Back in the beginnings of our company, my roommate and I had equivalently large and clunky laptops. His was a Dell Inspiron 5150 which experienced frequent heat problems. His video failed on him (fortunately under warranty) and he sent his laptop RMA'd to Dell. Dell repaired the laptop and shipped the repaired product -- not to the return address on the RMA, but to the address that the laptop was originally shipped to. Good one, Dell. Our old company refused shipment of his 5150, and it ended up somewhere in the bowels of Dell's shipping department. Fortunately for him, six months later, he called Dell, and his problems were solved by a particularly talented service rep with a thick Indian accent. My lucky roommate was shipped a brand new Dell laptop with the highest resolution display that I have ever seen on a laptop.
Unfortunately, during the RMA process, my roommate confused the power supply for his 5150 with my laptop's power supply, and sent it to Dell alongside his 5150.
As I had mentioned previously, we both had large and clunky laptops; his power supply was a 120 watt Dell, and mine was a 120 watt Toshiba. Earlier, I had tried a replacement power plug from Radio Shack, which was small and of terrible quality. After two years of procrastination, I finally ordered a replacement plug from All Electronics Corp. Last night I got around to soldering the plug onto the truncated Dell 19-volt output leads.
That brings us to the subject of undervolting. My illustrious
Toshiba A75-S209 was purchased along-side a Best Buy service agreement. The A70 and A75 laptop families were completely defective and they suffer from an electrostatic discharge issue that requires motherboard replacement. In fact, a class-action lawsuit against Toshiba on behalf of the owners has resulted in a settlement, and after showing the Best Buy employees the Toshiba service bulletin, they took it under their wing. I retrieved my repaired laptop from Best Buy a month later, and to my surprise they had even fixed the damage that I had done to the unit by a four-foot drop onto concrete (ouch!), and had even upgraded the laptop's hardware by adding an integrated five-function digital media reader (score!). By that time, I had purchased an IBM ThinkPad T42 (which I HIGHLY recommend). So there my poor Toshiba sat on the shelf for nearly a year, condemned to a lonely existence befitting a loud, inefficient, hot, unreliable piece of hardware.
Until last night. A while back, I had read an excellent article on Nordic Hardware talking about "undervolting". "Undervolting" reduces the power consumption and heat output of a computer. Over a year ago, I had successfully manipulated the CPU core voltage on my IBM laptop with great success. Remembering that one of my biggest issues with the Toshiba was that it produced a huge amount of fan noise under normal operating conditions, I realized that it may be possible to rectify poor thermal dissipation by reducing power consumption.
A few tidbits of interest to the Toshiba "undervolter:"
1. The latest RMClock utility, version 2.2, is completely useless because modification of the P-State Transition tables is impossible. It is important that you download RMClock version 2.05 or earlier.
2. Very few hardware monitoring programs work on this poor laptop. I was able to find a single program -- hmonitor -- that can display the CPU temperature.
3. A useful tool to help stress test your machine is the Video Card Stability Test from the FreeStone Group. I have noticed that my systems, when deemed reliable under heavy load executing the Prime95 torture test, will abort when exercising the graphics card.
I was exceedingly surprised to discover that I could reliably reduce my core voltage on the Toshiba's 3.06 GHz Pentium 4 Prescott D0 from 1.35V to 1.175V -- this is the largest reduction in core voltage that I have witnessed; consider that the vmin for this CPU is 1.150V. Under maximum load, the laptop makes an elevated fan noise. Now, under normal load, the laptop is very quiet.