Pre-Test Briefing: New Dual-Core and Quad-Core G5/2.5 Power...

[MacBytes]

http://www.macbytes.com/images/bytessig.gif (http://www.macbytes.com)

Category: Apple Hardware
Link: Pre-Test Briefing: New Dual-Core
and Quad-Core G5/2.5 Power Macs (http://www.macbytes.com/link.php?sid=20051023202550)

Posted on MacBytes.com (http://www.macbytes.com)
Approved by Mudbug

[gregnacu]

Quote:
"Apple says, in addition to moving from DDR 400MHz memory to DDR2 533MHz, the memory controller has been made for efficient than the previous model of G5 Power Mac, increasing the throughput to 8.5GB/s versus 6.4MB/s."

This can't possibly be right. 6.4 "Megabytes"... to 8.5 "Gigabytes". Is it just me or did this article have a lot of mistakes. "...been made for efficient than the previous..." Shouldn't that be, "...been made MORE efficient than..." An article with so many errors doesn't instill much faith.

[Eniregnat]

Quote:
"Apple says, in addition to moving from DDR 400MHz memory to DDR2 533MHz, the memory controller has been made for efficient than the previous model of G5 Power Mac, increasing the throughput to 8.5GB/s versus 6.4MB/s."

This can't possibly be right. 6.4 "Megabytes"... to 8.5 "Gigabytes". Is it just me or did this article have a lot of mistakes. "...been made for efficient than the previous..." Shouldn't that be, "...been made MORE efficient than..." An article with so many errors doesn't instill much faith.

Yes it does seem improbable. Mistakes of scale often happen. Perhaps an email to the author, with a link to this thread.

[Santaduck]

A typo here or there doesn't bother me in the least.

This author is one of the most thorough, systematic, and open benchmarkers that I have ever run across. I have full confidence in his eventual results, and if you don't know the quality of his previous benchmarks, or the reputation of his site, just browse through any of his articles. His mac benchmarking site barefeats.com (http://www.barefeats.com) is well-known, and for good reason. His pc vs mac (http://www.barefeats.com/macvpc.html) post was a great article earlier in the year, and he's known to consider every last variable in every bit of hardware and software, and will tell it like it is.

[sjk]

A typo here or there doesn't bother me in the least.I'm capable of forgiving typos while admitting they can be distracting. They make certain writing appear rushed and "unprofessional", which I'd hope isn't the impression the author wants conveyed to readers.

As gregnacu mentioned, "An article with so many errors doesn't instill much faith." I have to agree that typo-oversloppiness in documenting results where accuracy is important could make some readers wonder if there haven't been errors in the actual testing.

This author is one of the most thorough, systematic, and open benchmarkers that I have ever run across. I have full confidence in his eventual results, and if you don't know the quality of his previous benchmarks, or the reputation of his site, just browse through any of his articles.Rob-ART's testing provides a useful service to to Apple/Mac community, but that doesn't necessarily imply his information is always accurate and authoritative.

His mac benchmarking site barefeats.com (http://www.barefeats.com) is well-known, and for good reason. His pc vs mac (http://www.barefeats.com/macvpc.html) post was a great article earlier in the year, and he's known to consider every last variable in every bit of hardware and software, and will tell it like it is.
Rob-ART's test results are often quite informative, but claiming he considers "every last variable in every bit of hardware and software" is clearly an exaggeration. Let's get back to reality, eh? :)

[Santaduck]

ah well fair enough.

Typos in journalism do betray some dent in professionalism, but of course he's a one-man operation, not a publishing machine with layers of editing.

I enjoy his benchmarking compared to the print magazine benchmarking that we see all over, although to be fair to them, they have space considerations that may hinder listing every exhaustive detail that they did in fact consider.


ps- where are you in hawaii btw? I'm over in Makiki, and dropped by the Hawaii mug meeting for the first time a few months ago. Did you know Rob-ART when he used to live over here?

[sjk]

Typos in journalism do betray some dent in professionalism, but of course he's a one-man operation, not a publishing machine with layers of editing.Two-person, officially. And a few typos are expected from a chief mad scientist. :)

I enjoy his benchmarking compared to the print magazine benchmarking that we see all over, although to be fair to them, they have space considerations that may hinder listing every exhaustive detail that they did in fact consider.His resources are certainly more limited than larger operations with bigger labs, etc.

I've always enjoyed the casual style of his writeups, although I'd appreciate more details about how certain tests were run to arrive at the published results (e.g. how many passes of Xbench) and briefly suggested that a few years ago.

ps- where are you in hawaii btw? I'm over in Makiki, and dropped by the Hawaii mug meeting for the first time a few months ago.Windward. I used to go to HAPA meetings before they fizzled out a couple years ago after one of the main organizers moved to the mainland.

Did you know Rob-ART when he used to live over here?Yep. He was the first Mac user I made contact with after moving here ~6 years ago.