Fusion-io is putting a rocket up workstations in Hollywood with an ioFX flash card that features some of the technology used to create the visual effects in orphan-meets-robot family flick Hugo. The ioFX is a PCIe-connected card with 420GB of non-volatile memory on it, providing a tier of storage between the workstation's DRAM …
Dumb Question: Why not more RAM ?
So this card is basically acting like a large cache for the HD.
So why not just put a shed load of normal DRAM memory into the machine, and let the O/S filesystem caching kick in ? Or use the memory as a RAM Drive?
Surely DRAM's going to be quicker & cheaper than flash ?
Re: Dumb Question: Why not more RAM ?
Quicker to write, sure, but the idea of caching is that you're reducing the amount of writing. On the read side, they'll both be the same speed.
And not cheaper. Quick Google suggests £3 per GB for DDR3, and under £1 per GB for pen drives.
Re: Dumb Question: Why not more RAM ?
"So why not just put a shed load of normal DRAM memory into the machine, and let the O/S filesystem caching kick in ?"
Because having your cache reads and writes on the same memory bus doubles the traffic, effectively halving the data throughput.
As PCIe is on the memory bus anyway, you write once to PCIe. You presumably read it off the card without involving the system bus at all.
That should, in theory, prevent the system bus becoming a bottleneck....
this is silly
Are we seriously expected to believe that 420GB super fast storage is the only "magic ingredient" necessary to create special effects in CGI in real-time? And where is computing power coming from - since I refuse to believe that ordinary GPU is enough for cinematic movie-quality CGI.
I must have missed something important about this release, so it's perhaps me who is silly.
Re: this is silly
An ordinary GPU *is* enough. It took a lot less than a modern GPU to make the FX for everything from Tron to T2. Something with the same spec as WETA Digital's setup for the first LotR film is within reach for a keen and well-heeled amateur. The crucial ingredient is time. Halo doesn't look like Avatar simply bcos playing Halo needs a new frame at least every 100ms, and at that it's pretty ropey. But if you're rendering CGI, it doesn't matter if each frame takes a week to generate - you can generate frames a week at a time and print them as they arrive. The only questions are how long you want it to take for the render, and how realistic your CGI models are (both of physical items, and of laws of physics).
try getting 400GB of ram in a workstation?
ever try that? Yeah, didn't think so. Also having it be non volatile helps since you can turn your computer off at night, or suffer a crash and recover quicker. Staging 400GB of data from disk is going to take a while. The cost of $2500 seems really cheap if that is for 400GB of flash, a fraction of what most of their other products cost (at last check anyways).
As for Cisco becoming one of the largest Fusion IO customers - why would they be ? I mean Dell, HP and IBM have been customers/OEMs of Fusion IO for some time - any idea where they rank on the customer list? HP, to-date, as far as I know is the only one to have OEM'd fusion IO to put on a mez card on their blades.
Obviously HP/Dell/IBM have bigger market shares in servers than Cisco..
If Dell/IBM/HP are each already more than 10% then forget this comment !
- YARR! Pirates walk the plank: DMCA magnets sink in Google results
- Pics Whisper tracks its users. So we tracked down its LA office. This is what happened next
- OnePlus One cut-price Android phone on sale to all... for 1 HOUR
- UNIX greybeards threaten Debian fork over systemd plan
- Ex-US Navy fighter pilot MIT prof: Drones beat humans - I should know