THE CORE MEMORY FORUM
September 22, 2018, 02:48:56 pm *
Welcome, Guest. Please login or register.

Login with username, password and session length
News:
 
   Home   Help Search Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: Meaning of the Storage and Process  (Read 7525 times)
n8eyh with OCD-WM42 of Hilse
Jr. Member
**
Posts: 60


WWW
« on: January 14, 2008, 08:59:53 am »

Hi! All,
It's a memorial birth year of 4 decades for C-615-100 of NCR Century computer since the Press Release at March 5, 1968. I heard this date from Mr. Jim Taylor who attended to the official domo room as a member of NCR design center staff, even though I was still in a college studying at the day in Osaka, Japan.

In 1969, I had started my fresh system analyst job through learning the processing technology by the the Up-To-Date procedure of the father-son type under the B1 executive operating system.
When I made a program to fault the permanent while-looping, our coach used to suggest us to forcus into the event of each end-of-file.
However, it was very difficult for a fresh person to design the data flow procedure to be independent from the commands control flow procedure. Because in the data flow procedure we can not anticipate the sequence of each end-of-file on the multi-files. On the other hand, we could program easily the mechanism of the command control sequence with the flowcharts. But in order to handle the end-of-file, we needed to understand the event-driven architecture under the data flow.
Unfortunately we did not have any universal description method to define the event-driven system with the data flow in Japan, even though NCR Dayton had introduced the ADS (Accurately Defined Systems) for computer users in 1968. Only the computer design center of Toshiba corp. of Japan had noticed the sophisticated usefulness of NCR ADS designs approach. Today, it is easy for us to buy the book of NCR ADS through Amazon.com.

However, in NCR Japan, there had already been one person who noticed the necessity to get the effective technology to design the event-driven realtime processing and to reduce the overhead of multi-tasking.
His name is Mr. Ikuo Akiyama, former president of NCR Japan and the 1st generation realtime-online banking systems designer/implementor for the Sumitomo Bank. He had invented the way to define the event-driven technology by the multi-thread procedure with the Group-Queues definition and to reduce the 'move' instructions in computing for the task management, in order to maximize the computer execution of the algorism in late 1960s.

Generally the storage, as one of the resources, is needed by the processing. The storage would be most effective to compute, if it is used only for constructing the data. Otherwise, it would become an anchor for computing, if we copy and paste the data or decompose a task into small tasks meaninglessly.

During development of the operating software kernel from this year, I am going to confirm the usefulness of the above theory of Mr. Ikuo Akiyama in the case of the embedded software system and to transfer his idea to the next generation.

Please enjoy if you have any interests in the computer algorism and the data structure.

Best regards,
Katsuhiko
« Last Edit: January 31, 2008, 04:07:51 pm by n8eyh » Logged

Katsuhiko Hirai
Fan of the Century architecture under 63 index registers.
n8eyh with OCD-WM42 of Hilse
Jr. Member
**
Posts: 60


WWW
« Reply #1 on: January 20, 2008, 04:48:52 pm »

Hi! All,

In NCR Japan, Mr. Ikuo Akiyama had introduced the way to implement the threading under the realtime process on the NCR Century 200 system in 1970.

Generally, the more decomposition into multi-tasks causes the processor to make more heavy task switching.

In order to reduce the overhead of the task switching, he made an idea to minimize the switching rate with configuring the grouped queue. A task requests to get the new IBH (Inter-Block-Header) of message through the grouped queue that includes several queues. By this handling, the rate to encounter the emptied queue would be reduced. Then the minimized task switching used to make the processor better utilization.

  I would like to introduce another story to invent the 'Threads' on the UNIX system in the following URL:

http://www.faqs.org/faqs/os-research/part1/section-10.html

Please enjoy several histories of the software science since Dr. E. W. Dijkstra's report in 1965.

Best regards,
Katsuhiko
« Last Edit: January 25, 2008, 01:59:00 pm by n8eyh » Logged

Katsuhiko Hirai
Fan of the Century architecture under 63 index registers.
Pages: [1]   Go Up
  Print  
 
Jump to:  

Powered by SMF 1.1.4 | SMF © 2006-2007, Simple Machines LLC
Page created in 0.054 seconds with 16 queries.