CES 2012: LaCie Hub Connects eSATA Drives to Thunderbolt Macs

[Digital Skunk]

Very nice LaCie.

I still may opt for the Sonnet TBolt to Express Card 34 adaptor thought. I have a bunch of RAID arrays that need to be connected through eSATA . . . but I also have a good deal of capture cards with Express Card host adaptors.

 

[toke lahti]

I've already got a device like this. The Sonnet Echo. Been using it for a few days now and is fast and works like a champ.

If they only could add another ec slot, so it could do something more than my old non-TB MBP.

There isn't. At this time, you cannot get anything faster than eSATA for any kind of drive. And that's true of Thunderbolt drives, too -- inside the enclosure the drive is still using SATA, so that's going to be the bottleneck (unless there's an even smaller one somewhere else), not Thunderbolt. But you can connect multiple eSATA drives and/or daisy-chain the Thunderbolt to something else, like a monitor. Given that SATA is still nearly an order of magnitude faster than Firewire, it's still pretty useful.

TB has separate datapath to displayport, so you can't "loan" bandwidth from pipe to pipe.
Sata will not be a bottleneck. Sata3 has 6Gb/s, so even 2 sata3 ports can oversaturate one TB.

Hopefully, Thunderbolt also on LaCie d2 Quadra and LaCie Rugged drives. Soon.

Soon like 4 years after usb3?

 

[Thunderhawks]

No room for your fancy schmancy logic here. We're marching boldly into a brave new future of "portability" and "cleaner design". Thunderbolt...HOOOOOO!!

Don't know about "cleaner design"

This would never pass the "Hausfrauentest", i.e. it has too many areas to catch dust.

These kinds of devices are usually put behind or under the computer some place once the cables are attached.

Oh well, it will join the other dust bunnies.

 

[AidenShaw]

I've had bad experiences with LaCie drives.

Also, what's the point of Thunderbolt is faster than eSATA? To connect multiple eSATA things into one port? There has to be a new standard for internal drives that is Thunderbolt-speed.

Then T-Bolt would be the bottleneck, and you'd want faster T-Bolt!

That hub looks like the "Little Big Disk" enclosure with eSATA ports added - I wonder if it has internal 2.5" disk slots.

If that's the case, then that one little handsized package needs more peak bandwidth than T-Bolt can provide.

(Of course, unless they use two PCIe x4 SATA controllers it will be stuck at 10 Gbps peak, far less than the 24 Gbps peak that the drives have.)

 

[tbrinkma]

Why is it so big and why does the casing have to be so elaborate? I'm pretty sure all of that is costing money. I'd also have much preferred a USB 3.0 system. At least USB 3.0 drives will work at reduced capacity with the current Macs even if you take this adapter out of the equation.

My guess would be that they're utilizing an outer casing from another design in order to reduce costs. But that's just a guess.

 

[Joe The Dragon]

My Apple rumors for 2012: Thunderbolt was not developed for just a faster hard drive connection and not just so Apple could create a new display with network connectivity built in. It is much bigger than that. I think they are making an add on device that will connect to any Thunderbolt enabled computer to make it Mac Pro capable. It will have awesome graphics capabilities, much more disk expansion, and additional processors, say up to 12 more. This will provide pro level options for all of their computers. They will not have to sell a pro model anymore.

Thunderbolt is not Hyper Transport or QPI.

And less then pci-e x4 will slow down most video cards much less a cpu.

 

[*LTD*]

I've already got a device like this. The Sonnet Echo. Been using it for a few days now and is fast and works like a champ. Welcome to the party Lacie...whose drives I'll never use again after having 5 out of 7 die on me in as many months.

I've stuck with sturdy Micronet Fantom drives for a few years now. They usually come with WD or or Samsung drives (but mostly WD.) No complaints.

I'm beginning to learn that often it is not the hard drive that's the problem, but the controller.

 

[repoman27]

This is just a Thunderbolt Little Big Disk, but instead of having two 2.5" drives connected to the onboard SATA controller, it has 2 eSATA ports instead. It would appear to lack the noisy fan of the TB LBD, and the ability to add internal storage devices. I also imagine it uses the same Marvell 88SE9182 6Gbps SATA controller as the TB LBD, for which, AFAIK, there is not an available Mac OS X driver, hence it only provides 3Gbps performance and no hardware RAID capability. So total real world throughput will probably top out just under 500 MB/s.

TB has separate datapath to displayport, so you can't "loan" bandwidth from pipe to pipe.
Sata will not be a bottleneck. Sata3 has 6Gb/s, so even 2 sata3 ports can oversaturate one TB.

That's a bit of a misconception. Each direction in each channel can carry data and/or display packets. The TB protocol is agnostic to the underlying protocol of the packets it transports. The limiting factor with current generation TB controllers lies, as far as I can tell, with the on die protocol adapters. Light Ridge (the current 2 port TB controller) would only appear to have a single TB to PCIe protocol adapter, limiting the total PCIe throughput for all connected devices to 1250 MB/s (actually 1028 MB/s with a bare minimum of overhead.) AnandTech clocked throughput of 1002.7 MB/s using a single Thunderbolt Pegasus R6 and 4 SF-2281 based SSDs. I don't believe anyone has reported similar results using just two drives, since there aren't any available yet that can truly saturate a SATA 6Gbps connection (although they're getting close.)

Incidentally, Light Ridge also appears to have two DP 1.1a to TB protocol adapters and a single TB to DP adapter (DP having distinct source and sink devices.) This is why the Apple TB Display can't drive a directly connected DP display, because it is already using the lone TB to DP adapter to drive its own panel.

 

[landrew4]

For 99.99% of the population, the reason for this is because they have a drive or enclosure that supports eSata as its fastest connection (most likely offer USB 2 or 3 and maybe FW as well) or they want to buy such a drive/enclosure. If enclosures were shipping at an affordable price that supported TB, then this would be irrelevant. Buying a cheap USB 3.0 or eSata drive/enclosure but then having to spend $$$ to attach it to your mac at faster than USB 2.0 doesn't really help. If was bad enough when it was a roughly $70 delta to get FW800 support versus USB 2.0 and most people I know could not justify the difference. The hope with TB is that this new, faster, svelte interface would become as ubiquitous and cheap as USB so it didn't mean an additional "Apple Tax" the way firewire ended up. So far.... not impressed. Too bad, the future of really slim devices like the Air to become usable by more than light users depends on it.

 

[kamil97]

I like the look of the Belkin Thunderbolt dock.

US$299 for three USB, HDMI, Firewire, Audio in-out (I think) Ethernet and another TB port is pretty good.

I hope you're kidding

 

[toke lahti]

That's a bit of a misconception. Each direction in each channel can carry data and/or display packets. The TB protocol is agnostic to the underlying protocol of the packets it transports. The limiting factor with current generation TB controllers lies, as far as I can tell, with the on die protocol adapters.

Ok,
but if you observe the TB system as whole, you can carry about 10Gb/s "other data" and about 10Gb/s DP data. Not 15 & 5. Nevermind if bottleneck does not come from specs, but from controller design.

 

[S. Baldrick]

This sounds great. Sticker price will likely be a shock to those without a clue.

Could you give a hint as to what the relevant clues might be?

The cable is surely expensive, but that's because there's some circuitry involved at the connection ends: Now, surely one can infer from this, that everything involving TB circuitry will be outrageously expensive (and offer the Belkin hub as evidence); but I guess it is also possible to say, that since the cable takes care of at least some of the chips and whatnot (I know nothing about electronics, which I'm sure you've realized by now), a 'hub' wouldn't need to be as complicated, hence, less expensive.

There was a post on the OWC blog about this; they were saying that making a TB enclosure isn't as simple as throwing in the 'controller' inside one of their existing chassis—but not giving any further details as to why that is so. Is it a simple issue of the size of the controller? (They'd need to design new aluminum chassis, which would be why TB is 'difficult to implement') Is it because the controller itself is prohibitively expensive, and they don't want to release $300 enclosures as part of their sales strategy?

Anyone who knows about electronics—please enlighten an ignoramus! : )

Make this a 99$ product and we are talking.

I hope someone from LaCie is reading this...

The 'Sonnet Echo' adapter is ~$150, with TB on one end, and an ExpressCard slot on the other. I'm guessing that an ExpressCard controller is more expensive than an eSATA controller. Thus, maybe ~$100 could be a reasonable price for the LaCie hub?

Enough speculation for now... Sorry for the tech-ignorance displayed above.

 

[S. Baldrick]

Don't worry, all of these accessories will be competitively priced. Of course, there is no competition at the moment, so...

There's the Sonnet Echo TB-ExpressCard adapter ($150), and Seagate has finally announced the GoFlex TB adapter, $100 for the 2.5" model.