Yeah, flash memory isn't the best technology to use for this task. There has to be some other technology around the corner that doesn't have such a limited write cycle life.

There are new technologies being worked on. One is in the labs now. No news on when it might make it into production. Could be a year or two away still. Then again, it might show up tomorrow.

Depends on how the testing goes. I have the articles about it around here somewhere, but have to much to do right now to dig them up.

Just a word from your friendly neighborhood fear mongerer.

More on this topic, but this time, the information is directly from Intel:

http://enthusiast.hardocp.com/article.html?art=MTYxMSwxLCxoZW50aHVzaWFzdA==

I wanted to edit this, because, just by using Google, you can quickly discover that companies have produced software that can reduce this problem with SSDs:

downloads.diskeeper.com/pdf/HyperFast.pdf

And even more interesting, is the widely reported observation that Windows 7 has a similar SSD "optimizer" built-in.

Some one has even tried a SSD with Vegas.... and it appears to fail:

http://www.avsforum.com/avs-vb/showthread.php?p=15588354

The Impire strikes back:

http://news.cnet.com/8301-13924_3-10168084-64.html

Intel issues a statement saying that they cannot duplicate the results presented in the original review, that led to this thread.

I have an SSD drive in my Dell 6400. No problems at all.
The problems related in the AVS forum are not related to SSD, IMO. I've seen this problem in any number of systems, often times related to anti-virus (which isn't even mentioned in the referenced thread).

SSD's may not be the world's answer but they are also very fast, useful tools.

Some of the early SSD designs used a small cache(on the SSD card) to managed writing, this was a bottleneck issue if the OS was reading/writing, paging,etc at high IO rates. Here is an article that details the issue well for a PC site. But there are many issues being worked out on SSDs, like similar technology, every couple of months, things improve. ;-)

------------------------------------------------------------------------------------------------------
Avoid SSDs with Jmicron’s JMF602 Controller
3:40 PM - February 19, 2009 by Tuan Nguyen
Source: Tom's Hardware US – Category : Buyers Guides
21 comments

Solid state drives (SSDs) are the latest in hot tech upgrades these days, and rightly so. SSDs add huge boosts in performance to a system that is HDD bound, and the storage subsystem of a computer is the weakest link in its performance.

It's no wonder that SSDs are becoming increasingly popular. However, not all SSD drives are made alike. Don't think that you'll get great performance just by swapping out your old HDD for a new SSD. The first thing to consider is whether or not you'll want a MLC (multi-layer cell) or SLC (single-layer cell) drive. MLC SSD drives store more data, but at the cost of read and write performance, with a big dip on write performance. SLC drives rule in speed, but currently come in smaller capacities.

No matter which type of SSD you decide to pick up, you will no doubt witness instant performance gains after installing your new drive. However, with SSDs, things aren't so simple. Because of the way SSDs work, you will witness a steep degradation in performance over time. The performance occurs from a combination of things, but not because of file defragmentation like on a regular disc-based HDD.

With a regular HDD, you can simply run a good defragmentation utility to optimize the drive back to optimized conditions, and immediately experience the performance gains. On an SSD, you should never defrag the drive due to wear and tear of the flash memory cells. But this is only part of the problem. Another big problem with many SSD drives on the market is their use of the Jmicron JMF602 drive controller.

We'll just go ahead and say it right here: the controller is terrible.

First, the Jmicron JMF602 controller only has a 16 KB of onboard cache. That's barely enough to do anything. Contrary to this, the controller that Intel uses, the PC29AS21A blows the Jmicron controller out of the water with 256 KB of cache. The Intel controller also has significantly better wear leveling and write combining algorithms.

During heavy use, the Jmicron controller will literally choke on incoming data, and consequently report back to the operating system that it's buffers are filled and writes and reads need to be queued up. This puts a hold on incoming and outgoing disk I/O, causing applications to hiccup and hang while the controller chugs along. Worst, the performance degrades significantly over time as the drive is used.

Recently, OCZ swapped out the Jmicron controllers in its drives in favor of Indilink's Barefoot controller, offering significantly better performance and cell management algorithms. Other manufacturers are following suit as well.

If you're in the market for an SSD, do a bit of research to find out if the drive uses a Jmicron controller. News from Jmicron is that its updated to a new revision of its SSD controller that is suppose to solve many issues. Save yourself the problem of testing out these claims and just avoid Jmicron-based drives altogether.
----------------------------------------------------------------------------------------------

EEEK! A RAID SSD card:

http://www.theinquirer.net/inquirer/news/293/1051293/ocz-1tb-ssd-raid-module

some have suggested short-stroking up to 4 cheap hard disks for performance that approaches SSD:

http://www.tomshardware.com/reviews/short-stroking-hdd,2157.html

but I was not so impressed by the results

A friend of mine acquired a Dell system with 3x 15.5K SAS drives and controller in it which he doesn't really want, these should wind up my system very nicely.

Bob.

The cost of Western Digital 300MB 10Krpm Velociraptors has come way down, and man they are fast.