Whether to use a HLL depends a lot on the application. I have done many projects over the past 40 years where assembler was the best choice. Sometimes it was because of speed, many times because of lack of program space. Even the choice of processor depends on the application.
Just wish I had a Prop back in the late 70's onwards.
One of my projects used 2x MC68705P3S, 1x 4011, 1x 7404, 2x transistors and a ferrite coil. Could have done this with a Prop, eeprom, 2x transistors and coil. The first purchases of these MC68705P3S cost $150.00 each! Yes, that's what they cost for a while. 6 years later they were down to $8.00 each.
Both of these micros required very fast and specific timing. There were no uarts and the proprietary serial clocked at 17us. With 4us instructions, and resyncing mandatory, there was no time for interrupts! Assembler was the only solution here! Bit-banging was as hairy as hell.
A 4us/instruction propeller would have been a true gem. Even 4 cogs and a lot less memory. Unfortunately Chip wasn't around back then, or else in diapers
There is still plenty of new hand-crafted assembly written for the 8-bit PICs. People will keep buying the cheapest option that works.
Absolutely. While my workbench is supplied with hundreds of Cortex M0 chips from various manufacturers, there are times when some tight assembly code on the cheapest or smallest 8-bitter is the right fit.
BTW, I have some 16C54 chips here that I programmed 25 years ago. The same code now fits in a 16F54 which was made the 40th week of 2015. I imagine similar tales could be told of the 8051.
I like asm, forth, and wait for it... interrupts. Welcome to the 21st century, Prop
Had to chuckle. Not sure if you meant it as a joke or not. ASM, Forth and interrupts have been around since the dawn of time. That is certainly not the 21st Century.
Your selective editing did not go unnoticed.
interrupts have been around since the dawn of time.
Here are the key takeaways for Microchip from its billion-dollar deal with Atmel.
1. MCU Powerhouse: Microchip's pending acquisition of Atmel would move the Chandler, Arizona–based chipmaker to the third position in the MCU ranking following Renesas and NXP.
2. Manageable Overlaps: There is a lot of talk about different MCU architectures. However, in the 8-bit space, the hugely popular AVR and PIC families of Atmel and Microchip, respectively, enjoy a huge customer base. They can most likely coexist in volume embedded applications.
3. 16-bit MCU Complement: Atmel doesn't offer MCUs in the 16-bit space and that significantly eases the overlap concerns.
4. ARM Boost: The high-performance IoT designs are catalyzing a shift from 8-bit and 16-bit devices to 32-bit microcontrollers, and here the wind is clearly blowing toward the ARM ecosystem. Atmel is a strong player in the low-power ARM microcontroller market, and that can lead Microchip to new market segments.
5. Securing IoT: Security is a major concern in the IoT designs. Here, Atmel's expertise in security technologies like Trusted Platform Module (TPM) and crypto memories can complement Microchip MCUs in securely connecting objects and dealing with the key provisioning.
Atmel's on-chip SHA and AES crypto engines enable encryption of datastreams and accelerate authentication of applications and firmware.
6. Silicon Valley Outpost: Atmel is a Silicon Valley pioneer and it's tie-up with Microchip would bring the chipmaker from Chandler, Arizona a strategically important presence in the heart of chip industry's nerve center. Earlier, in 2015, Microchip had acquired another notable presence in Silicon Valley when it purchased Micrel.
Microchip's 32-bit microcontrollers use MIPS architecture and internal software tools.
Comments
Just wish I had a Prop back in the late 70's onwards.
One of my projects used 2x MC68705P3S, 1x 4011, 1x 7404, 2x transistors and a ferrite coil. Could have done this with a Prop, eeprom, 2x transistors and coil. The first purchases of these MC68705P3S cost $150.00 each! Yes, that's what they cost for a while. 6 years later they were down to $8.00 each.
Both of these micros required very fast and specific timing. There were no uarts and the proprietary serial clocked at 17us. With 4us instructions, and resyncing mandatory, there was no time for interrupts! Assembler was the only solution here! Bit-banging was as hairy as hell.
A 4us/instruction propeller would have been a true gem. Even 4 cogs and a lot less memory. Unfortunately Chip wasn't around back then, or else in diapers
BTW, I have some 16C54 chips here that I programmed 25 years ago. The same code now fits in a 16F54 which was made the 40th week of 2015. I imagine similar tales could be told of the 8051.
Had to chuckle. Not sure if you meant it as a joke or not. ASM, Forth and interrupts have been around since the dawn of time. That is certainly not the 21st Century.
Your selective editing did not go unnoticed.
Not for the prop.
finî
Exactly. By gaining interrupts the Prop is catching up with the mid, 20th Century. Hence the chuckle.
Nothing devious was intended by my quoting. Though I do seem to have dropped the word "Prop" off the end of your sentence in my quote some how.
http://www.allaboutcircuits.com/news/atmel-microchip-and-the-iot-writing-on-the-wall/