|
|
Wow, this came out about 4 days ago and I didn't even realize it. If you are a heavy Kali user, you almost certainly know about it already...it fixes the netplay hack which allowed cheating. Download it here. This patch only works with the expansion pack...versions for the original CD and Mac users should be forthcoming. 10:05 AM From It's Golgotha, Dammit! comes the Golgotha demo. A 3Dfx based card is required to play this preview.
Hmm, I wasn't thinking and didn't post this last night when I saw it available. A weird note, people were saying that the patch does not work if the MW2.PRJ file on your CD is dated 9/16/96 (Even if it was patched up to 1.05 later). It apparently does work if you have a MW2.PRJ file dated 10/29/96 on your CD (V1.05 out of the box). Wait, now I just read the patch has been fixed to deal with the different dates. Well, just download it, it's about 7MB. ComputerNerd has a chart of data which gives the maximum possible power dissipation for Intel-compatible CPU's. I don't know if the data is necessarily true or where they got it from, but you still might want to take a gander. The data for the Pentium Pro 150 looks right from what I saw on an Intel spec sheet, so the others might be too. I used to think these fans were really high-quality (as do many people, I think) but the more I think about it and the more I learn, the more I believe these fans are just hype and good marketing. From Intel's data on heat dissipation and PC Power and Cooling's own specs, it hardly seems like the heatsink/fan combos are any good. First of all, Intel's data (which I will post in table format shortly, perhaps this weekend), shows that you need fin heights of around 1-1.5 inches to get good figures for heat dissipation. PC Power and Cooling's Pentium Pro fans are only 0.74 inches high. Furthermore, the fan (or dual fan's, in the case of their best model) are inside the height of the heat sink, further decreasing the amount of surface area which can make contact with free air. The design of the fins only goes in one direction; typically pin fin heatsinks are better if designed correctly. And last, but not least, the Pentium Pro data they use for the specs of their Pentium Pro heatsinks assumes 23W of power dissipation. Intel did their measurements at 29 and 40W of power dissipation and recommends that thermal solutions be capable of dissipating up to 40W of power. Just so you know, the Pro 150 can dissipate up to 29 W of power and it only gets bigger from there. Draw your own conclusions, but I personally don't believe the PC Power and Cooling Pentium Pro heatsinks and fans are any good. The only good thing I can say is that the fans in the heatsinks seem to be pretty decent. The other cool thing about HP is that you can cross-ship your drive and get a replacement the next day, provided you secure the cross-shipment on a credit card so you can't rip them off. I got my drive today via FedEx...I'll probably give it a whirl tomorrow after work.
Well, HP is replacing my drive once again, no hassle, no fuss. The drives have been pretty damn bad in terms of reliability, but the customer service has always been effortless. This time they are replacing it with a 6020i! I guess they decided on that because this is my third replacement and they need to get it right. Well, I'm kind of happy about that. Hopefully this will fix the damn problem once and for all. Gil Amelio and Ellen Hancock, Apple's CEO and former chief techinical officer, have both resigned. This comes days before the announcement of Apple's financial results, which have been estimated to be around a $70 million loss for the quarter. It doesn't really look like Apple's financial results are going to be better than expected, does it? It just keeps getting uglier and uglier. I personally feel like it's been coming for a while, but I think Apple's number is up. Like it or not, I've never seen Apple being able to compete with the innovation and price competition and consequent market share that PC's have. My computer did seem fine for 3-4 days, but after I got home today, I turned on the computer and saw a big blue MACHINE_CHECK_EXCEPTION blue screen on my monitor. The sight of that error scares the willies out of me so despite my efforts, increasing the supply voltage to 3.4V and trying to keep the system cool at 233, it's gone back to 210 at 3.3V. And really, who am I kidding? 210 MHz is hardly what I would call "bad".
It should also go to show you that overclocking your CPU requires a lot of testing. A blue screen every few days is not likely to show up for you in an hour or two of heavy load. Always have a stable reference you can fall back on. Something else I've also learned is not to be in a state of denial when something goes wrong. If it looks like something is going wrong, it probably is. Blue screens are a definite no-no! My freakin CDR is busted again. This is the third time it's done this to me. Why am I cursed with this piece of shit!?!?! It's an HP4020i with the famous accursed Write append error. I hope to God HP replaces the drive for me again. I'm hoping to post some data on heatsinks and heat sink dissipation soon. The CDR thing kind of occupied my night.
Well, my computer did in fact end up crashing, but that was only after I put the case back on. I then took the case off and left the computer running stuff all day while I went to work. Came back, still going. So it's definitely a heat problem, which I was kind of expecting. I spent a lot of time today rearranging all the cables and drive to move everything out of the the way of the CPU/fan area. The fans on my CPU's are only 40mm, although they are ball bearing and spin very fast. I'm considering replacing them with larger fans, say 60mm or so. That should double the air flow through the heatsinks. I'm also trying to figure out a good way to force cool air into the case. I'm not sure exactly how or where is a good place to put a third fan (I already have two!). The problem there is that my second fan is currently blowing on the area below the CPU's, not over them. In my quest for better cooling, I've been looking through several manufacturers sites. Most of the reseller and commercial sites seem to be full of BS. I looked at NMB's site...you probably know them better as keyboard manufacturers, but they do make fans too, and quite a variety of them. I recommend taking a look...I'd sure like to replace these joe-shmoe fans of mine with the big-time ones! I also looked at Thermalloy, which seems to be the consensus for a good heatsink being arrived at by the net, and IERC, which just has some nifty diagrams and stuff. I found other info but I don't really have time to write about it now, I'll post more tomorrow.
I went to 233 MHz and jumpered the VRM's to 3.4V to hopefully stabilize operation. So far so good, I've been running for 3 hours now with no problems (Knock on wood, remember what happened the last time I thought it was stable!). I found a really neat device of my father's which allowed me to measure the CPU temperature of my PPro's. Wish I had gotten a chance to try this out back when I was running at 166. Anyhow, one of the CPU's measures at 51.3 C (touching the ceramic next to the heat spreader on the Pentium Pro) and the other measures at 56.5 C, probably due to airflow issues. There are a lot of cables and junk hanging around the hotter CPU. I hope that's not too hot. The device I used to measure the temperature of the CPU is a multifunction device manufactured by Wavetek and the model number is DM23XT. It has a little port where you can attach a thermocouple.
Here's a diagram of a Peltier cooler I found on the web. Probably the best source of heatsink information for computer-oriented people I've found on the 'net, The Heatsink Guide provides a decent amount of information on cooling your system and good ways to make or buy your own cooling solutions.
This one is from the "Clock Tower", a famous part of Castlevania legend that all you Castlevania junkies will surely know about. I especially like this one, it's got some synth guitar in it. I made some minor adjustments to the hardware pages. BTW, after some heavy research, I'm fairly confident I can stabilize my system at 233 by increasing the supply voltage to 3.4 or 3.5V. I seriously doubt the problem was heat since the processors were cool to the touch at 233. We'll see how that goes. | ||||||||||||||||||
