Okay, well I studied up on my multi-core and hyper-threading technologies and thought about how that will take part in the 128-Bit processor. You might think that I would suggest that more cores and threads should be added. No, I don’t! That technique, while it has worked great in the past, has made programs more difficult to program and some programs do not work any faster because they don’t support the hyper-threading technology. So where does that leave the future of the 128-Bit Processor?
Let’s go through the parts of the CPU and see where our changes need to be. The ALU (Arithmetic Logic Unit) is where the calculations take place in the CPU. Where does it get these numbers? Well, input from the mouse and the keyboard go into the Control Unit and that data or input is moved from the Control Unit to the ALU. Input, or data as it is now called, is calculated by the ALU and sent to the Control Unit then to the Memory Unit where it is stored until the ALU can access and process all the data. Then once the data is processed, or all the registries are full, The data is sent to RAM.
Now what is wrong with this picture? It’s the amount of memory your Control Unit can intake at once. It determines the amount of input that can be processed at any one time. Currently your CPU Memory Unit address sizes are 39 bit physical and 48 bit virtual. According to web.standford.edu a bit can only store a 1 or a 0. That means 8 bits make a byte and 8 bytes can store numbers between -9223372036854775808 and 9223372036854775807. So then it takes two storage addresses in the Memory Unit to store this number for calculation which would be used in calculations to the moon and trigonometry with a decimal in it. Or something like that!?!
It would seem that storing a number in memory divided into two segments (if we are only that lucky) would take more time and processing power. And that is just a math number, not a program like Microsoft Word! I couldn’t imagine how many registries are used to store MS Word or a picture. In electronics, the trend has been to get smaller but they changed that in phones. Some are getting a little bigger and I think Processors should go that route too. How, you ask? Well you have SoCs or Systems on a Chip that are computers the size of a credit card like the Raspberry Pi 4. So we can create the 128-bit processor with a bigger design in mind making it easier to increase the ALUs and work on programming the CPU Memory Unit so that each registry holds more data per address.
At this time I would like to talk about the CPUs cache memory which stores frequently used programs for the CPU to easily access. The L1 cache stores about 32MB of memory which L1, L2, and L3 cache can increase in size but to make that efficient you would have to make the works inside the CPU store more in each registry. All these techniques will create a better processor to run the current software programs without having to rewrite them. Having multi-core and hyper-threading 128bit processors will happen! The thing about it is to create a processor that runs the current software faster and better without rewriting the programs or code of the software.
Thank you for your time and bless you with your future plans. Just remember that when the 128-bit processor comes out you heard about it here first. Please refer your friends and family to my website to ask questions from the contact form. My website is still small and I take the time to read all the emails from my contact forms. If not and you just want to listen, join the email list. You will only receive emails when I post a new blog post. Have a Blessed Week and remember to check back in a week or so to see my next blog post that reviews and extends the study of the Snap Circuits Extreme Kit Project #3.