FPGAs: What makes them programmable?

It sounds crazy when you say there is a device called FPGA that might replace ASIC. Definitely it will be a great achievement for the electronics engineers if they can make a hardware which can be used for all the electronic devices. Field Programmable Gate Array (FPGA) specifies that  it can be configured by costumer or designer after manufacturing. Hardware description languages (HDLs), such as VHDL and Verilog are used to implement the logic.

FPGA

FPGAs have mainly four elements viz. CLBs, Interconnects, BRAMs and IOBs which makes them programmable.  Configurable Logic Blocks (CLBs) are the main weapons to make device programmable. CLBs are made up of LUTs, flip-flop and a 2:1 mux. 

FPGA

Look UP Tables (LUTs) are made up of small BRAMs that can implement any logic function. Each of inputs is essentially an address for the BRAM so that it can define the output logic to the block. 2 inputs LUT can become any logic gate. LUTs are then connected to the DFF which are the memory elements used to store the output from LUT. A 2:1 MUX is used in case you want to bypass the DFF (D-Flip-flop). Each logic block can connect through the Interconnects to their corresponding block to implement the desired logic. IOBs are used as an interface to the outer world. 

Author - Dharmendra Kumar

(Research Associate at Silicon Mentor)

 

From Brick sized transistor to a pin head sized SOC. What NEXT?

In 1947 the first transistor was developed at Bell labs. That was the foundation of solid state electronics as this invention expanded the possibilities in the design of analog circuits and some logic functions which were not possible with bipolar diode. A few years later Jack Kilby and Robert Noyce (Co-founder of Fairchild Semiconductor and Intel) built upon the foundation placed by Bell labs. The two researchers worked independently and fabricated integrated circuits with a couple of transistors. These bulky circuits proved the concept of integration of multiple devices on a single piece of silicon. Semiconductor industry took the invention seriously and worked on the concept to realize integration of devices at Small scale (SS), Medium scale (MS), Large scale (LS), Very large scale (VLS) and ultra large scale (ULS) progressively with evolving technology and fabrication techniques. The evolution took decades of learning and huge knowledge base of semiconductor physics and material science. The rate of evolution observed in semiconductor technology is even higher than the predictions under “Moore’s Law” by Gordon Moore (Chairman of Intel). Now we are at the stage where Ultra large scale integration has enabled us to fabricate millions of transistor on a single chip. The results of ULSI are System on Chip (SoC), Network on Chips (NoC) and Stock IC’s. Now we are heading toward a minimum feature size of 5 nano-meter (nm), which results in a channel length of a few atoms.

What is the future of semiconductor industry below 5-nm?

What NEXT?

Author - Krishan verma

(Research Associate at SiliconMentor)

transistor technology

                                    

FPGA Vs ASIC

FPGA- Field Programmable Gate Array    |    ASIC- Application Specific Integrated Circuits

This has become the most prominent discussion in VLSI industries and amongst VLSI engineers that if FPGAs can overtake the ASICs or not. FPGA is considered as a substitute of the ASIC and various research and development teams are working on proving it, but certain parameters are kept into consideration before making such comparisons in the technology i.e. performance , flexibility , power consumption and cost of the device.

If we talk about an FPGA then there is no doubt that it’s a revolutionary product if it can be proved an alternate to the ASICs as it has faster time to market because it is independent of the manufacturing steps like masking and layout etc. One just needs to burn the codes on readymade FPGA kits and we are done.

Apart from that there are some more factors on which FPGA beats an ASIC viz. reusability, design cycle, logic density and field programmability. An FPGA can be reprogrammed at no cost basis and a second is more than enough to reprogram an FPGA but an ASIC can take thousands of dollars and a bunch of time to reprogram or it cannot be reprogrammed.

There are a number of facets where FPGA lags behind an ASIC such as power consumption, limited design size, not fit for the production in bulk and most importantly the cost of an FPGA kit is much higher than an ASIC which helps an ASIC to put behind an FPGA as cost is the most prior and vital factor for an electronic industry.

Now, when the discussion goes round about then if isolated the manufacturing steps and cost then an ASIC would be of great use and vitality for the VLSI industry and for whole electronic industry. ASICs can be produced in bulk and operates at very low power compared to FPGAs and due to design flexibility of an ASIC it gives an enormous speed than the FPGAs, but the tools required for the designing of ASICs are available at higher or mammoth cost.

The fact cannot be neglected that even if the FPGA replaces an ASIC then what really is an FPGA made up of?

It is only an ASIC which is comprised in an FPGA kit. So, without an ASIC we cannot even think of the FPGAs. Every technology has its own merits and demerits but the real technology is what meaningful we extract from that particular technology that can be used for the welfare of humans and others.

Author - Aman Jangra

(Research Associate at Silicon Mentor)

FPGA Vs ASIC

MATLAB PROJECTS RESEARH GUIDANCE AND TRAINING IN DELHI-NCR

MATLAB refers to the Matrix Laboratory which includes the math works with an integration of computation, visualization, and programming or MATLAB can be expressed as an interface between user and algorithms. MATLAB has become a tool of preference for the industries and academia due to its vitality in the electronics and related facets.

Most of the B.Tech colleges and universities do not emphasize on the MATRIX laboratory due to lack of MATLAB acquaintance with them and as a consequence of this the students gets deprived of the MATLAB concepts. After the completion of B.E/B.Tech students finds themselves running after the various institutes and training centers for the MATLAB training and other certified courses in MATLAB and same is the case with M.Tech students. However it becomes worse for the M.Tech students who are assigned with the MATLAB projects as an academic project for the fulfillment of university criteria.

So, the importance of MATLAB cannot be neglected when it comes to the engineering and instrumentation. Apart from the academics the students can opt for the training and research guidance on MATLAB projects and IEEE MATLAB projects along with the conceptual and advanced tool training at SiliconMentor which is known for its quality of work in the Delhi/NCR and entire northern INDIA.
SiliconMentor not only directs and assist the students in MATLAB projects but also advancing in VLSI domain of ECE and other sub-domains viz. Analog Design, Digital Design and Verification, Antenna and Digital Communication etc and also for the PhD research projects in VLSI and MATLAB. We SiliconMentor team delivers the work with full justifications and try to forward the work of associates up to patent level and gives an opportunity to the students and professors to own the patent of their work

Author - Aman Jangra

(Research Associate at Silicon Mentor)

MATLAB Projects Training

 

 

 

Tags : VLSI Projects Training , M.tech Projects Training, MATLAB Projects Training , VLSI Backend Training Institute

Static Vs Dynamic Logic Circuits

 Comparison Between Dynamic Logic Circuit & Static Logic Circuit

• Dynamic logic circuit requires small area in comparison of static logic circuits.

• Dynamic logic circuit generates smaller parasitic capacitance, due to the smaller area on
   silicon chip.

• These dynamic circuits has smaller dynamic power than static design because the less
   parasitic capacitance.

• The common clock system in dynamic circuits in all over the circuits enables the
   synchronization.

1) Static logic Gates: The output of these logic gates depends on its applied input levels.

The output can be keep as long as the power is supplied to circuit.

2) Dynamic Logic Gates: The output operation of these gates depends on the temporary

storage of the charge in the load capacitance. This stored charge does not stay for

indefinitely, it updates with the different voltage levels.

  

Author- Dinesh Chauhan

(Research Associate at Silicon Mentor)

Dynamic Logic Circuit

Static Logic Circuit

Design Of an Echo Canceller using adaptive filter with LMS technique

In this particular project, we had introduced a concept of LMS (i.e. least mean square) in the adaptive filter designing. And finally, this particular adaptive filter is used in our communication system such as to cancel the generated ‘ECHO’, during process of communication between transmitter and receiver. It has been observed that this particular “Echo canceller” is very successful in reducing the undesired signal beside the upcoming speech signal form the transmitter end to the receiving end.

Beside the adaptive filter using the least mean square technique, one more component is also used within our design i.e. Vocoder. All these above discussed architecture is designed on MATLAB. 

Author - Rahul Bhardwaj

(Research Associate at Silicon Mentor) 

Acoustic echo cancellation setup using an adaptive filter