Originally posted by kenohki
The only benchmarks people have mentioned here (besides silly Photoshop bakeoffs) are Xinet server benchmarks. This proves what? That the G4 can keep up with 1GHz PIIIs. Okay, great, and while the P4 is at 2.53 GHz on a 533MHz FSB (and 3GHz by years end or sooner), we'll sit around and pat ourselves on the back about how we've finally caught up with LAST YEAR's Wintel machines. Granted, the P4 does less per clock than the PIII or G4 but you can't reason away 3GHz processors to the "megahertz myth". [/B]
I'm not going to defend Apple, or more specifically Motorola on these new releases. They are certainly an improvement, but they are more like a stop gap between the Old towers and modern PC technology. The latest processor to chipset bus is NOT DDR! Only the memory to chipset is DDR.
but... don't dismiss the Xinet benchmarks or the photoshop benchmarks. The xServe is a SERVER! The benchmarks done by Xinet are the types of benchmarks that server vendors routinely tout to sell boxes. They are perfectly valid, they are important indicators of design-house server needs, and they are compiled by an independent source. By the way, check the Xinet benchmarks again... you don't seem to appreciate how well Apple fared, especially in price to performace (lower price, more performance than Sun OR Dell).
As for photoshop, it is an important benchmark for a variety of reasons.
1) it has a common, cross platform, code base
2) it is optimised for a variety of platforms so you can't claim it is tweaked for one platform and not another
3) it is mathmatically intense, but it also stresses the memory subsystems
4) it can stress interger, FP, or SIMD performance... or all of the above depending on the batch of tests.
5) and for Apple, it is the 'holy grail' of one of their core markets. Lost time in Photoshop is lost productivity and lost money.
Here is the latest blurb from the G4 site...
The twin-engined 1.25GHz G4 runs professional applications like Adobe Photoshop up to 90 percent faster than a 2.53GHz Pentium 4-based PC
Not independent, but I bet Apple can back up that statement.
Right now, a Dual G4 offers up to 2.5GHz of clocks (minus the overhead associated with SMP). This seems to fare very well to the 'architecturally challanged' P4. I'm sure even Motorola could ratchet up the clock a lot more if they put 23 stages in their pipeline. Apple's problem is price... I'd say the performance of the low end dual is fairly close to a similarly priced (not garage built) P4s in Apple's core market applications, but if you look at what a high end Dual G4 costs you... then you have to start comparing to dual Athlons and Dual Xeons where Apple isn't going to keep up.
The 3GHz you mention has not been released yet. I bet they could get real close now but Intel is screwing the consumer and waiting for AMD to get closer before they release new processors. As for the Mac, there is still a 7470 processor on the road map (likely for this year) that WILL clock higher, if only due to a smaller process, and which will perform much better overall due to Rapid I/O DDR support and a 512K L2 Cache. If you want to talk 3GHz P4s, you should compare them to Dual 1.5 GHz 7470 processors. It is only fair to compare vapor to vapor isn't it?
... but of course, in the end you should look at TCO when you look at price or even price/performance shouldn't you? Could you show us one, just one, report that states Windows PCs have a lower TCO, or a higher level of productivity than Apple computers? If you do, I guarantee the rest of the posters here could find ten independent reports that state otherwise.
Even INTEL released a report a few years back that their macs cost less to operate than their PCs. I'm sure they fired the guy that leaked that one.
Just my 2cents.
(where's my freaking 1.6+ GHz G4s!!!) ... ffakr
