Well it’s been a while since I published a new post. In the interim I have nearly been knocked on my face by several low flying pigs. Just when you think the computer industry is predictable, along comes some new bit of weirdness to make go “gee, never thought I’d see that happen”.
Apple has switched to Intel starting with the next generation of Mac hardware. Now, the strange part about this really isn’t Apple’s decision, it’s how belligerent some (Mac) people have become. I have read more than a few predictions from the Mac faithful that this would spell the “end of Windows” or the “end of Linux”. No, I kid you not, certain Mac users relish the thought of head to head competition against the big boys of the x86 world.
Sorry, but it ain’t gonna happen. For one thing, a Mac will still be a Mac. It might run Windows, but you can bet the farm that MacOS X will not be running on anything other than Mac hardware (at least not in anything resembling a supported fashion). I could see the ability to use Mac hardware for Windows or general purpose X86 Linux as being a major boon to Apple hardware sales (in that some people and organizations would purchase one hardware platform to meet all their OS needs), but make no mistake about it, MacOS will not be running Windows or Linux out of town.
For every crackpot technoweenie wannabe, like Robert Cringely, there seems to be a crazy whodunit scenario for why Apple made this decision. Guess what, it’s simple really, Intel had what Apple needed: A guaranteed source for a full line of general purpose CPU’s. That’s it. Period. Why didn’t they choose AMD? Because Apple doesn’t want to go through this process again even if it is a somewhat simpler switch in the future. Yes, AMD is doing well today and has really embarrassed Intel (did you know the install source directory on all version of X64 Windows editions is labled AMD64?), but AMD is still a relatively small player compared to Intel. And, beyond a few percentage points of performance difference with AMD today, Intel has everything Apple wants. All said, it was the right business decision to switch to Intel. Viva la Mac!
Which brings me to my rant of the week:
Why is the computer industry press so freakin ignorant, inept and just generally wacky? Why is it that people like Robert (I) Cringely, John “No one will call it the Pentium” Dvorak and Maureen “The Stalker” O’Gara are allowed to represent the industry?! These people haven’t the technical credentials, experience or basic aptitude to be commentators for this business! Most of the time they either recycle misinterpreted press releases, garble some one else’s comments and, quite often, simply troll for publicity.
There are no standards for who can legitimately claim to be a technology journalist, but I would ask managing editors in the industry trade rags to perhaps use the following list to help them narrow down the list of those who truly qualify as “technology journalists”:
Does the person exhibit an in-depth understanding of Computing concepts. For example, does person have significant technical experience and aptitude or is a subject matter expert in a particular sphere of technology?
Does the person attempt to be a true journalist through uncovering new technology trends, reporting on specific technology and helping readers better understand technology?
Does the person attempt to present a balanced and reasonably unbiased view?
The fact is, most of the technology industries better known “journalists” don’t match any of the three criteria. Most of the big names couldn’t get a “real” job in the industry if they tried. C/Net, ZiffDavis and TechTV are littered with these worthless babbling heads.
There are exceptions, of course; most of the contributors over at Windows IT Pro and Linux Journal are fantastic and I can’t think of anything bad to say about the folks over at the two Imagine Media magazines I read (MaximumPC and PCGamer). Never-the-less, it’s time for a lot of these ignoramuses in the general tech industry press to simply go away.
Now for some geeky stuff:
At a customer site, I got my first set of Windows Server 2003 x64 boxes up and running and so far I am tremendously pleased with the performance. We are using 3 DL580’s, all with 4 3Ghz (8MB L3 cache) EMT64 Xeon’s. Two have 32 Gig’s of RAM, the other has 16 (and is actually a loaner from the fine folks at HP). We had wanted to use the DL585, but couldn’t get them in time for our project.
So what are we doing with these monsters? We are running VMWare GSX 3.2 (release candidate) on them. 10 days in and we have yet to run into a single problem with either the OS or VMWare GSX 3.2 on 2k3 X64. Now the main reason I chose X64 over 32bit version of 2k3 EE is primarily due to the potential OS performance improvements, specifically in that the 2GB kernel address space limitation disappears, which given that the actual direct execution core of VMWare executes entirely in kernel mode means that there may be some improvement in performance, especially in that kernel address space is not constrained. With that said, the actual process heap lives in user land with calls passed to the vmx86 driver, so the improvement would come from a reduced pressure on the non-paged pool set from other OS components. If nothing else, it’s nice to know that the OS itself is not doing two-level page translation tricks to address all that wonderful physical memory and that the kernel land components can address it all. We will be scaling these servers up to around 30 VM’s per box so it will be interesting to see how it pans out, but so far, my bleeding edge X64 bet hasn’t turned out too bad.
Last but not least, I have decided that I will post my Linux AD integration whitepaper sometime this coming week. I just need to verify that recent changes to Samba and Windows 2k3 SP1 don’t break my documentation.
