BreakPoint
Delusional statements in article comments in NI.com
Once in a while NI gets comments such as the following found here:
A Closer Look at 32-Bit vs. 64-Bit Windows
- May 27, 2011
Wrong!
I am sick and tired of misinformation being spread around the internet by self- appointed 'experts' in the field, who have a scarry half-truth knoledge of whats really going on!... in this article.....First of all, 64 bit operating systems will ALWAYS out perform a 32 bit system... There is a DOUBLING of data flow....Therefore a halving of speed.......Duh!..... Double the resource flow, divide the time....=faster....Why some people refuse to believe this is beyond me...the facts are out there....Mathamatically intensive programs used to require a math co-processor upgrade to perform well....Not true today...ANY of todays processors will be far more than needed for any 'math' needs Mhz are needed....not math capibilities.....So stop it!....If you don't know what you're talking about, keep your mouth shut....Get another hobby, you are obviously not very good at computers. So stop spreading your weak half-baked knoledge around. Does it make you feel good to pretend that you know something?
Now, this guy is delusional beyond belief and does not understand that 64 bit refers to the address size, which has nothing to do with how much data is moved at a time. I wonder if there should be a clarifying response in order not to confuse the readers or is it obvous to most that this guy is full of hot air? Given the high chances that he's just a troll, responding in the thread might just be what he wants.
Should NI delete misleading comments, especially if they use degrading and foul language?
(Of course his accidental backwards statement that a doubling of dataflow equates to halving of the speed reminds me of a student who wrote long ago in his report that he did signal averaging to reduce the signal/noise :D)
Well, at least we know he must feel really good since he seems to be really good at pretending he knows something. 
As to the question: that's always a gray area. Technically, the user is not violating the terms of use of the site. They're just being a jack-ass. Does that mean they should be censured? No. That's a path I don't think we want to head down. One of the unfortunate side-effects of all this "user comments" on articles is that ... you get comments like that. It's sort of like the news. I want to read the news. I don't want to hear people's comments. But news agencies thrive on user comments because, well, people like to talk. So, we have to deal with the occasional blip in order to have the freedom to talk. Or in this case, the privilege, since there's no inherent right to be able to say something on NI's web site. I don't know to what extent NI monitors these comments, given the number of articles in the NI KnowledgeBase. In this case it might be worthwhile to add a "correction" comment, and if the user starts ranting, NI can always shut it down.
It has been quite a while for me to dig down into the low-level details of OSs.
Speaking partially in defrense of that poster, I suspect there are few people still around today that have seen where an OS leaves off and the hardware begins.
Much of those comments apply to 64 bit hardware vs 32 bit hardware.
Yes in in a 64 bit OS you can address more than 4 Gig (note only 2 Gig availabel without speial switches) and comes down to addressing details.
4Gig requires 32 bits.
To push pass 4 Gig the adress bus to memory has to be pushed passed 32 bits. even if your 32 bit machine supported more than 4 Gig (by extra bits availbale to address the memory hardware) it would require two fetches from memory to get the adress info.
Running on 64 bit hardware both adress words can be fetched in the same memory read operation.
That is just one example of how 64 bit can be faster than a 32 bit machine.
Most modern processors implement a form of "pipe-lining" which a hardware implementation that attempts to fetch data from memory before the CPU realizes it nedds it. This feature leverages of the fact that instructions are often executed in assending memory address order so it you are fetching the 32 bit machine instruction at memory address 100, the data located at adress following that are also fetched and cached. (Aside: Pipe-lining will examine the intruction stream for jumps, branches etc. to abort tyhe pre-fetch).
how many memory bus read cyles are required depend on the bus width. The last processor I looked at closely was the VAX-Alpha that implemented a 32 bit machine while it bus width was 128 bits to speed up execution.
So a 64 bit machine should out-perform a 32 bit machine of equal capabilities.
This topic only draw a more distinct light on an issue I asked out-loud about 20 years ago when all of those "Crash Dump Analyzers" who could trace a stack dump back to a bad data paths module in the CPU where getting laid-off.
There few left that have stood with one foot on hrdware and the other on software and could answer the question;
"Is it hardware or is it software?"
Which reminds me about another one of those technology forgoten subjects;
"What exactly was Damascus steel?"
And since I rambling this morning let me also remind readers that the United States has lost the plans of the Saturn 5 booster and even if we wanted to, we could not duplicate it.
Personally speaking...
I am learning how to grow my own food. 
Ben
@Ben wrote:
Much of those comments apply to 64 bit hardware vs 32 bit hardware.
Ben, thanks for the details. Yes, it is the distinction between an X-bit processor in general and an X-bit OS.
(Similarly, I don't think you can find a typical 8 bit processor that has only 8 address bits, right?)
The article is about an X-bit program running on a X-bit OS, so the quoted commend does not apply and could confuse the reader.
They actually have gone to the Johnson Space Flight center in Houston and pulled parts off of the Saturn 5 on display there to measure, etc. I was astounded when I heard that, figured that these big rockets laying and standing around were all empty shells! It is tragic that so much of that amazing engineering effort is lost. I grew up during the "space race", really wanted to be an astronaut, started getting the NASA Tech new letter in the 7th grade (when Gemini capsules were the US manned space craft). By high school the astronaut "thing" was gone, I passed the height limit then in place for astronauts (they had cramped seating before the airlines figured it out, another innovation for the air industry from NASA?), and moved on. But when I saw the movie 2001 back in 1968, a "real" space station, outpost on the moon, all seemed possible, why we were only a year from actually landing on the moon, and it had taken less than ten years to go from relatively rudimentary ballistic missiles to the Apollo. So I figured "in 30 years more think were we will be! " 
As to having the plans filed away, I worked for a number of years for a major US defense contractor, building pretty complex electronic systems. When a competitor won the next round of contracts to produce that system, they were, under the government rules, given all of the "plans" for building these systems. It essentially put them out of business, because despite the mountains of schematics, assembly instructions, test plans, etc., what was lacking was the oral history that had never been capture in a formal document, where out on the shop floor the foreman would tell the new guy, "that part has to be positioned this way for it to really work". This was a fact that amazed me when I first started working there, was against the company's own policies, but was how it worked, but the competitor, who would have lost money on the first few systems even if they were able to build them immediately, wasn't able to get one built that worked in a timely manner.
So I figure that for what was described a the time as the most complex assembly ever built by man, having the plans, and the parts, only gets you about 70% of the way to having a working Saturn 5.
We probably shouldn't hijack this thread, but I vaguely remember that series. I'll also post, in a new thread, a link to a youtube of a guy making his own vacuum tubes. We could probably get back to the technology of the 1930's relatively quickly from scratch, although even acquiring tungsten for filaments (much less the thorium that more modern tubes used) might be tough. Though one of my childhood favorites "A Connecticut Yankee in King Arthurs Court" was from one of my still favorite authors "Mark Twain", who did a fair amount of his writing just down the road from here in Elmira, NY.
