Slashdot Mirror
Battle of the SATA 3.0 Controllers
Deathspawner writes "Think that all SATA 3.0 (6Gb/s) controllers are alike? As Techgage explores, that's not the case. While most SATA 3.0 controllers do deliver the performance promised, the most popular offering on the market does not — at least where bandwidth-busting SSDs are concerned. The controller comes from Marvell, and was bundled on all motherboards prior to AMD and Intel launching their own SATA 3.0 solutions. In some cases, Marvell's controller is half as fast as the others, making it no better than a SATA 2.0 controller. For those with motherboards using a Marvell controller, the solutions are few; build a new PC, or invest in a super-expensive add-in card."
Ever since they got bought by Disney, Marvell's disk controllers were never the same.
There's no -1 for "I don't get it."
I initially tried the Marvell controller on my Gigabyte X58 board for a new Agility 3. It was barely getting 230MB/s reads, and it was capped. It eventually failed to detect the drive, so I tried the Intel SATA 2 controller instead. Not only did the drive detect, but I now get ~250MB/s reads (faster random too I've read). I should've known that the company notorious for their freezing SSD controllers would do no better with the SATA controllers.
This happens all the time with computers, but especially with drive controllers it seems. The guy who rushes his half baked solution to market first at the lowest price ends up with millions of copies in nearly every computer in the world. Then a couple of years later when people start really using them, they discover that in fact the chip is full of bugs and slow and corrupts your data. It happened with the CMD 640 back when IDE first came out, the SiI 3112 when SATA first came out, and now it's happening against with SATA2. Most early Firewire controllers were total crap too, and the cheap ones still are.
The worst part is that nearly every peripheral card manufacturer is going to use that same chip because it's the cheapest. So even if you try to get around a buggy chip on your motherboard by buying a PCIe card, you'll just end up with a second copy of that broken chip. It's infuriating and I don't expect the situation to change anytime soon. That is why I always wait when a new storage access standard comes out, it's just a solid bet that the first generation chips will be way more trouble than they're worth.
I read the internet for the articles.
Don't be at the front of the technology curve when buying stuff. Let the other guy take the brunt of it all (thank you other guy for testing these things for the rest of us).
I'd be shocked to see that work. An enormous number of data links are named for, colloquially identified by, or associated with, their theoretical maximum speeds; but not hitting those often, if at all, is more or less standard. Unless Marvell or the motherboard vendor actually made specific performance claims that they failed to meet, rather than just claims of SATA revision 3.0 compatibility, their lie factor would be no greater than that of numerous other protocol silicon vendors(ethernet, wifi, etc.) who have gone legally unmolested over that fact...
Virtually everything JMicron has ever released should probably be mentioned here as well. Those guys really know how to crank up the quality...
I think you're misunderstanding: they only promise to deliver "up to" the advertised speeds!
Douglas Whitaker
It was the first controller on the market. It's got two ports each rated for 6Gbit/sec and it's connected via a single PCI-E x1 lane that's theoretical maximum is 5Gbit/sec.
Nobody should have been surprised by this at all. The information was readily available.
Basically, it's suitable for a single device that's sata6, and won't outperform the sata3 controller in some areas.
The device was only meant as a stop gap for bleeding edge users to get the capability.
Only when the expected behavior involves going to some other value...
Here's a rule of thumb I've found useful: whenever an ad says "up to" you need to replace it with "usually less than".
If God forks the Universe every time you roll a die, he'd better have a damned good memory.
It's not always the fault of the controllers, it can also be the way they're connected to the system.
These onboard controllers are connected to the system using PCI Express x1 - it's literally just like plugging them into a x1 slot only they're directly on the motherboard. The problem is there are two versions of PCI Express - the older PCI Express 1.0 provides 250 MB/s in each direction, while PCI Express 2.0 provides 500 MB/s in each direction.
AMD motherboards only had PCI Express 2.0 lanes but Intel had a mix of 2.0 lanes and 1.0 lanes - the most common was 32 x 2.0 lanes (for 2 x x16 lanes for graphics cards) and about 6 x 1.0 lanes coming from the southbridge. So motherboards manufacturers had to either use 1 lane from southbridge and get only 250 MB/s in each direction or resort to using some multiplexing chips that take 2 or more lanes and create a x4 path for the controller. More recently, motherboards detect if there is a card on the second pci express x16 and if there's nothing there, they "borrow" a few of those unused lanes to improve the performance of the various controllers integrated on the motherboard.
See this Anandtech article, it explains better than I can explain: http://www.anandtech.com/show/2973/6gbps-sata-performance-amd-890gx-vs-intel-x58-p55/2
But the point is even if the pci express 2.0 is used, there's only 500 MB/s in each direction, SATA 6 gbps means that a maximum of 750 MB/s should be reachable - very few motherboards connect the controllers to more than one 1x lane so even if the controller could reach 750 MB/s, you won't get it.
This is nothing new - remember the gigabit network cards on PCI? The whole PCI system on your computer can do 133 MB/s and a gigabit link can do about 110 MB/s - would you sue anyone if you plug 4 pci cards in your system and can't reach a throughput higher than 133 MB/s ?
Does Marvell do anything right ? I know their network interfaces are pretty dodgy, as were their SATA 2.0 kludges.
They know they're a shit company, which is why they rush things to market. Think of all the asian motherboard and add-on manufacturers that are dying to be the first to stick another starburst buzzword on their shiny boxes. Marvell released a shit product a few months before the good ones came out, so they sold millions of chips.
If the manufacturers had any standard of quality, we wouldn't have bottom feeders like Marvell, VIA, Broadcom and friends. Like all other things made in China, it's a race to the bottom. Why should we expect otherwise, when their time is so cheap compared to ours ? If I lose a month's work due to corruption, I'm out a good $5k. If they lose a month's work... well they lost less than the cost of the board.
-Billco, Fnarg.com
You mean the Corporate states of america where nVida just got the crap kicked out of them in a class action lawsuit about a year ago? To the tune of having to give all affected brand new laptops?
Yea, kindly refrain from spouting nonsense. Companies regularly get hit and found liable in class action lawsuits.
To get good SATA 3 performance you will need to spend ~ $160 or more for a controller card with a decent chipset and multi-lane PCI Express support. If you actually look at all the cheap SATA 3 controller cards for sale on Amazon (or Newegg) you'll find they're all quite similar.
Here's some examples from Amazon in the price range you mentioned:
StarTech PEXSAT32 $38
* Marvell 9128 Chipset
* PCIe x1 lane
Sybausa SY-PEX40032 $32
* Marvell 9128 Chipset
* PCIe x1 lane
ASRock 2-Port SATA 3.0 $26
* Marvell 9123 Chipset
* PCIe x1 lane
HighPoint RocketRAID 622 $39
* Marvel 9128 chipset
* PCIe x1 lane
Generally speaking if you want SATA-III to operate satisfactorily you need to use the AHCI controller built into the cpu chipset bundle. That is, the one that Intel and AMD bundle. That will get you a reliable 32-tag-per-port controller. You definitely do not want to use an external controller or a third-party chipset controller (aka Marvell), at least not if you can help it. You won't have a choice if you want hardware RAID, AMD and Intel's controllers don't do RAID (BIOS-based fakeraid doesn't count).
All chipsets have bugs, even AMD and Intel chipsets. Intel AHCI controllers have problems probing Intel SSDs (go figure) and require a driver workaround to unbrick the port when the problem occurs during probe. AMD chipsets don't mask phy errors during initial training, which creates a lot of superfluous interrupts. Both controllers play fast and loose with the AHCI spec and the AHCI spec itself is pretty badly designed, with tons of issues (though not as badly designed as the immensely idiotic USB HCIs).
Another big problem is that the firmware controller that runs the chipset side of the AHCI is typically responsible for ALL the SATA ports, which means that hotplug on one port can actually interfere with operations on another. It pisses me off, but there's no avoiding it.
The external chipsets are even worse. Marvell is a joke. Silicon Image chipsets are full of HARDWARE bugs (not just firmware bugs) which require a lot of workarounds in driver code (for example, you can't abort a soft-reset sequence reliably on a SIL chipset and you can't access the on-chip shared memory while commands are in progress without corrupting any DMA that happens to be occuring).
The stuff is getting better, slowly. The manufacturers of these chipsets have traditionally not really cared about these sorts of bugs because 99.9% of their users are consumers who don't care. The remaining 0.1% professionals who do care aren't a big enough crowd to make the manufacturers actually fix their firmware.
SATA at least has the AHCI spec, too bad more chip manufacturers don't use it. If you want to talk wireless and ethernet chipsets matters are far, far worse.
-Matt (who wrote and maintains DragonFly's AHCI driver)