Bluewater Systems

It has been a bit over two years since the Nvidia license was announced. At the time we were promised high definition video and high performance chips with a graphcs focus. It takes a couple of years for a new licensee to get their feet under the table, so it’s a good time to have a quick look at the results.

This video shows some of the technology at work:

It certainly looks compelling. From the specs, it can drive an large HD display as well as a Blu-ray player. Another video is here (HTC Touch Diamond) showing the Opera browser.

Nvidia has recently named their ARM line ‘Tegra’. This is a 700MHz ARM11 MPcore (multi-core CPU) with high-end graphics capability. Support on the web site looks weak, and it seems to be aimed at Windows. The HTC Advantage is one consumer product I can see with this technology. It uses Windows too. The target market seems to be tier 2/3 Asian manufacturers of electronic gadgets.

Nvidia sees its competition as the Intel Atom, but this is a strange choice. Sure, Tegra looks great against the Atom - 1/10th the size and 10th the power, they claim. But most ARM SOCs look good against the Atom. How does it stack up against TI’s OMAP3530, for example? Surely this is the real competition - no one in their right mind is going to put Atom into a small mobile device.

The focus on Windows perhaps just reflects the market they are in - as reported on Linux devices:

Asked about future Linux support, NVidia Spokesperson Andrew Humber replied, “Market demands will always dictate the direction in which we take a product, and this is true of all of NVidia’s businesses. However, at this time we are focused entirely on Windows Mobile and Windows CE.”

This is in strong contrast to both Intel and TI, who go to great lengths to support Linux on their products.

Is Nvidia persuing the right strategy here? There is no need to argue with Intel about the merits of using ARM versus x86 in mobile devices - that argument was surely won years ago. And the focus on Windows (while it has advantages in concentrating Nvidia’s limited resources into a single platform) plays into Intel’s ‘we are x86′ story, and allows people to pidgeon-hole Nvidia into a customer base with limited software expertise and little chance of producing an iPod-killed. Could Nvidia aim higher?

There are 10 types of people in this world……. Those that understand binary and those that don’t.

While developing some in-plane software, we recently had to determine the best way to do end-to-end testing of our system. Unfortunately access to an actual plane was limited, and obviously both time consuming and costly. Fortunately we discovered the hugely extensible X-Plane.

It provides a highly detailed flight environment (the scenery data requires 6 DVDs), and allows for easy writing of plugins to manipulate the environment, providing full access to all of the plane controls & environment. This allowed us to easily test our system on the desk, in a completely simulated environment, whilst still testing the entire system.

Question: How many Software Engineers does it take to change a light bulb……..?

Answer: None. Its a Hardware problem.

Recently I was working on the power supply for our upcoming SnapperDV board. I needed a small efficient power supply to generate 1.3, 1.8 and 3.3volts for TIs TMS320DM355 which is the CPU on SnapperDV. On our Snapper270 design a TPS65021 is used which is a tripple DC/DC converter with two LDOs. Something similar would be ideal. After looking at all the options, strangely the same power supply chip seemed most suitable. Current limit looked good, the power fail detection comparators can be used instead for voltage sequencing, VRTC can be used as a system monitor voltage, the nINT output can be used as a voltage supervisor, The LDO can be used to power the CPUs PLL. It actually works out to be a very good fit - even though it is recommended for a PXA270 series CPU and it is not being used ‘as intended’. Taken into account that we have these in stock given they are used on our SN270 product and it looks to be a very good proposition indeed.

Apple was the company, in 1990, which put the money up for the formation of ARM. Acorn put up the founding staff and the technology. Apple’s thinking then was that ARM would power the Newton, which would revolutionise the world.

Well we know what happened with the Newton. Looking at it now the main thing that strikes me is how enormous it was. And I can’t forget a comment made by a colleague that the handwriting recognition software on the Newton was the most efficient cache-flushing algorithm known to man.

But anyway, my point is that Apple had an associate from ARM since the first. And now there is a rumour that Apple has perhaps purchased an architecture licence from ARM, meaning that it can implement its own variant of the microarchitecture. The evidence is sketchy so far:

[Talking about DEC who implemented StrongARM based on an architectural license]

And who was the engineer that had led the Digital Equipment team that developed StrongARM under an architectural license? It was Dan Dobberpuhl, subsequently CEO of Palo Alto Semiconductor Inc. With the recent acquisition of P.A. Semi by Apple, Dobberpuhl has become an Apple employee. And perhaps now, with ARM’s disclosure, we are beginning to see why Apple was interested in Dobberpuhl and his capable design team.

[The Apple rumour mill is running hard also].

And this is supposedly the motivation:

The Apple iPhone is thought to have as many as five ARM processor cores inside it, but those processor cores are contained in multiple chips from several different chip vendors. Wouldn’t that be perfect for rationalizing into a multicore ARM architecture, if Apple chose to go down that route?

Certainly there is merit in this - it is what Nokia have done with their communicator, which, believe it or not, used to have a 386 CPU and an ARM. The 386 was for keeping your hand warm. Most Bluetooth and WiFi devices include an ARM core.

The iPhone is ARM11-based, not the latest thing but still pretty nifty. Sadly, if Apple do produce an amazing new microarchitecture, it’s unlikely we will see it in anything other than an Apple product. On the plus side, as StrongARM vanished, the prize for building a high performance ARM chip, well out in front of the industry, has not been awarded for years.

“Commodore International have announced the release of their newest product, the Commodore 64 (C64). The new player in the home computing market is based on the 8-bit MOS Technology 6510 running at the incredible speed of 1MHz. It contains 64KB of memory and runs Commodore Basic version 2.0. Graphics are provided by the highly advanced 16-colour VIC-II (Video Interface Chip II) and sound though the 3-channel SID (Sound Interface Device) chip. The introductory price for the C64 is US$595 although there are rumours that the production cost is only US$135 providing Commodore a profit margin of over 75%!”

17 Million units sold.
Peak production of 400,000 a month.
Production cost eventually dropped to ~US$25.
Best-selling single computer model ever.

What’s the right machine for this time?

Once upon a time, in a kingdom not far from here, a king summoned two of his advisors for a test. He showed them both a shiny metal box with two slots in the top, a control knob, and a lever. “What do you think this is?”

One advisor, an engineer, answered first. “It is a toaster,” he said. The king asked, “How would you design an embedded computer for it?” The engineer replied, “Using a four-bit microcontroller, I would write a simple program that reads the darkness knob and quantizes its position to one of 16 shades of darkness, from snow white to coal black. The program would use that darkness level as the index to a 16-element table of initial timer values. Then it would turn on the heating elements and start the timer with the initial value selected from the table. At the end of the time delay, it would turn off the heat and pop up the toast. Come back next week, and I’ll show you a working prototype.”

The second advisor, a computer scientist, immediately recognized the danger of such short-sighted thinking. He said, “Toasters don’t just turn bread into toast, they are also used to warm frozen waffles. What you see before you is really a breakfast food cooker. As the subjects of your kingdom become more sophisticated, they will demand more capabilities. They will need a breakfast food cooker that can also cook sausage, fry bacon, and make scrambled eggs. A toaster that only makes toast will soon be obsolete. If we don’t look to the future, we will have to completely redesign the toaster in just a few years.”

“With this in mind, we can formulate a more intelligent solution to the problem. First, create a class of breakfast foods. Specialize this class into subclasses: grains, pork, and poultry. The specialization process should be repeated with grains divided into toast, muffins, pancakes, and waffles; pork divided into sausage, links, and bacon; and poultry divided into scrambled eggs, hard-boiled eggs, poached eggs, fried eggs, and various omelet classes.”

“The ham and cheese omelet class is worth special attention because it must inherit characteristics from the pork, dairy, and poultry classes. Thus, we see that the problem cannot be properly solved without multiple inheritance. At run time, the program must create the proper object and send a message to the object that says, ‘Cook yourself.’ The semantics of this message depend, of course, on the kind of object, so they have a different meaning to a piece of toast than to scrambled eggs.”

“Reviewing the process so far, we see that the analysis phase has revealed that the primary requirement is to cook any kind of breakfast food. In the design phase, we have discovered some derived requirements. Specifically, we need an object-oriented language with multiple inheritance. Of course, users don’t want the eggs to get cold while the bacon is frying, so concurrent processing is required, too.”

“We must not forget the user interface. The lever that lowers the food lacks versatility, and the darkness knob is confusing. Users won’t buy the product unless it has a user-friendly, graphical interface. When the breakfast cooker is plugged in, users should see a cowboy boot on the screen. Users click on it, and the message ‘Booting UNIX v. 8.3′ appears on the screen. (UNIX 8.3 should be out by the time the product gets to the market.) Users can pull down a menu and click on the foods they want to cook.”

“Having made the wise decision of specifying the software first in the design phase, all that remains is to pick an adequate hardware platform for the implementation phase. An Intel 80486 with 16MB of memory, a 300MB hard disk, and a SVGA monitor should be sufficient. If you select a multitasking, object oriented language that supports multiple inheritance and has a built-in GUI, writing the program will be a snap. (Imagine the difficulty we would have had if we had foolishly allowed a hardware-first design strategy to lock us into a four-bit microcontroller!).”

The king had the computer scientist thrown in the moat, and they all lived happily ever after.

A recently released OMAP3530 based development board is now available for purchase from Digi-Key. The BeagleBoard is a small, cheap and simple development kit. Using the Cortex-A8 core, it provides a 600MHz processor (the equivalent of more than 1GHz of ARM11 MIPS), 128MB ram, and 256MB flash.

The OMAP3530 core also provides features such as OpenGL ES 2.0 capable 2D/3D graphics, and an HD capable DSP for video compression/decompression.

For purchasing information, visit the Digi-Key site: http://www.digikey.com/Beagleboard/

For more technical details, visit the BeagleBoard homepage: http://beagleboard.org/

Many years ago while at ARM I remember then CEO Robin Saxby getting us involved in his vision to sell one ARM chip to everyone on the planet every year. At the time I think we were at a few million, so it seemed an impossible goal. Nevertheless it was an inspiring thought, particularly as we knew first hand what the technology had to offer.

ARM is now shipping 90 processors per second globally, which is around 3 billion per annum. Given the rate of growth in ARM’s volumes, this is getting very close to achieving his vision.

Part of the change in recent years has been a larger focus on lower-end micros. Luminary Micro didn’t exist four years ago but now have a range of 134 ARM micros available, several priced at US$1. Other existing players such as ST Microelectronics, Atmel and Philips (now NXP) have also made efforts in this area.

It looks as if Mr Saxby’s goal will be achieved in the next few years. I wonder how they are motivating the troups at ARM these days?