ReadyBoost

Started by WAS, June 11, 2018, 02:19:05 PM

Previous topic - Next topic

WAS

Is it possible that ReadyBoost actually slows Terragen down? I'm on a much older system, a Xeon X5450, with a old, and recovered HDD, so I thought I'd use a dedicated 32gb USB 3.0 for ReadyBoost to get better performance considering my HDD, it "seemed" to have worked when I was on my other PC, but I never used with without ReadyBoost, I just activated ReadyBoost, and renders which would take 3 minutes (the spires I shared) take 6-10 minutes. The only thing I can think of is ReadyBoost, or the stability of the system altogether (been getting strange errors).

Does anyone have any knowledge into ReadyBoost and Terragen? Maybe you do, Matt? It was popular back in the day.

Was also thinking of downgrading to XP for possibly better results, but not sure.

Matt

I've never heard of it.

Matt
Just because milk is white doesn't mean that clouds are made of milk.

WAS

#2
Quote from: Matt on June 11, 2018, 10:58:40 PM
I've never heard of it.

Matt

Oh really? It's write cache. Was super popular in the Vista days due to it being a memory hog. It turns a USB flash driver into a intermediary cache between memory and disk, since mine, and at the time, HDDs were mainly slow with 7200rpm drives like what I have.

For example my USB 3.0 has 6+gb/s dedicated compared to my HDDs like 1 - 2.4 gb/s max.

I have basically 32gb dedicated cache/swap instead of default like 2gb on my local drive slowing things down. Some programs done utilize it well I have read though.

If it does use it it may be why I can run scenes that people max at 8+gb ram before my objects, since I only have 8gb ram I general. But dunno if it helps with speed.

https://en.m.wikipedia.org/wiki/ReadyBoost

Matt

#3
Thanks for the info. If the system is not accessing the disk while rendering, I imagine there is no benefit. But I don't know why it would slow down your renders. Maybe you could run some benchmarks with other rendering software and see if it affects those too?


Matt
Just because milk is white doesn't mean that clouds are made of milk.

WAS

#4
Quote from: Matt on June 12, 2018, 08:50:34 PM
Thanks for the info. If the system is not accessing the disk while rendering, I imagine there is no benefit. But I don't know why it would slow down your renders. Maybe you could run some benchmarks with other rendering software and see if it affects those too?


Matt

I believe I figured out the issue, my external HDD with my multimedia is on the same USB bus, and is set to always active, performance, while the USB 3.0, while being used as a ReadyBoost, is still subject to default USB and disk settings. I flipped settings, allowing the HDD to go to sleep and the flashdrive always active and performance mode.

Not a bad idea though benchmarking. I do notice that even when ReadyBoost says it's dedicating all of the 32gb, that is not seen in system resources. I also not sure how to make sense of compressed memory in contrast to what's used. If it's compressed to 444mb, why 3.7 used?

Matt

Much of what Windows Task Manager reports is a mystery to me.

Quote from: WASasquatch on June 12, 2018, 09:04:38 PM
I do notice that even when ReadyBoost says it's dedicating all of the 32gb, that is not seen in system resources.

Maybe because it's acting at as a layer between the drive and the system?

Matt
Just because milk is white doesn't mean that clouds are made of milk.

WAS

Quote from: Matt on June 12, 2018, 09:46:09 PM
Much of what Windows Task Manager reports is a mystery to me.

Quote from: WASasquatch on June 12, 2018, 09:04:38 PM
I do notice that even when ReadyBoost says it's dedicating all of the 32gb, that is not seen in system resources.

Maybe because it's acting at as a layer between the drive and the system?

Matt

That's a good point. There's probably also more going on than just "cacheable" space.

I also have issues with windows task manager, a lot. I used to use a edited/replaceable version which was very comprehensive, I've just lost the name to time and I cannot think of it. It was open source, or if it wasn't, it at least had no bloatware and was entirely free.