-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
USB 3.0
- Thread starter SickBeast
- Start date
RebateMonger
Elite Member
imminent domain technologies don't qualify as "inventions" because there is nothing novel or exotic about them, and in particular USB 3 will not be filling any unoccupied niches. high-speed flash drives are not offered in USB because you must aggregate the bandwidth of multiple NAND chips in order to achieve the ~200MB/s performance they offer in the 2.5" form. This cannot and never will fit in the "pen drive" form factor without a revolution in NAND technology, NOT an evolution of bus technology. Furthermore it is even more senseless to claim/hope that high-speed USB drives will allow you to boot an OS in seconds. Portable OSes on external disks are unconventional and will remain unconventional until an amorphous hardware abstraction layer becomes predominant such as that used with VMware, virtualbox, etc, but even in this exceptional case, no one is going to be installing windows with one HAL onto a USB drive and then plugging it into a machine with a different HAL. It just doesn't work that way. Hot-swappable virtual servers are already occupied by SAS. Almost every possible niche is inhabited by a well-established master, and in comparison USB 3 is a clumsy, inefficient imposter.
You can blame intel all you want for holding all of the cards, but the cards are mostly theirs to hold and they aren't going to let the cat out of the bag until they get their money for the cat and for the bag. They could've released USB 3 in 2007 and integrated it into ICH10 but it wouldn't have made a damn thing faster and no one would've spent the money to integrate it or install discrete controllers. Widening your band does not make your printer faster or lower the latency of your input devices. USB 3 doesn't seem attractive at all considering 6gbit SATA/eSATA is also on its way, and the USB 3 maximum cable length isn't much longer. After 3 meters you're switched back into USB 2.0 mode.
The only application for ultra-wide USB is if you have a hub with 4 or more external storage devices running from a single bus to the host, otherwise you're wasting CPU on overhead and are better off in 2.0 mode. As far as peripherals are concerned, USB 2.0 covers all of our bases, and as far as storage is concerned, eSATA covers the remaining bases.
You can blame intel all you want for holding all of the cards, but the cards are mostly theirs to hold and they aren't going to let the cat out of the bag until they get their money for the cat and for the bag. They could've released USB 3 in 2007 and integrated it into ICH10 but it wouldn't have made a damn thing faster and no one would've spent the money to integrate it or install discrete controllers. Widening your band does not make your printer faster or lower the latency of your input devices. USB 3 doesn't seem attractive at all considering 6gbit SATA/eSATA is also on its way, and the USB 3 maximum cable length isn't much longer. After 3 meters you're switched back into USB 2.0 mode.
The only application for ultra-wide USB is if you have a hub with 4 or more external storage devices running from a single bus to the host, otherwise you're wasting CPU on overhead and are better off in 2.0 mode. As far as peripherals are concerned, USB 2.0 covers all of our bases, and as far as storage is concerned, eSATA covers the remaining bases.
RebateMonger
Elite Member
I think it's quite possible that USB 3.0 could overwhelm eSATA if USB 3.0 is done right and hardware and drivers are properly done. eSATA hasn't gotten either quite right yet.
Remember how USB 2.0 trampled Firewire, even though Firewire was "technically superior"? USB immediately ended up on every single motherboard and Firewire went bye-bye.
If you would just look on newegg you will not find a SINGLE pen/keychain flash drive that can even use half the bandwidth of USB 2.0.
You have to use multiple memory chips and accumulate their bandwidth in order to get beyond this per-chip limit on throughput. Such a device simply will not fit into the pen drive form factor. USB 3.0 pen drives will be a tad larger since the USB 3.0 connector is about double the height of a 2.0 connector. So perhaps there'll be a PCB with two NAND chips on either side. Still, that is only 60MB/s as a best-case scenario. Nothing like the 250MB/s you get from X25/vertex/etc. changing the bus isn't going to help you if your memory chip(s) can't deliver.
You have to use multiple memory chips and accumulate their bandwidth in order to get beyond this per-chip limit on throughput. Such a device simply will not fit into the pen drive form factor. USB 3.0 pen drives will be a tad larger since the USB 3.0 connector is about double the height of a 2.0 connector. So perhaps there'll be a PCB with two NAND chips on either side. Still, that is only 60MB/s as a best-case scenario. Nothing like the 250MB/s you get from X25/vertex/etc. changing the bus isn't going to help you if your memory chip(s) can't deliver.
taltamir
Lifer
You are talking about a pie in the sky... i said only a monster SSD will max out USB2. your keychain certainly does NOT... you take any keychain in the market and fit it with a USB3 plug and it will be the exact same speed as before. only much more expensive.
BTW, there are a bunch of a keychains with a USB on one end and eSATA on the other... and their eSATA side is significantly faster... but I wouldn't attribute it to the speed of the PORT, rather this is a controller issue and an issue of conversion between different data transport formats. AFAIK SATA just takes less computational power.
Forumpanda
Member
I'm just stating the obvious, but wouldn't that be a really good reason to want USB3 if it does away with some of that overhead? (which at least I thought it did).
It seems a common opinion on these forums that there are no point in improving interfaces and computational power (SATA3, USB3, Dx11, 6+ cores, SSE17, etc.) unless something that can be bought here and now exists to take complete advantage of it.
When time and time again we have seen that people with a bit more imagination than the average forum poster will eventually create a product or program to take advantage of said features.
I expect no less of USB3, I'm sure that when it eventually comes out it will be used for 'something' that wouldn't be possible (or slower) with USB2.
Its just how progress works, the infrastructure has to be there before it makes sense for a company to invest into making a product which takes advantage of it.
I personally pay more attention to new features like this, because I know eventually the products will follow.
taltamir
Lifer
If it does, then it would be worthwhile
Well, there is certainly reason to CREATE such things and buy them... but spending a LOT more for "future compatibility" does not make sense when you can just wait a year or two and get it for peanuts... I WANT USB3... but if the only way to get it is to pay 100$ extra, then i will wait two years and get it then on a new 100$ mobo.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 24K
-
-
