Computer Specs for best case Calibre Usage

[LMF]

Originally Posted by ilovejedd

Whether it's easy to take out the optical drive depends on the laptop. If it uses normal slim optical drives, then yeah, your idea will work. I've seen some slim 5.25" ODD to 2.5" HDD adapters on Newegg. What you could also do is keep the Calibre metadata database on the SSD while keeping the actual books on a separate disk. There's a DOS batch file somewhere in the Calibre install directory showing you how to do this.

Thanks a bunch for this advice. The more I look into ssd's the more I think it'll be the most beneficial upgrade in terms of cost/performance it's justy working out how to manage the size factor. The 128gig crucial m4 is only about $170 (seems to be fluctuating between 160-180) on amazon though I'd prefer the 256gig so might have to save for that or cut costs out of the laptop in other areas.

Thanks to everyone for the the advice. I should have asked here sooner! I'm a lot clearer or what I'm looking for now - of course what I'm looking for and what I can afford will be a battle but at least I know the things I want to prioritize now.

[speakingtohe]

I bought a lenovo 2nd generation I5 laptop this week to take to the north for 5 months.

I spent a lot of time deciding on price/performance options as I wiil probably sell it in the fall. Paid $429 CAD versus $749 for I7 or $699 plus for higher spec I5.


Comparing it to my I7 desktop (Bought Oct 2010) I was pretty pleased.

Calibre loading time appears to be the same for 40 gig library. Less than 20 seconds on both.

Converting 1000 books about 10% slower on I5. ( I did not time the overall process, but looked at the job list and compared the first 10 and last 10 and converted same books on same external drive.) I was impressed that the laptop did not overheat at all during this time.

The overall time difference on 100 books was negligable (less than a minute).

My understandings are:
Processor type is more important than clock speed.
A newer I5 is comparable or better than an older I7 and far better than a non Intel mobile quad core.
A newer I7 is better than an I5 of course and fantastically better than 1st generation I5.
An SSD can make a big difference if it is 2nd/3rd generation but the price is pretty steep still for what you get. In many SSD equipped laptops it is the only storage and can get so cramped that you need an external drive, kind of defeating the difference in speed.

The processor in my new laptop is a 2410. Bottom of the line pretty well for 2nd generation I5 but I am overall pretty happy with it.

I am very comfortable with its performance with calibre and games so far. Video processing apps and photo processing with photoshop work well (can't detect a difference).

I used to keep my laptops till they were totally obsolete, now I try and sell them or give them away. Still have a couple of boat anchors hanging around And my first ereader dammit.

Prices keep dropping but one has to make a decision. I made mine on it being more than reasonable performance wise, good chance of selling for 75% of purchase price as I don't need it when I am at home.

Helen

[kacir]

Originally Posted by HarryT

A couple of years ago I had a Dell netbook with an SSD. Reading from disk was OK, but writing was mind-numbingly slow.

I have had the "pleasure" of seeing and [trying] to use one machine with such disk. One big chain of stores here was bundling them for one euro with large purchases, such as luxury lawn-mowers. The machine was slow to the point of being totally, utterly, completely unusable. Unfortunately, it wasn't standard IDE or SATA compatible SSD but some module that looked more like RAM modules.

I have also used EEE netbook from Asus with very cheap SSD that was even older and the performance was OK.

It is VERY important to do your homework before purchasing such item. Have a look at this table:
http://en.wikipedia.org/wiki/IOPS
A good SSD can have 100 times better read AND write speeds for random data.


One more thing. SSD has limited number of writes to a single memory cell before it starts to fail. This is not that bad with modern units, but still, it is something to keep in mind. So have enough RAM and try to eliminate needless swapping in OS. There are many HOWTOs on the net describing how to optimize system for an SSD disk.

[ilovejedd]

Originally Posted by kacir

One more thing. SSD has limited number of writes to a single memory cell before it starts to fail. This is not that bad with modern units, but still, it is something to keep in mind.

That's around 10,000 write-cycles for 50nm, 5,000 for 34nm and 3,000 for 25nm. Another thing, if by some chance you do reach the maximum write cycles allowed for an SSD, the SSD wouldn't just stop working. You won't be able to write on it anymore but you should still be able to read your data from the device.

The SSD is more likely to die due to controller failure or some other thing before you reach the maximum write cycles.

Originally Posted by kacir

So have enough RAM and try to eliminate needless swapping in OS. There are many HOWTOs on the net describing how to optimize system for an SSD disk.

Beware of the advice given in some HOWTOs. While they do give sound advice such as turning off hibernation, etc, some of the advice is just overkill. For example, moving the pagefile to a mechanical drive. The pagefile is one of the things that stand to benefit a lot from an SSD so moving it to a slow mechanical drive would decrease the benefits of getting an SSD. There are folks who keep nothing but the OS on their SSD. No temp files, no programs, no games - just the OS. I wonder why they even buy SSDs. So they can boast about fast boot times?

I did calculations on SSD endurance some time ago. On a 128GB 25nm SSD, you would need to write 375TB on it before you run out of write cycles. Even at 50GB/day, it will take 20 years before you run out of write cycles. I'm a bit of a power user and the only time I even write 50GB in a day is when I defragment virtual hard disks to prepare them for compaction. Take note, I'm defragging virtual hard disks stored on the SSD, not the SSDs themselves. Defragging SSDs is totally useless because of the way the controller works. Besides, SSDs actually work better fragmented. On my OS drive, I only write around 5~10GB/day on average. That's with the pagefile on the SSD. I've tested with pagefile enabled/disabled and it didn't seem to make a noticeable difference in writes. That's considering I'm using a 32-bit OS and only have 4GB of RAM (3GB addressable). At 10GB/day, it'll take 100 years before my SSD runs out of write cycles.

[kacir]

Originally Posted by ilovejedd

... some of the advice is just overkill. For example, moving the pagefile to a mechanical drive.

This is what I would do ;-). With windows, even if you have enough memory, windows still writes to the pagefile - so just in case it might need more memory it would already have some data moved to the disk. Until you can precisely set your prefferences ...
At home I use Linux, so I would probably look for ways to optimise swap space use. With Linux YOU are the master of the system. It does what you tell it to do. With Windows it is often impossible set certain behaviour.
I would also consider placing swap file on an USB memory stick. I have seen memory sticks that claimed that they are optimised for such use. But I would have to test/investigate this before use.

I have also seen installation of Windows on a mil. spec. SSD that was set to "read only" by physical switch (a jumper) located on disk. It is *highly* non-trivial operation to make Windows run from a read-only disk and you have to make ramdisk overlay for certain directories and files so Windows "thinks" that it can write to files (otherwise it goes bonkers). User data is written only on demand on D:. This brings great benefits. You do not need to run antivirus, antispyware, antimalware, updates ..., users are not capable of screwing up anything. You just reboot machine and it works like new.
This would be much more easily done with Linux, because this is how many "Live" distributions work (*), and Linux is built to have only certain directories mounted read/write (such as /var or /etc or /home)

Originally Posted by ilovejedd

There are folks who keep nothing but the OS on their SSD. No temp files, no programs, no games - just the OS. I wonder why they even buy SSDs. So they can boast about fast boot times?

Again. This is almost what I would do ;-)
Place just OS and those few programs and those few sets of data that would benefit from being on a fast disk.
Data that do not affect critical performance can be on slower disk. But, an SSD disk is very expensive item for me. And those write numbers DO look scary for the price. (**)

Originally Posted by ilovejedd

I did calculations on SSD endurance some time ago. On a 128GB 25nm SSD, you would need to write 375TB on it before you run out of write cycles.

I have two observations:
1. You would need to install special program that would trap and record every single write to some log. I suspect that programs write to the disk more often than you think. Once I was tracing complicated problem with a Visual Foxpro application that was opening some tables. So I installed a special program that has recorded every single access to the disk to a log file. I was *astonished* how many files how many files get accessed before a complex application, such as Foxpro program even starts thinking about actually opening tables with data.
2. What you write is correct, in terms of math, but you would need to have a perfect wear-leveling mechanism, and by that time you would exhaust the last writeable byte ;-)


(*) see owerlay FS, for example here: http://askubuntu.com/questions/10941...-use-overlayfs , Slax distribution uses something similar.

(**) my main desktop is still Pentium 4 with 2GB of RAM that I have purchased second-hand quite a few years ago. Until recently I have considered it to be great and powerful machine. Oh ... and many parts inside are scavenged from other systems.

[ilovejedd]

Originally Posted by kacir

I have two observations:
1. You would need to install special program that would trap and record every single write to some log. I suspect that programs write to the disk more often than you think. Once I was tracing complicated problem with a Visual Foxpro application that was opening some tables. So I installed a special program that has recorded every single access to the disk to a log file. I was *astonished* how many files how many files get accessed before a complex application, such as Foxpro program even starts thinking about actually opening tables with data.
2. What you write is correct, in terms of math, but you would need to have a perfect wear-leveling mechanism, and by that time you would exhaust the last writeable byte ;-)

You can easily view the amount of Host Writes on Intel SSDs (wear-leveling included) via CrystalDiskMark or Intel SSD Toolbox. No need for a special program to log it for you. As for other brands, you don't really need to monitor every little disk access. You just need to see the writes. For SandForce drives, write amplification can be even lower depending on how compressible the data is.

I don't have access to my main OS SSD right now but one of the SSDs I use for my virtual disk images is currently at 545GB host writes at 1151 hours of operation. That equates to just 11.4GB/day. That drive contains 3 Windows XP virtual machines (all 3 with 512MB or 1GB pagefiles) which get defragged and compacted every 2 months or so (lotsa, lotsa writes).

Yes, there's a lot of disk access but I reckon most of those are just small 4KB or so writes. Besides, it's all those small disk accesses which make moving to an SSD such an upgrade.

Here's a thread from folks trying to kill their SSDs:
http://www.xtremesystems.org/forums/...e-25nm-Vs-34nm

Several have succeeded in rendering their SSDs unusable (at upwards of 200 TiB). Surprisingly, the 2nd gen 34nm 40GB Intel (Kingston SSDNow 40GB) is still trudging along (albeit with mapped sectors).

Sure, trying to minimize the amount of data on the SSD makes sense if you don't have a lot of space on it (e.g. you only have a 40GB SSD - barely enough for Windows 7 after you factor in OS rot). However, moving everything off the SSD to "prolong" its life? Kind of a non-issue since the SSD is likely to die first due to something else (controllers seem to be particularly delicate, moreso than the NAND anyway).

[kacir]

Originally Posted by ilovejedd

I don't have access to my main OS SSD right now but one of the SSDs I use for my virtual disk images is currently at 545GB host writes at 1151 hours of operation. That equates to just 11.4GB/day. That drive contains 3 Windows XP virtual machines (all 3 with 512MB or 1GB pagefiles) which get defragged and compacted every 2 months or so (lotsa, lotsa writes).

Here's a thread from folks trying to kill their SSDs:
http://www.xtremesystems.org/forums/...e-25nm-Vs-34nm

Thank you for a very informed post.
Sooner or later I will have to upgrade my main rig and this kind of info will be extremely valuable. With falling SSD prices and improving technology they are becoming a *very* interesting option.