Slashdot Mirror
Ten Lies About Microprocessors
cloudkj writes "Processor selection too often turns into a religious war. Debunking the dominant myths is the first step towards making a rational choice. Embedded.com has an article highlighting the 10 most common lies and misconceptions about microprocessors."
Based on actual performance.
"PPC by design handles better than x86. The Ghz is meaningless."
Except that it is one of the ways that x86 way outperforms the PPC. PPC is so far behind; the speeds of the current ones were "state of the art" for x86 back in 2001: downright languid.
Yes, actual performance does not matter. That is one of Apple's main marketing messages when marketing machines with sluggish processors in them to try and fool buyers into thinking it does not matter at all.
#3: Instruction sets don't matter
I would say:
#3. The clock speed is a good estimate for processor performance
They implied it here, but even in this world today, there are competent people that think clock speed actually matters when comparing one processor over another. I had an IT person who controls a pretty big budget actually compare a processor in a 8way Sun server to a 21264 alpha chip using only the magnitude of the clock speed as the only performance benchmark. As most (should) know, clock speed works for ranking processors within a family, but mean very, very little in the real world. It's obvious, but as long as purchasers think this is true bad decisions are being made...
-Sean
Best quote from the article: There aren't many strong brand reputations in the microprocessor business but ARM enjoys one of the best. According to their reputation, ARM's chips are endowed with an almost magical ability to run on bright sunlight or the energy released by rubbing a cat. An ARM processor, two lemons, and some copper wire are all that's needed to build the latest PDA, it seems.
Some of Slashdot's trolls would do well to pay attention to the cynical wit present in that statement. Overall though it's a very well-written, concise, and informative article. I'll be quoting his statement on MIPS next time a cpu discussion comes up.
Take a 486 processor and the latest pentium 4 and then run them for 4 or 5 years. Due to the way that modern processors are manufactured as the die features are much smaller, then there is a high probability that the Pentium 4 has now stopped working:- The 486 is still likely to be going strong.
The problem is this, it is fine having the latest technology fitted in the a piece of equipment. However, if you expect that piece of equipment to be operational for an extended number of years, don't select the most powerful processor avialable. Fit one that is reliable and is likely to still be in production for the foreseable future.
Ahh, some of you would say, surely you can emulate an old microprocessor in the future if you need to. Good premise, but if you are building for an application that is incorporated into a safety device on an aircraft or even a car, you will spend an absolute fortune re-qualifing the replacement software. An if you are only building a a few units for spares, it is one easy way of going bunkrupt.
and this is before I start talking about whether the microprocessor will operate in the temperature, vibration environment and the effects of atmospheric radiation at altitude.
Working outside the PC industry, there is a lot more to consider than you would think....
Plusser
For those in a position to select one of these embedded processors, they already know all of this.
Umm, I believe this assumption is not a good one. Lots of people working in embedded systems believe some of these myths. This article is aimed at making sure everyone in the field is aware of these pitfalls. For example, I've seen lots of CPU selector charts specify Dhyrstone MIPS to indicate CPU horsepower.
K.
The author seems to enjoy dispelling myths but, in this and one of his other articles (RISCy Business) that he links to, he seems to cling rather stubbornly to a couple of his own. For example, I just read twice about how "some RISC chips" don't have multiplication instructions, even though most do nowadays. But hey, it's a convenient club to bash with, just like the one about code density. The guy's a troll. He hates RISC for some reason, and perhaps he has some good points to justify that dislike in the embedded space, but in his zeal he just goes too far into exaggeration and misrepresentation. He also needs to read H&P to understand the real rationale behind RISC, instead of the strawman rationale he gives in the article.
Slashdot - News for Herds. Stuff that Splatters.
the wee ARM6 consumed less total energy than the others gave off as heat.
-1 Redundant
Nothing to see here; Move along.