Why is my PC so slow?

Howdy all. Why does it seem that my PC is so slow?


Back when I was a Wind River customer (yeah, yeah, dinosaurs walked the Earth and the planet’s crust was still cooling) VxWorks 5.1 developers used the command line in a ‘shadow’ directory of the product distribution. Editing was done with ‘vi’ and BSPs and apps were built with ‘make’. We lived in the era of ‘burn and churn.’

All of this happened on big (for the time) shared Sun 3 boxes. PC’s were a novelty — the exclusive toy of the accountants and finance people (who didn’t have any real work to get done.

These days the masses are blessed (or is it burdened?) with more computing power than they can figure out how to use. This has been a boon to distributed projects like SETI@home and Folding@home. Who would have though that there would be teraflops of performance in a video card?

But developers are (always) different! With the distribution of full OS and network stack source in the VxWorks 6 era, customers are building more of our product. At the same time apps are bursting in size. There never seems to be enough performance.

So why does compiling this take so long? I have a fast CPU and lots of RAM – what could it be?

HD Tune to the rescue! With a tiny bit of attention to detail and the addition of a $15 PCI RAID card, I now have a disk I/O screamer!

Pio0_custom_1I started out in PIO mode because I had a problem with the chip set driver. Download pio0.JPG


I then went from UDMA2 Download udma2_40pin.JPG


to UDMA5 by simply swapping the 40-pin cable for a high quality 80-pin unit. Download udma5_80pin.JPG

Sil680a_raid_customAll the best goodness comes from adding a $15 RAID card to the pair of drives I already had! Download SiL680a_RAID.JPG

This is really only a solution for local copies of files which are maintained in a reliable configuration management system!

Keep in mind with increasingly larger capacity disk drives (driven by serial ATA drives), the risk of drive failure is increasing. Additionally, bit error correction technologies have not kept up with rapidly rising drive capacities, resulting in higher risks of encountering media errors. In the case where a failed drive is not replaced in a RAID 0+1 configuration, a single uncorrectable media error occurring on the mirrored hard drive would result in data loss!

Given these increasing risks with RAID 0+1, many business and mission critical enterprise environments are beginning to evaluate more fault tolerant RAID setups that add underlying disk parity. Among the most promising are hybrid approaches such as RAID 0+1+5 (mirroring above single parity) or RAID 0+1+6 (mirroring above dual parity).

1 Comment

Comments are closed.