Any reason to use Readyboost on Windows 7 X64 with 8GB?

Ive been reading about how windows 7 is supposed to improve and remove the restrictions of the readyboost that was in windows.

My question is if its even worth it on an X64 system with 8GB DDR3 RAM?

Would it improve the system speed even more having stuff cached right on the usb?

I have a 4GB sitting around here I would need to clean off but wanted to know what your thoughts were?

Dorfdad said:

Ive been reading about how windows 7 is supposed to improve and remove the restrictions of the readyboost that was in windows.

My question is if its even worth it on an X64 system with 8GB DDR3 RAM?

Would it improve the system speed even more having stuff cached right on the usb?

I have a 4GB sitting around here I would need to clean off but wanted to know what your thoughts were?

Hey dorfdad

Readyboost is designed for systems with little ram. If I had 8 gigs readyboost would be off and so would the page file.

hope this helps a bit

Ken

Yep I was thinking that, but being new to 7 and the new changes... Ok thanks installing RTM now :)

how do you even use all 8GB ram. I have 8GB and have not found a way to use it all. Even with virtual machine and gaming.

ReadyBoost does more than emulate RAM or Page. It complements SuperFetch. It caches certain disk I/O. I admit that I have not read any white papers on changes in Windows 7 implementation of ReadyBoost, but I doubt that MS made it dumber. It is my leap of faith that MS tweaked the algorithms that determine where small random data accesses are better served via flash versus HDD (page).

I have a self-modified USB splice inside my case, intercepting the lead to a flash reader. I soldered it to the back plane of the reader assembly. Here, I attach a 4GB thumb drive. It is internal and out of the way. If 7 ever needs it, it is there. I also keep a page file on a secondary controller RAID0 volume.

Let me define "Systems" as meaning hardware + Software. Some systems will never need the page file. Some systems will never need ReadyBoost. Some systems will experience a performance degradation using either or both. Some systems will benefit from either or both.

Many of the existing processes for analysis of page file efficiency will determine degradation vs. beneficence. Only direct examination of a particular system under a range of operational modalities will reveal the answer to the OP's question.

I think you youngsters call it YMMV.

crazlunatic said:

how do you even use all 8GB ram...

Video editing. The 8 GB @ 667 limitation of my board renders my platform virtually obsolete.

Patiently waiting for proven USB3 implementation to upgrade.

I have 8 GB and in Vista I had disabled pagefile once. Then Matlab would not work and I deisntalled it and tried all kind of things untill I realized that it needs a pagefile. Just leave it at default, if W7 doesn't need it, it won't hurt. If needed, it will use it. Unless your HDD really is small, leave it at the size W7 determines.

Unlike vista, W7 doesn't really fill up my RAM either. I haven't been using heavy software yet, though. I suppose once I do, superfetch will do more caching. In vista going from 4 GB to 8 GB sped up visrusscan, defragmentation and file transfer significantly.

I still have my 4 GB USB stick for readyboost. Not sure if it does anything. It sure doesn't hurt. If you have one, use it. If you don't have one... don't waste money buying one. I think W7 now can use much more than the vista-4GB for readyboost. That makes me think it is useful somehow.
My board has 4 RAM slots. 2 GB modules are inexpensive. 4GB modules are very expensive. So I figured readyboost would be the only feasible way to get more RAM (or fake RAM . no matter what you see how much RAM is free, you never have enough RAM! Especially not at $ 100 for 8 GB

7im? If the page file is a fixed size or range, and the Windows systems requires more memory than is available, the system will generate an out of memory error. It will NOT expand the page file.

Windows is designed to use a page file. The design aspect deals with memory allocation, not just memory usage.

I simply cannot wrap my head around "page it to kernel."

You are coming to a valid conclusion under some circumstances, but your roadmap is severely wrinkled.