Wednesday, June 08, 2005

Parama Networks, RIP

founded in 2000, funded in 2001, finished in 2005I received my official notification today of the intention of Parama Networks to "merge" (read as fire sale) with Bay Microelectronics and, therefore, cease to exist. From discussions with former employees who were there to the end, no one from the original team will be moving along into the new company. It is a sad end to a truly marvelous company with amazing and groundbreaking technology. Of all of the things lost in the tech bubble, this one bothers me the most - and not just because I worked there for a few short years.

Aside: This post is aimed at the telecomm community. I apologize if you're not a member and find this uninteresting. Please click here for something more amusing.

Parama was originally conceived as a chip company by a group of ASIC "rock stars" who had successfully put a DVD onto a chip for use in computers (you probably have one of theirs in one of yours). They wanted to put their expertise to work in telecomm. However, their investors convinced them to "wrap some sheet metal around" the ASIC and come up with a complete system. (Note that these are the same investors who later pushed them back to being a chip company, but that is neither here nor there.) From those beginnings, a world class systems company was born.

Parama was unique in that they didn't look at the market and try to change it (like many of the optical companies), they looked at the market and tried to make it do the same things better. And the reality was - and still is - that SONET (and SDH) rules the telecommunications world. Parama took SONET and put it on afterburners. Consider some of the following innovations, now assigned to the scrapheap of history.

Fully flexible overhead access

The Parama ASIC allowed complete and total acceess to the SONET/SDH overhead. On an OC-3 that's not such a big deal, but on an OC-48 or OC-192, there's a huge amount of wasted bandwidth in the overhead. With the Parama box, you could use that overhead for some pretty amazing purposes.

For example, true transparent transport of lower speed signals onto higher speed signals was possible. Sure, other elements claim this feature, but they transfer only a small percentage of the most used overhead bytes. The Parama box would allow you to pick as many overhead bytes as you wanted and pass them through the network. We called it "Electric Wavelength Services". Amazing.

Another example is something Parama called "Ethernet Orderwire". There's a white paper on the website - read it before it goes away. Instead of just providing the user with a voice channel between remote sites, the Parama box provided a full 10Mbps Ethernet channel. It could be used for remote monitoring, file uploads, or just email and instant message connections. Customers loved it, and no other company then or since has come up with anything even close. Marvelous.

Finally, even if the customer didn't want to use the extra overhead, the Parama boxes would use it to create huge management pipes throughout the network. Instead of the puny DCC channel between elements, Parama provided a 10Mbps channel. Software download, PM retrieval, and everything else moved around the network at speeds an order of magnitude ahead of the competition. Incredible.

Unbelievably compact and low power

The Parama box was the size of a Cisco 15454, but could handle up to 12 OC-192 BLSRs simultaneously and switch between all of them. This was a result of all of the SONET capabilities being combined onto a single ASIC. No one had ever tried that before. Plus, by having the hardware, software, and chip teams developing simultaneously, each was able to fully capitalize on the others' expertise. The resulting synergy was a SONET element that was - and still would be today - much more compact and power efficient than anything else out there.

Very low cost

When we went out to customers to present the Parama box, many of them thought that we had made a mistake in our pricing calculations. In M&A discussions, other companies accused us of lying - or at least of having negative margins. On the contrary, we never quoted a price with less than 60% gross margins. And we never lost a price competition. It was never close. Again, when everything is put onto a chip, and that chip is controlled by the system designer, there is a lot of room for cost savings.

Concentration on the basics

Despite all of the innovations from Parama, the company never lost sight of the basics. Not interested in Electric Wavelength Services? Fine, don't use them. Want full Telcordia compliance for basic SONET transport? Great, that's our number one focus.

Parama started from the concept of improving what was already in practice, not trying to change minds about the future of technology. It worked, and a Parama presentation is the only one I've ever given where I went into every meeting knowing for certain that we could meet any requirement that the customer threw at us.

And so much more

SONET-based mesh (Verizon co-wrote a paper with us on that one), ports with software-provisionable rates, per-STS BLSR/TOPS support, G.MPLS provisioning across networks, and so many other innovations in software. Not to mention the overwhelming performance of the hardware. The hardware team managed to build a half-height card with all of the functionality of a 1999-era full-bay OC-192 multiplexer. The backplane carried over 2Tbps (not a typo - terabits). Most of this lost to the ages.

Regrets, we have a few

Could Parama have survived as a systems company? It's unclear. The telecom market is a tough one, and there is not as much money around as there used to be. However, other systems companies with less customer acceptance are still hanging in there. And Parama had incredible customer traction - to a level that I've never seen before or since in a small company. If nothing else, the final outcome as a systems vendor could not have been much worse than the terms of the "merger" with Bay.

I wish nothing but the best to Bay Microsystems. From what I read, their ambitions are much smaller than ours were. What was intended to be incidental business to Parama (as a systems company) will be the main focus of Bay. I hope they do well.

Most of the Parama employees have landed on their feet - eventually. I got to move back to Atlanta, where my family is happy, and I get to work in the exciting and fast-moving FTTP part of telecom now. So, everything eventually worked out in the end.

But I will forever regret what was lost.


Swami Srinivasan said...


If Parama addressed a market need, why could it not have survived?


Scott said...


Sometimes it's not a matter of meeting a market need. Sometimes it's a matter of timing.

There were many companies funded and some that even went public during the telecomm bubble that could only marginally claim to meet a market need. There were a few - like Parama - that were terrific ideas and well-received but suffered from the telecomm crash. The best ideas in the world still need funding to bring them to light.

Parama's last round of funding was required just as its investors were getting out of telecomm and just as capital spending at major telecomm companies was tanking. It would have taken some forward thinking and risk taking from investors to put up money in that market.

And the reality is that even with funding, Parama may not have been successful due to the drop in capital spending throughout the industry. But I would still, even today, be proud and willing to promote that box, that set of capabilities, and that engineering team to any customer in the business.