Yeah, USB 2 has quite a bit of overhead, CPU wise, something people should consider...
Is there more then one different optimizing utility? How long does it take to optimize a drive? Is there a GPU assisted optimizer?
Why don't plotters do this automatically?
I'll give it a go on one of my 5TB drives and see if it improves the block mine time.
Any answers to the Qs?
- Modification of DCCT Optimizer
https://burstforum.com/index.php?threads/plot-optimizer-gui-1-0-4.589/ - GUI for said Optimizer - beta
http://burstcoin.eu/download/3-dcct's-miner-and-plot-tools - DCCT Optimzer
These are the only Optimizers I am aware of.
There is no GPU optimizer because optimizing is not about doing any calculations or hashing but about re-organized the Nonces within the plot file. Maybe a GPU can do this better than a CPU but I have not seen anyone release anything. Consider Plot optimization as similar to defragging your hard-drive. The Optimizer takes all of your nonces and places them in a nice neat order so when the miner looks for a particular nonce the remaining search will be sequential instead of having to bounce around the drive in blocks that are equal to the stagger size. ( my basic understanding of the process)
Speed will be dependent on your CPU and drives. Optimizing using a read on one drive and a write on another drive should be faster than read and writing on the same drive. The J6 Optimizer allows for this method, were as the original DCCT optimizer is only capable of reading and writing to the same drive. So when using the DCCT optimizer you have to have room on your drive for both the original plot file and the optimized plot file. I use a old E5800 3GHz dual core Intel and it takes I think an 1hr 100GB to optimize using DCCT. (I forget the exact time, but this sorta of sounds right).
The GPU plotter will create an optimized plot file if you use the direct write method instead of the buffer write method. The direct method has its draw backs with increased memory usage and slower speed. I think so far it is customary to just plot using a comfortable stagger and then optimize.
Well, that's my take on your questions, maybe some-one else can give more detail or correct any of my possible inaccuracies.
Good-Luck
