SATA 6 Gbps: The Right and the Wrong Way Options

[Zertz]

With the recent and upcoming release of drives utilizing the latest SATA 6 Gbps interface, manufacturers are rushing out motherboards to support the new standard. Since the P55 chipset, or any other for that matter, only support up to SATA 3 Gbps so companies have to resort to a third party controller. The only one being used, that I know of at least, is the Marvell SE9123. Using a single PCI-Express 2.0 lane (500 MB/s in both directions), it supplies two ports, but no RAID.

You probably noticed the potential bottleneck here, third generation SATA can provide bandwidth up to 600 MB/s, while the PCI-E lane will only do 500 MB/s. Well, that’s hardly a problem considering even the fastest SSDs aren’t reaching those speeds, for now at least. Anyway, I’m heading off course.
For some reason, Intel only gifted the P55 chipset with PCI-Express 1.0 lanes (250 MB/s in both directions) and there lies the issue. For a SATA 3 Gbps controller, it’s not too bad, but for a next generation controller, it causes a major bottleneck. There are two ways around this; one of them being quite obviously the better solution.

The easy, scratch that, wrong way is to take PCI-E 2.0 lanes off the processor and feed them directly to a SATA 6 Gbps (and USB 3.0) controller. However, there is a major downside. Remember that Lynnfield only has 16 lanes and you need 8 of them for a graphics card. Once these two controllers are enabled, you’re left with 14 lanes and the hardware is only designed to give 8 of them to the graphics card so you’re essentially wasting the remaining ones. The ugly part is that you lose CrossFire and/or SLI since the second slot can’t live off 6 lanes.

Then there’s the other way - the right way. It involves a PLX chip, similar to what ATI uses on dual-GPU cards. Its job is relatively simple, take 4 PCI-E 1.0 lanes and turn them into a pair of PCI-E 2.0 lanes. Simple enough. It’s just as fast as the other implementation, bare a few nanoseconds of latency, while retaining CrossFire and/or SLI capabilities. With the remaining 4 PCI-E 1.0 lanes, you can still have two PCI-E 1x slots, a PATA/SATA 3 Gbps controller and whatever else you can think of.

I’m not naming anyone, but you should be able to figure it out fairly easily

[Waco]

You could easily saturate a 500 MB/s link with two decent SSDs though, or one good one (the theoretical max of even a single 6 Gbps link is ~768 MB/s).

I'll wait for a decent implementation before I buy any of the 6 Gbps stuff.

This post has been edited by Waco: Nov 3 2009, 08:47 PM

[Zertz]

You could easily saturate a 500 MB/s link with two decent SSDs though, or one good one (the theoretical max of even a single 6 Gbps link is ~768 MB/s).

Yes, but chances are you won't be doing simultaneous 250 MB/s reads and writes on both drives at the same time. Keep in mind PCI-E is a full-duplex interconnect, you can do 500 MB/s both ways at the same time.

6 Gbps / 10 = 600 MB/s (8/10 encoding)

This post has been edited by Zertz: Nov 3 2009, 08:52 PM

[Compxpert]

Hmm so what I gather is there is support but its bottlenecked from the start.... suppose we'll have to wait for Intel or another third party to make something better but it seems the playing field is wide open.

[Waco]

Yes, but chances are you won't be doing simultaneous 250 MB/s reads and writes on both drives at the same time.

If they are in RAID 0 you will...and it's not 6 Gbps / 10, it's 6 Gbps / 8 to get GBps. 6 Gbps / 8 == .75 GBps == 768 MBps. Base 2 >> base 10.

Regardless...I'm going to wait for a non-kludged implementation before I buy anything. If the interconnect between the CPU and the HDD controller can't support the full 6 Gbps * (number of channels) then I'm not going to buy it.

This post has been edited by Waco: Nov 4 2009, 11:00 PM

[Zertz]

it's not 6 Gbps / 10, it's 6 Gbps / 8 to get GBps. 6 Gbps / 8 == .75 GBps == 768 MBps. Base 2 >> base 10.

I am fully aware that a byte is made of 8 bits. Your math would be correct if it wasn't for 8b/10b encoding.

Anyway, it seems like Intel won't be implementing it until 2011 so we're left with these alternatives for now. No word on AMD's plans. Intel's own Lightpeak is most likely the reason behind the delay...

This post has been edited by Zertz: Nov 5 2009, 12:18 AM

[Waco]

I am fully aware that a byte is made of 8 bits. Your math would be correct if it wasn't for 8b/10b encoding.

'doh! Sorry. I was pretty drunk when I wrote that.

This post has been edited by Waco: Nov 5 2009, 11:06 AM

[Guest_Digitalis_*]

First; HDDs with 6GB/sec are not performing any better than current top HDDs; so, they do not need this SATA "3".

Two Intel SSD's in Raid 0 hit till 500Gb/sec in read and you do not need SATA 3.

If you need sata 3 than there is the Asus 920 motherboard with two N200 chips added.

This post has been edited by Digitalis: Nov 7 2009, 12:54 PM

[The Smith]

Anyway, it seems like Intel won't be implementing it until 2011 so we're left with these alternatives for now. No word on AMD's plans. Intel's own Lightpeak is most likely the reason behind the delay...

AMD will be releasing their next chipsets in Q2 2010, according to VR-Zone. The SB850 will most probably support the two new standards, being USB 3.0 and SATA III.

'doh! Sorry. I was pretty drunk when I wrote that.

Reasons, reasons.

If you need sata 3 than there is the Asus 920 motherboard with two N200 chips added.

I'm not sure about the PLX chip used, but it probably only has four lanes, which is enough to provide two SATA III ports, as Zertz said. I don't see what's the advantage of using a 24-lane NForce 200 when it's probably much more expensive. If it uses the last four lanes from the PCH chip to provide an additional 2 SATA III ports, there are none left for other PCI-E devices.

Now if the NForce 200 chips are connected to the CPU, there are still only 16 lanes that communicate to the system that already prove to be a slight bottleneck for a high-end CrossFireX/SLI combo. I am using motherboards that have an NForce 200 chip to get 24 lanes for a 3x8 configuration out of 16, and when in x16 single card configuration, most of the time the board already gets 1-2 less FPS than the boards without.

Could you please link me to that motherboard? I am really intrigued about this design.

[Zertz]

First; HDDs with 6GB/sec are not performing any better than current top HDDs; so, they do not need this SATA "3".

Two Intel SSD's in Raid 0 hit till 500Gb/sec in read and you do not need SATA 3.

If you need sata 3 than there is the Asus 920 motherboard with two N200 chips added.

Your first "fact" is wrong. GB/sec != Gbps and nobody claimed HDDs need SATA 6 Gbps.
Your second "fact" is also wrong, I suggest you look up bit and bytes.
And I'm waiting on a link for your third "fact".

Like Smithy mentioned... NF200 just moves the bottleneck from the processor or chipset to NVIDIA's chip, there's nothing to gain with this chip.

This post has been edited by Zertz: Nov 9 2009, 08:26 AM