My most interesting failures, #2: The pre-ipad

Somewhere in 2000 or 2001, while working for MontaVista, I got ahold of a webpad design built around Transmeta's intel clone chip. By 2001, I had it booting from (32MB!) of flash, and running Linux, Xwindows, Mozilla and an mp3 player in 64MB(!) of ram. It had a virtual keyboard using a hack from the Xtest library. It had wifi - I used it to stream internet radio all the time, in loving memory of the Kerbango Radio. It was easy to hold in your hand and on your lap - it had a much softer edge and was lighter than the ipad, as best as I remember.

It was cool. Oh there were flaws: The battery lasted about two hours. It also cost - retail - over 1500 dollars. It was slow. It didn't have bluetooth. The onscreen keyboard was a bitch to type on. It found a limited market in the medical field, but nobody I talked to could see the potential I saw in it, if only it could do more stuff, wirelessly, and had more memory, both flash and ram. Back then 32MB of flash cost serious money.

I showed it to everyone I could find - My bosses - Sony - Nokia - NEC - Apple - Mozilla - and it went nowhere. All people could see was a slow, and very expensive laptop without a keyboard, even though I would plug in (and velcro) a logitec wireless keyboard on the back. It sat by my bedside, velcroed to the wall, for a year, before I had to give it back.

Later on we starting seeing stuff like the smart door for meeting rooms, and dynamic picture frames, but they were very specialized applications that used less general purpose hardware.

Recently I waited - with great anticipation, for the crunchpad. I was upset when I heard that they were using intel architecture - the low battery life and size of the chipset were going to hurt them - I could have told them that by using Intel architecture they were barking up the wrong tree, but nobody asked me, and in the end the project disintegrated before shipping due to internal politics. With the right confluence of circumstances they could have beat Apple in many ways.

I finally got a chance to play with the ipad last month. (I know I'm behind the curve on this, but the nearest Apple store is 1200 miles away)

First impression: It needs velcro, and it's heavy. Maybe I could hang one from a rope in the ceiling so I can watch movies comfortably, lying flat on my back. Second impression:

Slam it hard up against a wall at a good focal distance and I'd use the hell out of it; get some desk space back. Maybe it works with a bluetooth keyboard. I have one of those around here somewhere. Or maybe someone makes a dinky little usb slave to bluetooth adaptor so I could plug in ANY good keyboard, like the , and make that talk to it. I've googled for that little box - no luck. Sounds like a market segment someone (else) could address... I tried my bluetooth headphones on it, they didn't work worth a damn.

And while so many devices support bluetooth, the technology that logitech uses for their wireless keyboards and mice is much lower power and longer range, and more reliable. Why haven't people licensed that?

But the ipad is cool. I'd get one if I wasn't broke, and the screen was a little larger and it came with velcro. I expect that we'll see a lot of competitors in the next year.

Labels: embedded, failures, ipad, linux, webpad

  ¶ 5:45 PM 2 comments

  My most interesting failures, #1: Anagram architecture

Back in 1993 I came up with a paper design for a multicore CPU architecture that was strikingly similar to SUN's Niagra architecture, which came out over a decade later. My design probably shared some aspects with multicores that shipped before Niagra did, too.

I wrote about it publicly, only once, but by the time I stopped working on it, a year later, in Dec, 1993, it was well developed enough to file a few patents, had I chosen to do so, but I didn't. Someone that "knew" that the crossbar switches inherent in the design were impossible to build, convinced me I was barking up the wrong tree. (That person worked for SUN, amusingly enough)

It used an ISA roughly derived from the MIPS architecture and a stack based register set much like the transputer T800. Probably the most innovative feature from my perspective was that I used RAMBUS chips hooked up serially, rather than in parallel, and divided the memory up into as many 8 bit paths as I had pins available. Getting the RAM industry to produce sticks of ram in this segmented fashion would have been nearly impossible! Soldiering the chips directly on the buses was my answer but this put a hard outer limit on the amount of memory you could have with RAMBUS....

Still, I figured that the narrow memory buses would be a way of producing a series of CPU chips of various performance levels. I thought, also, that you could eventually produce smarter ram sticks - with cache and a MMU on board each one - if you did things this way.

The design also could do DMA at the virtual memory level, rather than at the physical addressing level. Memcpy was basically an op-code that suspended that CPU and handed off the operation to the MMU. This took best advantage of the slow memory buses because it could do burst transfers - so common in the message passing operating systems in academia at the time.

I called it the Anagram Architecture because, being stack based, and highly threaded, you could rewind and restart various sorts of instruction in a super-scalar fashion, and run more jumbled up processes in parallel, more simply than you could with a classic Van Neuman architecture.

I don't know when the crossbar switches I envisioned became buildable, but they are everywhere now. I've dabbled with hardware design ever since, and played with a few FPGAs, but ideas like 32 bit and 64 bit memory busses are firmly embedded in everything, and developing a competitive CPU from scratch a pointless, expensive, exercise. Not that that stops people. Two interesting CPUs I've seen recently are the Propeller and David May, creator of the Transputer, is working on Xmos.

Labels: failures, patents

  ¶ 5:32 PM 0 comments