Open-Silicon and the Business of Semiconductor Advancement
We are the music makers,
And we are the dreamers of dreams...
-- Arthur O'Shaughnessy (1873)
Some of us dream that we can fly. Some of us dream that we’re swimming in an endless ocean. Some of us dream we’re in high school again and can’t remember the combination to our locker or how to get to seventh period math class. In this week's episode of Fish Fry, we're talking about how we can make our silicon dreams into semiconductor reality. Taher Madraswala (CEO and President, Open-Silicon) joins Fish Fry to discuss how Open-Silicon is helping engineers realize their dreams and translate those ideas into silicon. Taher and I also chat about the trends and challenges facing the semiconductor industry today and the new developments that Open-Silicon is working to create custom SoCs for high bandwidth memory applications.
Machine Learning Platforms for AI
Last year, Jim Turley wondered why we have ranges of different processors. Now I want to bring another TLA to the processor table – the IPU (Intelligent Processing Unit). This is the brainchild of Graphcore, a company that came out of stealth mode last November with the announcement of $30m Series A funding from investors including Bosch, Samsung, Amadeus Capital, C4 Ventures, Draper Esprit, Foundation Capital, and Pitango Capital.
New PowerVR 8XE Plus Adds Guts to Previous Glory
In the automotive industry, it’s called a model-year midcycle refresh. In the fashion trade, it’s an update. Cosmetic surgeons call it a facelift. In the software world, it’s a patch, an upgrade, or a “user-focused feature enhancement.”
For Imagination Technologies and its PowerVR line of GPUs, it’s simply called Plus.
Key Xilinx and Intel/Altera Strategies Diverge
When there are only two competitors in a race, the tactics change dramatically. Winning is no longer necessarily a matter of simply going as fast as possible. In bicycle match sprints, the winning strategy is actually to stay behind, drafting the leading bike until seconds before the finish line, then catapulting past for the win with a burst of speed built in the wind shadow of the unfortunate leader. In yacht match racing, “covering” is the proven way to victory - mimicking the moves of the rival, and only rarely taking the risk of diverging in order to gain the advantage.
The high-end FPGA market has always been a match race between Xilinx and Altera. For decades, the two companies have jockeyed for position, each trying to outsmart and outrun their adversary with both technological prowess and marketing cunning.
Silexica Picks Up Parallelization and Mapping
It’s been quiet on the multicore parallelization front for a while. We looked at one example a couple times several years ago: Vector Fabrics*. We looked at the overall problem of extracting parallelism and then some advances in Vector Fabrics’ tools at the time. I include those links here because they go into some detail on the nature of many of the challenges – details I won’t repeat here.
But Vector Fabrics went out of business after they finally ran out of runway. Following a couple quiet years, I had a conversation at ARM TechCon with a company called Silexica doing surprisingly familiar things.
Micro Analog and On-Demand Power Management
He'd been here before, many times. He had procrastinated until he couldn’t put it off any longer. It had to be done. After a few cans of Mountain Dew and a giant sigh, he dove in. The experience wasn’t terrible, but it wasn’t great. I know you’ve been through it too. Choosing the right power supply for your next design can be a drag, searching through part numbers until your eyes glaze, only to come up with a solution that (most of the time) is still not quite right. In this episode of Fish Fry, we welcome Kapil Shankar from AnDAPT. Kapil and I take an in-depth look at AnDAPT 's on-demand power management solution. Also in this week’s Kickstarter Corner, we check out a new fundraising campaign called JeVois: Open-Source Quad-Core Smart Machine Vision Camera that could bring open-source machine vision to your next Raspberry Pi or Arduino project.