Information Technology

SETI@home – Mac vs. PC

I recently started running SETI@home again, first on my Pentium 4 3 GHz and then on my G4 800 MHz iMac. The SETI@ home website says that the the program uses about 16 MB of RAM while running and that above 64 MB will not affect how quickly data is processed. Since my iMac has 768 MB of RAM and my PC has 1024 MB of RAM memory, that isn’t an issue.

However, the SETI@home web site says that running the program in graphical mode can cause it to be slower. Here my PC clearly has the advantage, it has a ATI Radeon 9800 Pro video card with 128 MB of RAM. The iMac has a NVIDIA GeForce2 MX video card with 32 MB. However, Apple has always maintained that Macs and faster than PCs, even with the megahertz speed gap. But in the time it took my PC to process 10 data blocks from SETI, roughly 48 hours, the iMac had only processed 4 data blocks. So at least in terms of data processing power, a Pentium 4 beats a G4 no contest. I wonder how a Pentium 4 versus G5 SETI@home contest would go?

I’ll admit my test might have been a little unfair. But when Apple claims, “Even the single-processor 1.6GHz Power Mac G5 was 21% faster than the 3.2GHz Pentium 4-based system”, I would expect the systems to be a little more even.

5 replies on “SETI@home – Mac vs. PC”

Despite the megahertz gap? As if megahertz wasnt a scientific unit of measuring frequency of clock cycles….or, in other words SPEED. God I hate mac propaganda…

By the way, Im sure you’ve seen Pirates of Silicon Valley? I just watched that recently for the first time.

This is for ‘Danny’, in rememberance of that day we journeyed with Goose to the Ilink head office and made internet history.

Let’s see, your PC scored a value of 10, while your Mac scored a value of 4. So that is 10 to 4, 2.5 or 250% faster than the iMac.

Your mac is 800Mhz and your PC is 3000Mhz. Thus your clockspeed is 3.7x or 370% faster than the Mac.

I am not sure if that is an equivalent comparison, at least break it down….

The PC scored 1 per 300Mhz (10/3000Mhz), and the the Mac scored 1 per 200Mhz (4/800Mhz). Could we conclude that the Mac did more work per Mhz?

We could also extrapolate a 3.2Ghz PC versus a Single 2Ghz Mac. PC = 10.8 and the G5 = 10. That’s pretty close between the top Mac and PC.

If SETI is optimized for dual processors, then a dual Xeon could be compared to the G5/2Ghz with both processors.

It’s not propoganda, when it’s math, and it’s probably not really math when i do it :-0

Just to be clear, I never said anything about Mac propoganda. To be honest all I really wanted was one more data block out of the iMac and I would have called it even. I happen to like Macs or else I wouldn’t have bought one. 🙂

one reason like you stated that the mac failed was becuase you used the graphical interface. This is much slower than using the command line. The mac would have seen a huge increase in work units with the command line one. I know my mac almost doubled its output. secondly, the G4 processor is very fast, but it is partly due to the altivec support, which is in photoshop and FCP, ad other high end programs that the G4 shines. These programs take advantage of the altivec support. Seti, doesn’t. So while seti was optimized for the PC it was basically running on the G4 as if that G4 was a slow G3. So i think you would see a fairer battle if you used the command line version, also your PC would then really haul. We just have to wait for Seti to realease a altivec support client, so us Macs can catch up.

Comments are closed.