Will Thunderbolt make user upgradeable external graphic solutions a reality?
Let’s face it, no traditional hard drive will ever saturate Thunderbolt’s 10 Gbps bandwidth (1.25 GB/s throughput). Even a solid state drive will have a hard time doing it. That’s why the new LaCie Thunderbolt offering is a pair of SSDs in RAID 0. Because anything less is not enough to show end users the difference.
So setting that aside. What could be an immediate use of Thunderbolt that we can take advantage of? NVIDIA and AMD take note: external video card solutions.
We have seen them before. Proprietary solutions of external video adapters for notebooks. None of them really materialized or caught much fan-fare. Well I don’t see why they would, they’re expensive and often are restricted to certain manufacturers.
AMD has attempted this once before, but it was still a proprietary solution. Their solution is still active and provides a much more handsome 4 GB/s. That translates to a bandwidth of 32 Gbps.
Intel’s plan for Thunderbolt is to eventually reach 100 Gbps. 100 Gbps bandwidth = 12.5 GB/s. Remember, Intel’s figures are not theoretical, these are not the peak transfer speeds. This is the actual sustainable transfer speed. If AMD’s solution work with 4 GB/s throughput, what can they do with 12.5 GB/s throughput?
This is the dream device: Enclosure that can house an enthusiast class video card using Thunderbolt as the interface.
Make it, and people will buy.