words, the average time that each drive took to fulfill each IO operation. In order to make the data more understandable, we report the scores as an average number of IO operations per second so that higher scores translate into better performance. This number is meaningless as far as hard disk performance is concerned in our tests, as it is just the number of IO operations completed in a second. However, the scores are useful for comparing "pure" performance of the storage controllers in this case.
The performance patterns hold steady across both Multimedia Content IO and Business IO, with the ULi and ATI based disk controllers providing the fastest IO operations followed by the on-board NVIDIA nForce4 SATA controllers. The Silicon Image 3132 controller is able to provide performance almost equal to that of the NVIDIA nForce4 chipset.
Facebook
Twitter
RSS
42 Comments
View All Comments
LoneWolf15 - Tuesday, April 11, 2006 - link
There are TV tuner cards based on ATI's Theater 550 chip, Powercolor makes one, details can be found here:http://www.powercolor.com/product_series_Theater.h...">http://www.powercolor.com/product_series_Theater.h...
Gary Key - Tuesday, April 11, 2006 - link
We are currently reviewing the Powercolor T55E-P03 for an upcoming HTPC article. I think the results against the PCI cards will be interesting. ;-)LoneWolf15 - Wednesday, April 12, 2006 - link
From what I've heard, there's little or no difference in performance. However, the Powercolor would be the card I'd consider for future-proofness.My only disappointment is it doesn't use ATI's Remote Wonder line of remote controls; they include an iR remote of their own choosing instead of the ATI RF model, IIRC.
nullpointerus - Tuesday, April 11, 2006 - link
Interesting, how? Better, worse, wierd, or just unspecified in a frustratingly vague kind of way? ;-)Gary Key - Wednesday, April 12, 2006 - link
Actual throughput was different than the PCI based card, not trying to be vague but I think the article we are putting together will explain it best, new benchmarks, software versus hardware, TV Tuners- single, dual, SD, and HD, single core CPU , dual core CPU, AMD, Intel, NVIDIA, ATI, MCE2005, Linux, PCI, PCI-E, USB, you know just the basics. ;-)
nullpointerus - Wednesday, April 12, 2006 - link
Cool, thanks. I'm looking forward to reading it.ceefka - Tuesday, April 11, 2006 - link
I don't quite get this one. I can imagine it would say something like 1394b can be had on a s939 Gigabyte board for less than $ 100,00 e.g. GA-K8NF9 Ultra. It seems Gigabyte is the only one with 1394b for s939.
Fact is though that there are few F800 devices out there. If you do have one of these, your mobo options are limited.
JarredWalton - Tuesday, April 11, 2006 - link
It should have been "can't be very large" of course. I'm a bit befuddled on how that slipped in there, because I know I corrected that once before. Must have accidentally pasted over the original text at some point.... Ah, well - fixed now regardless.Myrandex - Tuesday, April 11, 2006 - link
I agree too, there is no reason for manufacturer's to not include this. Firewire B devices will not be mass produced without the users with Firewire B ports. I have a Giga-byte s939 SLI mobo with Firewire B on there and I do want to purchase an external enclosure that supports the standard (along with A and USB), but I also wish we would get some highly OCable boards from the likes of Asus or Abit etc. that provides this feature for the future. And also I believe the article is wrong about the price difference being very large, or else you wouldn't see Giga-byte squeezing these into ~$100 boards with other manufacturer's at the same price point including only A (or no firewire at all).Jason
Duplex - Friday, April 14, 2006 - link
1394B to the people!!! Couldn't agree more!--
I also must give credit to MSI for including a parallel and serial port.
There aren't that many people with a printerserver or USB-printer at home (I think).