My Weblog

Blog about programming and math

My experience with online learning

I am one big confused soul who is always curious for learning. I found Coursera and Udacity very helpful. I try to take as many courses as possible but recently I took the Heterogeneous Parallel Programming by Professor Wen-mei W. Hwu and it was fun course. The video lectures were excellent and assignments were not very hard. They were designed to make us familiar with basics of cuda programming and apart from that I got certificate signed by Professor Wen-mei W. Hwu :). While doing some more research about heterogeneous computing which involves CPU, GPU and FPGA all together in single system, today I found FPGA Programming for the Masses. There are two promising technologies, one from Altera and other is Liquid Metal from IBM.

Altera OpenCL-to-FPGA Framework.
Altera proposed a framework for synthesizing bitstreams from OpenCL. The framework consists of a kernel compiler, a host library, and a system integration component. The kernel compiler, based on the open source LLVM compiler infrastructure, synthesizes OpenCL kernel code into hardware. The host library provides APIs and bindings for establishing communication between the host part of the application running on the processor and the kernel part of the application running on the FPGA. The system integration component wraps the kernel code with memory controllers and a communication interface (such as PCIe)
Liquid Metal.
The Liquid Metal project at IBM aims to provide not only high-level synthesis for FPGAs,1 but also, more broadly, a single unified language for programming heterogeneous architectures. The project offers a language called Lime, which can be used to program hardware (FPGAs), as well as software running on conventional CPUs and more exotic architectures such as GPUs. The Lime language was founded on principles that eschew the complicated program analysis that plagues C-based frameworks, while also offering powerful programming models that espouse the benefits of functional and stream-oriented programming.
Heterogeneous parallel programming looks very promising but still long way to go. My certificate 😉



March 9, 2013 - Posted by | Programming | , , , , , ,

No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: