Charlie White (author of the infamous, and now missing Adobe "PC Preferred" article), has written an editorial piece for Digital Video Editing. Not surprisingly, it takes Apple to task on their PR this week.
Although not likely to curry favor with the Macintosh crowd, it is not a bad article. It strikes at a number of things I commented on when watching the roll-out in person: claiming the first 64-bit personal computer, aggressive use of SPECmarks, and the lack of mention of the AMD Opteron.
Recognizing, as Mr. White does not, that the PR was set long before the June 4 release of the AMD Opteron-based BOXX computer, it is somewhat forgivable that Apple didn't mention this, but I can't find a reasonable excuse for the use of SPECmarks.
The numerical benchmarks were uninteresting and are always up for debate (most of the PC folks are centering their complaints on the use of the GCC compiler for both Apple and PC tests). Although the use of Intel-optimized compilers and IBM-optimized compilers would likely have shown bigger numbers for each side, the real meat of the demonstration for me was the real-world speed test with Adobe Photoshop and Mathematica.
Those two just rocked and show the G5 for what it really can be, an excellent competitor against the Intel and Intel-derivitives for some time to come.
Another interesting point about Mr. White's comparison is that I went to the Boxx Technologies web site and priced the G5 against their 244 (the "fastest" machine that he used as an example) and a similarly configured machine was:
- $4160 MSRP
- contained no Firewire 400
- contained no Firewire 800
- contained only 4 DIMM slots
- uses 333MHz (as opposed to 400MHz) DRAM
- had 5 slots, instead of the G5's 3
- all 33/66MHz 32-bit slots, instead of the G5's PCI-X slots
- Used IDE-133 instead of SATA
- Had a TON of drive bays (9 in comparison to the G5's 3)
- Included a Floppy Drive
- Used ECC DRAM
- Lacked digitial audio I/O