Template talk:Parallel computing

Notifications

Please don't add stream procesing onto Parallel Computing. Stream Processing is a method of procesing warp (aka threads) base on / originated from event-driven orientated programming, these technologies are thus use on some GPU processors, and they are more concerned with the Kernel Engineering and Scheduling mechanism of the microarchitecture of a CPU than directly related to the Parallel Computing Template. --Ramu50 (talk) 18:21, 22 October 2008 (UTC)[reply]

Regarding Hardware section

I am wondering should we decrease the amount of links in the Hardware section, because most of the technologies present in Hardware are directly related to Processor Technologies and if we expand it the template will be overloaded. Also when adding Programming section, I think we should try to add only navigation articles and not expand every single one of them. --Ramu50 (talk) 18:23, 22 October 2008 (UTC)[reply]

I think all the links currently under Hardware are reasonably relevant to parallel computing, except maybe Vector processing. Letdorf (talk) 21:35, 22 October 2008 (UTC).[reply]

Vector processing is most certainly a parallel computing topic. Raul654 (talk) 21:36, 22 October 2008 (UTC)[reply]

Questions

Is Preemptive Multitaksing part of parallel computing. --Ramu50 (talk) 22:18, 20 October 2008 (UTC)[reply]

No. It's an operating system technology, not a CPU / computer architecture technology. Completely unrelated. Georgewilliamherbert (talk) 20:07, 22 October 2008 (UTC)[reply]

I guess we should we delete Computer multitasking from the navbox then? Letdorf (talk) 21:26, 22 October 2008 (UTC).[reply]

Since the template has a lot too technical CPU technologies that aren't even relevant to Parallel Computing at all. Things like NUMA, COMA, distributed memory, shared memory, we should just shorten up to Coherency and each of technical details be navigated under the subsection of the article. What you guys think. --Ramu50 (talk) 02:07, 23 October 2008 (UTC)[reply]

What are you talking about? Everything you just named is related to distributed computing. Raul654 (talk) 02:10, 23 October 2008 (UTC)[reply]

Those are all relevant to parallel computing. Why would you think they weren't? Georgewilliamherbert (talk) 02:11, 23 October 2008 (UTC)[reply]

Ramu50 - can you do me a favor? Please go read Pfister's "In search of Clusters", and get back to us... Georgewilliamherbert (talk) 02:13, 23 October 2008 (UTC)[reply]

What make you thinks that all other have a less knowledge than you to even begin with, load of crap on bias contributions attitude. --Ramu50 (talk) 02:24, 23 October 2008 (UTC)[reply]

You are not showing evidence that you're familiar with basic college student level conceptual understanding of several technologies you're trying to classify here. It is not safe to be trying to organize things you do not clearly understand. Either there is a concepts / understanding / education gap here - in which case you need to go spend a while studying, or there's a horrible communications gap here. If you have studied the topics and have the educational background, please explain so. If you have not, please stop editing templates in these fields. Georgewilliamherbert (talk) 02:58, 23 October 2008 (UTC)[reply]

And do us a favor and keep that bias comment to yourself before spreading more of your own mental problems upon others. My expertise of knowledge is totalyl none of buisness, if you got a problem contribs at my talk page for later discussion. By the way if one mistakes account for all the reason to judge one person's every deeds of an entire life of that closeminded thinkings you mine as well not participate in a discussion at all. --Ramu50 (talk) 03:27, 23 October 2008 (UTC)[reply]

The first sentence there is a personal attack, which is against Wikipedia policy, about which you have been repeatedly warned.

The second sentence is entirely my and everyone else's business. You are contributing and trying to classify highly technical, complex articles in Wikipedia and making contributions that appear to be destroying or confusing information. This is disruptive and could be called vandalism. I assume that this is not intentional, that you just don't understand it, but your resistance to listening to many many people pointing this out is stretching credibility now.

The third sentence is not parseable. Whatever you were trying to say, the sentence was not properly constructed english and is not understandable. Georgewilliamherbert (talk) 03:37, 23 October 2008 (UTC)[reply]

I think he was trying to say: "By the way, if you judge someone's entire life from one mistake, you are close minded and might as well not participate in this discussion at all." Raul654 (talk) 04:00, 23 October 2008 (UTC)[reply]

That was the closest I could come to parsing it as well, but there has been no one mistake in the edit dispute histories and it didn't make sense in local or global context. If that's what he meant, I am not clear what one thing he's referring to. Georgewilliamherbert (talk) 04:13, 23 October 2008 (UTC)[reply]

That is totally not true, Flynn's Taxtonomy, MIMO, SIMO, LIFO, FIFO...etc are all methodologies of processors microarchitecture implementations technologies, the fact that Parallel Computing, Distributed Computing, and Grid Computing utilize those technologies doesn't prove they are parallel computing technologies at all, even though the industry say so. What type of technologies you computational technologies you utilize from other types of sciences is a personal opinion or a corporation opinion of how a product is being built. I think too many people is mixing it up, I just look through some of the talk page and there has been a lot of problems on the talk on LIFO and FIFO.

Well if my above statement is wrong, then consequently anything that is a microarchitecture arrays design such as should also be Parallel Computing which is not true. Multiplexing

SISO (Single Input, Single Output)
SIMO (Single Input, Multiple Outputs)
MISO (Multiple Inputs, Single Output)
MIMO (Multiple Inputs, Multiple Outputs)

--Ramu50 (talk) 02:21, 23 October 2008 (UTC)[reply]

Ramu, for the love of god, read the Parallel computing article before recommending we remove links from this template that obviously belong. (I wrote parallel computing article, it's completely technically accurate to the best of my knowledge, and it's a featured article). You can't "do" Flynn's taxonomy - therefore, it is not a methodology. It's a system of categorization, based empirically on the kinds of technologies that were prevalent in the 70s when he devised it. And it is most certainly related to parallel computing - it's the pedagogical basis for teaching parallel computing. I learned about it during my first week in my parallel architecture class. That's what it's relevant here. Raul654 (talk) 02:52, 23 October 2008 (UTC)[reply]

Yes, Flynn's taxonomy is one of the most fundamental concepts of parallel computing theory. NUMA, COMA and distributed/shared memory are also very much parallel computing concepts. On the other hand, LIFO and FIFO are very basic concepts in computer science and don't belong here. MIMO and SISO are terminology used in RF electronics, so I don't know why you mention them here. Maybe you are confusing them with MIMD and SIMD?! Letdorf (talk) 09:43, 23 October 2008 (UTC).[reply]

Ramu50, as I mentioned on the CPU Technologies talk page, you misunderstand what Flynn's taxonomy is about. It is not describing "technologies" that a designer might use or not. It is a classification system: You can work on one or multiple instruction streams, and you can work on one or multiple data sets at a time. That gives four possibilities and there aren't any others; every CPU or multi-CPU complex must fit into one of these. Flynn's taxonomy is a useful model for analyzing and describing any CPU. Jeh (talk) 16:17, 23 October 2008 (UTC)[reply]

Distributed shared memory?

Should the navbox have a link to distributed shared memory? Rilak (talk) 06:13, 27 October 2008 (UTC)[reply]

But it does already! Letdorf (talk) 11:00, 27 October 2008 (UTC).[reply]

Oops! So it does! Would it be more appropriate for it to be under "Memory" with the other memory-type links? I guess I didn't see it because it wasn't there. Rilak (talk) 05:23, 28 October 2008 (UTC)[reply]

Yes there's definitely scope for tweaking the categories in this template for the sake of clarity. Letdorf (talk) 13:07, 28 October 2008 (UTC).[reply]

PVM?

Should the navbox have a link to Parallel Virtual Machine? 85.124.239.162 (talk) 15:31, 27 August 2009 (UTC)[reply]

Fair point; done. Letdorf (talk) 12:55, 28 August 2009 (UTC).[reply]

Thread safety

I was thinking: should this template have a link to thread safety? --Adamd1008 (talk) 01:33, 23 August 2010 (UTC)[reply]

Add distributed frameworks

I am not sure if distributed frameworks should be added to parallel computing. Distributed frameworks such as Apache Hadoop and Apache Spark could be deemed useful in this template page. Please let me know or scribe in my talk page about your thoughts. -- LokeshRavindranathan 05:13, 16 February 2015 (UTC)[reply]