![](/Content/images/logo2.png)
Original Link: https://www.anandtech.com/show/3211
Thoughts on not being immune, construction, IDF, consoles and Apple
by Anand Lal Shimpi on August 22, 2005 3:03 AM EST- Posted in
- Anand
So I'm pretty sure I'm immune system limited. I managed to get sick on our anniversary, so we spent longer in NC than expected (which wasn't necessarily a bad thing, even though it meant that I was doing a lot of sleeping). The trip itself was quite wonderful, and the drive back up wasn't bad at all thanks to the fact that Vinney drove all but 2 or 3 hours of the way as I slept. But I'm better once more, and got back to work after a couple of days back here in CT.
We got a lot of house-related decisions done in NC. I never thought choosing cabinets could take so long, but it did in fact take a very long time. The fundamental problem both of us have with interior design is that we are both fans of more modern/contemporary designs, which are either not very prevalent in things like kitchen cabinetry or are ridiculously expensive.
The more and more I'm exposed to it, the more I'm blown away by some of the profit margins to be found in anything related to construction. Thanks to the fact that suppliers for new home construction have moved online, the homeowner now has a bit of a reference point to how much things should cost. We've caught a number of people charging hundreds (sometimes thousands) of percent over what online vendors were charging for the exact same parts. It's like what would happen if the only place to buy computer parts was CompUSA, and one day you ran across Newegg and your world was blown wide open.
I'm sure by the end of this project you all will have heard more than you'd like to about our struggles during the construction process, so I'll cut this part of the update a bit short out of sheer interest of saving our readership at AT :)
This week Derek and I will be at IDF, and what a unique IDF it will be. This type of IDF really happens once every 5 years, it will be the first IDF where a new microprocessor architecture is introduced. The last time this happened was back in 2000 at the Spring and Fall Intel Developer Forums, where the now infamous Willamette processor and its accompanying NetBurst architecture were introduced. With Intel's next micro-architecture a step in a very different direction from NetBurst, this IDF will be very important, mainly to see exactly how different that direction will be. Intel was surprisingly forthcoming at the Spring IDF earlier this year, let's hope the trend continues this week. I've got high expectations for Intel here and I'm hoping they don't disappoint.
Review-wise I've got a few items on my plate, including a pair of Gigabyte i-RAM cards that will work in RAID-0 (on certain chipsets). I unfortunately won't be able to play around with them until I get back from IDF, but I'll be thinking of test scenarios while I'm out in San Fran so the cards will be on my mind for the next week even if they are not in front of me. I'm also working on a budget CPU performance article and I'm trying to make it more interesting/useful than the usual, so we'll see how that turns out as well. Outside of what I'm working on, Derek has a piece on Creative's X-Fi in the works and Johan has his long awaited follow-up to his No more mysteries: Apple's G5 versus x86, Mac OS X versus Linux article.
I've been running out of storage on my home and work network, so I've been toying with the idea of building a hefty but quiet terabyte storage server for quite a while now. All of my test images, benchmarks and content for testing ends up taking up a great deal of space, and right now I'm running very low on space and very much without redundancy. As lucky as I have been with having hard drives not die on me (except those 75GXPs from a few years back), I'm not looking to push that luck any further with such important data so I think I'll start working on piecing something together after IDF.
John Carmack's Keynote from Quakecon 2005 proved to be very interesting, especially his discussion of the next generation of consoles. There are a few points that he made during his keynote that I found particularly interesting:
1) Development on Xbox 360 is much easier than on the PC thanks to fewer layers of abstraction between the software and hardware.
2) Running ported x86 code on the PowerPC cores of the Xbox 360 ends up being about half the speed of the fastest x86 processors today.
3) The point at which the transition to multithreaded game development will be able to take advantage of the incredible parallelism of the Xbox 360 and PS3 probably won't be until the next-generation of game consoles (e.g. Xbox 360 + 1, and PS4).
The third point in particular sounds very familiar to me :) Which once again leads me to believe that Microsoft/Sony didn't architect these next generation consoles in accordance with what game developers asked for, but for other reasons. It would seem that an Athlon 64 X2 or even a Pentium D (ignoring the heat issues of the latter) would be much better suited for offering great CPU performance in the way today's games are developed while still allowing for the transition to multithreaded game development.
I personally fear that the progress towards more realistic physics and AI in games has been hampered, not accelerated by the CPU architectures of these two consoles. Don't get me wrong, I think the games of the next 5 years on these consoles will prove to have better physics and AI models than what we have had in the past, but I do not think that the rate at which those areas improve will be as great as it could have been had the decisions been different.
My final point of rambling is about Apple's OS X Tiger, and no, not about the leaked x86 version. When I originally reviewed Tiger I felt that it was a very rushed project, shipped much earlier than it should have and full of little bugs that were (at least in my short history with the company's recent products) uncharacteristic of Apple. Now that my Macs are running OS X 10.4.2 I can finally say that the OS is where it should have been on day 1. Almost all of my complaints have been addressed, and the OS is just in far better shape than it was back in April. The reason I bring this up now is because when I was doing the testing for Friday's 9600 Pro Mac & PC Edition Review I fired up the G5 testbed that was still running the release version of Tiger, and it was like I was using a completely different OS. The interface felt the same, but things like OS and application stability were significantly worse than with 10.4.2.
Apple really has come a long way in these past two major updates to Tiger, but there still is room for improvement. My two complaints about Tiger happen to be about Dashboard and Safari. Dashboard's widgets continue to take far too long to become active if you just start up your computer or if you have lost your net connection. And the number of website incompatibilities (or websites where Safari stalls for no good reason) has grown considerably since I last wrote about it. I realize that this isn't always Apple's fault, but it is an annoyance of using Safari that has pushed me to Firefox on numerous occasions.
It's late and I'm sure there's more to be done before our flight to San Fran, so I'll call it quits for now. There's going to be a lot to talk about this week, so you'll surely hear from me again shortly. The fun begins on Tuesday.