Dave's Workshop - Workbench August 26th

Workbench

Week of August 26^th - September 1^st 2001

Sunday

Monday

Tuesday

As if on cue, Intel has released a 2 GHz processor. If it seems like only yesterday that 1 GHz machines were available you're right and wrong. About 18 months ago, both AMD and Intel "launched" 1 GHz processors, but it was some time before they were generally available (especially from Intel). Interestingly, 18 months is about what is expected to double the speed of processors these days, as indicated by Moore's Law. This is somewhat misleading though, as there is more to a processors speed then raw MHz. As any Mac fan will tell you, the PowerPC processor is far superior clock for clock then Intel's x86 architecture. And only at 2 GHz is Intel pulling away from AMD's 1.4 GHz processor in most benchmarks. But MHz, now GHz, sells, so it will be interesting to see how AMD fares in this "MHz Gap".

Apparently, Microsoft is getting support from the Dead in its attempts to settle its anti-trust cases with the government. The Seattle Times is reporting a "grass roots" campaign directed at the state and federal government. Now I have nothing against a company or organization rallying there supporters to make their voices heard in Washington. Now I know the materials to do this are often supplied (form letters, postage, etc...), but this seems to have crossed the line.

Also about Microsoft, Ars has a nice article talking about the differences between Windows XP. As I've said before, I'm surpassingly interested in XP. However the recommended system requirements are fairly steep. What's more, I really don't see a good reason to upgrade. So unless you're buying a new machine that comes with it, I would thing twice about making the move.

Wednesday

Tom, of Tom's Hardware Guide is reporting that AMD will re-institute a form of the old PR Ratings. In the mid to late 90s AMD and others created PR Ratings to compare the relative performance of their lower clocked processors to Intel's higher speed processors. For example, I have an AMD K5 PR166 in a machine at home. What AMD was saying about the chip is it performs as an Intel Pentium 166. In fact it runs at around 133 MHz (I think), but AMD felt it performed on par with Intel's 166 MHz chip and thus should be marketed as such. Or maybe more correctly, had to be marketed as such because AMD had nothing to counter Intel in direct MHz speed. And that's what's happening again. With Intel's releasing a 2 GHz P4, and a 2.2 rumored on the way by the end of the year, AMD finds itself behind the eight ball. Currently, AMD's fastest processor runs at 1.4 GHz. They have a new "core" or version that should scale to 1.7 - 1.8 GHz by the end of the year, but they're afraid the 500 or so MHz gap will spell doom in the eyes of consumers. And they're right, that's why it looks like they're going to start referring to their processors with model numbers and not MHz speed. Now I agree with AMD that their processors are MHz for MHz faster then Intel. I'm planning to buy an AMD Athlon 1.4 GHz chip in a week or two and one of the reasons is it's faster then it's Intel counterparts. But how much faster? Certainly it's faster then the equivalent P4 at 1.4 GHz and probably a 1.5. What about against a P4 at 1.8 or even 2 GHz? Unfortunately, it's hard to tell, and that's where the whole PR Rating or Model numbering scheme braked down. There's no good way to directly compare different processors. You can run benchmarks until your blue in the face, one will show you one's better, and another shows the other. Comparing the Athlon at 1.4 GHz to the P4 at 2 GHz is a perfect example. In most benchmarks the P4 has pulled ahead of the Athlon, some by a considerable margin, but a few have AMD coming out on top. Does AMD now say it's 1.4 GHz chip is a Model 2000 (a 2000 MHz equivalent, or 2 GHz)? Certainly not! And from Tom's article it looks like their going to tack on about a 15% addition to the clock speed (1400 MHz will equal Model 1600) which is probably pretty close overall. That would also let them stay within striking distance to Intel, as a 1.8 GHz Athlon could be sold as Model 2100, close to a P4 at 2.2 GHz. Personally, I'm split on this issue. I certainly sympathize with AMD as the Athlon is definitely faster then a P4 at the same speed. However, using an arbitrary rating just seems... well arbitrary. What's to stop Intel from countering with their own advantageous numbering scheme. In fact, one could argue that Intel has done just that in designing the P4 to scale better in terms of MHz at a cost to what the performance should be for marketing reasons alone. For just as AMD's Athlon is faster then Intel's P4 clock for clock, and so is it's own P3 in a number of cases.

Thursday

Friday

AnandTech has an interesting piece on Linux in the corporate world. It's a short look at a company who left UNIX/Novell world for greener pastures in Windows NT, and the difficulties they encountered. Of course Linux is introduced in the end to save the day. This scenario is exactly where I see *nix making a come back. The company has around 7,000 employees, which means it's big enough to have resources to spend on IT, while at the same time it's small enough to more easily adapt to changes larger companies would shy away from.

It's been what, a week, sense the last virus scare. Fear not, a new one has surfaced. I usually don't talk a lot about viruses because in reality new ones are constantly being found. However, this one comes in the form of a notice from Microsoft about patching your system. Like the "I Love You" virus, an email supposedly from Microsoft might slip past people's defenses.

Finally, another rumor about Dell using AMD chips. Like others that have been false, this one looks good. Who knows, maybe someday Dell will branch out from only offering Intel processors. As we use Dell PCs at work, I would love for this to happen and AMD could really use the support, but if I were Dell and getting a sweetheart deal from Intel (as they are rumored to be) to stay away from AMD I probably would too. I can't help thinking if Dell was going to embrace AMD it would have been when AMD hit 1 GHz and kept going while Intel was stuck waiting for the P4.

Saturday

HOME

WORKBENCH

COMPUTERS

HOBBIES

ABOUT

Workbench

Week of August 26^th - September 1^st 2001

Sunday

Monday

Tuesday

Wednesday

Thursday

Friday

Saturday

HOME

WORKBENCH

COMPUTERS

HOBBIES

ABOUT

Workbench

Week of August 26th - September 1st 2001

Sunday

Monday

Tuesday

Wednesday

Thursday

Friday

Saturday

Week of August 26^th - September 1^st 2001