comparing a couple of older CPUs

[Seeker47]

I'm about to swap some parts. (This pertains to dual-core, onward.) If there were some great cpu relative-power comparison charts to consult, I would have done so, but I found nothing online that made this clear to me. From what I did see, my working assumption is that this Xeon 3230 266 was a fairly low-end cpu, and that something like a Core Two Quad Q93xx, Q94xx, or Q95xx series would represent a major upgrade ? In particular, I'm looking to maximize video-encode performance -- short of stepping up to an i5 or i7 based box. (There are overall design-preference reasons why I intend to stick with the Shuttle P2 line, for a good while to come.)

There are also some Core Two Extreme models listed, which I think may represent the upper performance limit of what the mobo in question supports, but some other reading indicated that the Extreme series was short-lived on the market, ran considerably hotter and at a higher wattage (which I try to avoid in these small HTPC type boxes), and that most of its advantages got incorporated into the Core Two Quads.

[jagabo]

You won't see much difference between a Xeon x3230 and a Q9550. The FSB speed and L2 cache sizes don't make much difference to video encoding. Otherwise you're just looking at a die shrink and a slight bump in clock speed. You really need to move up to a Sandy Bridge part (i7, i5) to get a significant bump in encoding speed.

About the only direct comparison I've seen:
http://www.cpubenchmark.net/high_end_cpus.html

[aedipuss]

that xeon was a quad core at 2.33ghz - not a slouch by any means. for socket 775 stuff that's still available a q9550 will get you a quad at 2.83ghz for about $300. maybe a 20% increase in cpu speed.

[jagabo]

The Xeon X3230 was a 2.66 GHz part:

http://ark.intel.com/products/30797/Intel-Xeon-Processor-X3230-%288M-Cache-2_66-GHz-1066-MHz-FSB%29

So the increase with Q9550 (2.83 GHz) will only be about 6 percent in terms of CPU clock speed. Not worth $300 in my opinion.

[aedipuss]

hehe - really need to clean my glasses!

[Seeker47]

Thanks, guys. And thanks for those links. I was unfamiliar with this line of CPUs, thought this one was a 2-core rather than a 4, and just assumed that the box was sold with lower end components, as can be the case with an eBay purchase. (The Atapi burner in this thing sounds like a Cessna taking off right beside your ear.) But that is good news on the processor, since I welcome not having to spend $$ on critical parts or going to the trouble of swapping them out. I have a few boxes from the P2-chassis line -- discontinued by Shuttle, but which I regard as the best overall design of theirs that I've seen over the years -- and each of them now has something that needs to be fixed or replaced. I've already replaced the PSU on a couple of them. (Much better if it's something inexpensive, like a cable or a burner.)

In regard to the processing power thing, it occurs to me that there is more than one valid approach to multitasking. The approach I'm likely to take is to have more than one box working in more than one location. If one rig is left to just a video job, and that job takes 70 minutes instead of 40, I won't care about it that much if I can keep working on something else -- plus web and email -- in the adjoining room.

[jagabo]

If one rig is left to just a video job, and that job takes 70 minutes instead of 40, I won't care about it that much if I can keep working on something else -- plus web and email -- in the adjoining room.

Just run the encoder at a low priority. You'll hardly be able to tell it's running in the background:

[Seeker47]

You won't see much difference between a Xeon x3230 and a Q9550. The FSB speed and L2 cache sizes don't make much difference to video encoding. Otherwise you're just looking at a die shrink and a slight bump in clock speed. You really need to move up to a Sandy Bridge part (i7, i5) to get a significant bump in encoding speed.

About the only direct comparison I've seen:
http://www.cpubenchmark.net/high_end_cpus.html

I would not have guessed there was only a 6% difference between those two chips.

That's a very useful chart. At a glance, it does not seem to extend back to dual-core and earlier.
(In some earlier models, and based on running a variety of common apps -- all with XP and the same amount of RAM -- I noticed a rather perceptible difference in performance as between the AMD 4600+, the 5600+, and the 6000+. {I'm not referring to any games, or anything that really stresses a GPU.} In that example, I think we're talking about more than 6%.)

Re the background encodes, I'm always wondering if some computer glitch or crash in some other app might ruin an ongoing video-processing task, so I'm thinking a rig that is doing nothing else might be the best bet.

[jagabo]

That's a very useful chart. At a glance, it does not seem to extend back to dual-core and earlier.

They have benchmarks for older processors too. Go to the main page:
http://www.cpubenchmark.net/

Re the background encodes, I'm always wondering if some computer glitch or crash in some other app might ruin an ongoing video-processing task

That's extremely unlikely in modern Windows systems. Each process runs in its own protected memory space and can't see or touch other processes. Of course, Windows could crash but I've hardly ever had that happen since starting to run XP. I almost always have a rendering task running in the background. I don't worry about it at all.