x
Routing

Pluris Shutdown Confirmed

Core router maker Pluris Inc. has shut down amid financial turmoil, several of its former employees say. The company held an "all-hands" meeting on Monday morning in a vacant building adjacent to its locked offices to break the news, sources close to the company say.

Pluris announced that it had closed a $53 million funding round in February and, at that time, boasted that it had raised some $215 million since its inception (see Pluris CEO Bolts and Washout Rains $53M on Pluris). Now, however, several close to the company say that Pluris only received about half the money it was supposed to get during its fifth funding round before investors decided to pull the plug.

Pluris executives have not returned calls seeking comment. Certain Pluris board members have agreed to talk on the record later today.

As part of its most recent financing, Pluris reworked its entire capital structure to entice investors and employees to stick around. One month later, however, Pluris CEO Joe Kennedy left the company and Pluris has been steadily thinning its ranks ever since (see Pluris Hangs On).

For a brief period, Cynthia Ringo, the former CEO of CopperCom Inc., took the top job at Pluris. All the while, though, Ringo and the company made it clear that she wasn't the permanent replacement for Kennedy.

Pluris first announced its product in the spring of 1999 but then went quiet for several months while it frantically built the product it had already hyped (see Pluris Is Back). In June 2000, then-CEO Joe Kennedy admitted that exposing itself too early was a mistake. "The programmers were still writing code when they launched the product," Kennedy told Light Reading. "Then we had nothing else to announce. We couldn't announce that our last announcement was wrong."

The company announced it had closed a funding round of $100 million in November 2000. At the time, it believed that amount to be enough to see the company through to an initial public offering (see Pluris Preparing for Its Public). In February 2002, the company reorganized its capital structure and expected to receive $53 million in new funding from its investors.

As part of its capital reorganization, Pluris sent its shareholders paperwork about two months ago, informing them that it was executing a reverse split, according to one shareholder who asked not to be identified. A reverse split occurs when a company reduces the number of outstanding shares and increases the price per share proportionately.

Pluris shareholders were taking part in a 20-to-1 reverse split, according to the shareholder. This means that for every 20 shares they owned prior to the split, they would have one share afterwards.

Adding to Pluris' financial strain were several factors, say those close to the company. For one thing, the company had fallen out of favor with Mission West Properties, a Silicon Valley landlord, over a property dispute. Mission West sued Pluris in June 2001 and the two companies later settled.

There are also unconfirmed rumors that IBM Corp. (NYSE: IBM) required a hefty upfront fee -- upwards of several million dollars -- to produce the custom chips that would power Pluris' core router products.

As if that weren't enough, carriers who typically buy core routers (well, those that have survived) are having financial worries of their own. Pluris joins core router startups such as Coree Networks and IPOptical as another casualty in an overcrowded market.

In the past few months, former employees say the company tried to sell itself to at least two competitors -- Procket Networks Inc. and Avici Systems Inc. (Nasdaq: AVCI; Frankfurt: BVC7) -- but neither were interested.

Eventually, the uncertainty of Pluris's future began to take a toll on the more than 100 employees left. "After the last round of layoffs, people had pretty much given up hope," says one former Pluris engineer. "We were just going through the motions."

— Phil Harvey, Senior Editor, Light Reading
http://www.lightreading.com
<<   <   Page 2 / 5   >   >>
Outsider 12/4/2012 | 10:07:49 PM
re: Pluris Shutdown Confirmed Any startup or even public company who is looking for an CEO should contact Joe. The fact that he was able to receive round after round of major funding for a company that never really had a chance of competing against either Cisco of Juniper, even Avici too is absolutely amazing.

Also lets be honest, even though Pluris had some very nice engineers they were not the most talented ones in the Silicon Valley. Over the years the best ones went to other startup companies. Maybe these engineers saw the writing on the wall during their interviews, but Pluris had a very high number of offers rejected by potential candidates.

Joe whatever company you are a part of next, I want to be there.


Outsider
cyclical 12/4/2012 | 10:07:49 PM
re: Pluris Shutdown Confirmed First of all Pluris is a JUNK company in executing plans. Always slip in plans and never convinced any body they have a good product.
The Engineering is managed by idiots. Idiot bring idiots, idiots promote idiots. so on..

Any body can believe how bad their product is by not selling their IP yet.
It's shame on the company by taking all retirement money from mutual funds. All the top should be tried before court under Bush's new plan.
Marmaduke 12/4/2012 | 10:07:47 PM
re: Pluris Shutdown Confirmed How long where you there ? Assuming you're not an idiot, why did you stay there if you knew that "The Engineering is managed by idiots. Idiot bring idiots, idiots promote idiots", etc... ? And if it took you years to find out, then maybe my assumption is wrong.

Every company (startup or not) is weak in executing plans and managing things because everyone pushes itself to it's limits (specially startups); no one settles for simple plans... because these can't be financed!

I'm always surprized how much failure is always someone elses fault... "IF" it's the case, and assuming you are not an idiot and assuming you learn from your mistakes, then you will avoid similar situations in the future and find a Utopian Company.

When you do, let the others know... I'm sure many people feel that the company plan or management isn't as good as they would like.

As for myself, I stopped looking for Utopia a while back... I'm the best person to make my environment better; I can't wait for the others to do that. I run much more on what I can bring to the table for a given set of cards than on the limits imposed by others.

Marm
netchap 12/4/2012 | 10:07:39 PM
re: Pluris Shutdown Confirmed Isn't Procket going to be a resounding success
story?

ExPlurisEmployee 12/4/2012 | 10:07:39 PM
re: Pluris Shutdown Confirmed Dear fabric,

your post only shows that you either never worked with Mr. Antonov, or were on the receiving end of his "go away you idiot". I bet my employee # is smaller than yours :-)

First of all, most people think that the early Pluris box somehow was intended for production. In fact, it was a proof-of-concept prototype for experiments and for software development (and for convincing carrier people that all performance claims for this technology were real). Its design was brutally simple and efficient: Pentium-based routing engines (which were shown to handle 120kpps, about the same as then-biggest Cisco box, 7000 with SSE) and mixed optical/electical 256Gbps fabric (using 1Gbps links). It had two custom PCBs (the fabric card and the backplane), had one (one!) kind of FPGA in it, and no custom silicon whatsoever (compare that to a Cisco box). If it were somehow productized it could undercut 7000 1:3 cost/performance wise. The choice of butterfly switch was dictated solely by its simplicity. If you can call that kind of design "incompetent", we have different definitions of what competence is.

The fully-functional prototype software was demonstrated (including all routing protocols, distributed routing, CLI, SNMP, and some hot-swap support). The hardware was fucked, mostly courtesy of the former Director of Hardware, George S., who managed not only to choose FPGA barely big enough to fit a PCI core, leaving no space for the actual logic), but also had "final" PCBs made when it was clear that they'll have to be reworked. He was eventually fired, but the time was lost. That was subsequently used against Mr. Antonov, whose strong (or even arrogant) opinions and complete disrespect to corporate politcs made him less than popular with other managers.

I participated in some of the "production box" design meetings; I remember he was strongly against extensive use of custom silicon (long cycles and supply problems as Pluris later run into with IBM), or going directly after 10Gbps. (In retrospect, I think he was right because with IP bonding the speed of trunk interfaces doesn't matter much and price/performance-wise OC-24c transport is still better than OC-48c since it is significantly more tolerant to various kinds of dispersion. That OC-24c routing can be done in off-the-shelf CPU was demonstrated by BBN folks in 95). The choice of VRTX (in place of the original FreeBSD) was also a serious mistake. If I remember correctly, Mr. Antonov was ousted when the production design was still in the early discussion stage.

Considering the claim that he invented most of the Pluris technology, I would like to ask you to provide examples of significant differences between his concept and the final device. Note that the oft-quoted difference between butterfly and the later fabric design is irrelevant; butterfly is simply a way to implement a hypercubic network. The use of hardware routing was by that time an industry standard practice, so the later Pluris folks cannot claim they invented that feature. Note that I was talking about inventions, not about implementation or detailed design.

Mr. Antonov pioneered use of serial communications (and optics) in fabric, virtual routers, IP bonding, and parallel routers in general. I think you can still find on the Web his original papers (which, by the way, remain the only concise and ungarbled explanation of the rationale behind Pluris design), and check the USPTO database. By the way, I think his insistence on doing things in software had serious reasons, as he was talking about having backbone routers to do more intelligent things like transparent content cacheing instead of multicasting, and some other stuff I didn't really understand ("liquid routing" as an alternative to MPLS and TE in general, something about proactive flow control, etc). As far as I know, he's not in telecom anymore.

To put the record straight, if you want to call him incompetent, I would agree that he was a very inexperienced manager, and didn't have a good CEO to back him. I think his fault was that he completely mismanaged relations with V.C. and allowed politicos to take over.
inlight 12/4/2012 | 10:07:38 PM
re: Pluris Shutdown Confirmed Marmaduke,

Wow! That's wonderful! you do whatever you want in your org! P........lease let us know about your company, may be we want to join it ... may be NOT!

IL
skeptic 12/4/2012 | 10:07:37 PM
re: Pluris Shutdown Confirmed By the way, I think his insistence on doing things in software had serious reasons, as he was talking about having backbone routers to do more intelligent things like transparent content cacheing instead of multicasting, and some other stuff I didn't really understand ("liquid routing" as an alternative to MPLS and TE in general, something about proactive flow control, etc).
-----------------

The problem is that those serious reasons he had,
while they have merit as ideas, were/are
considerably in front of what real customers
were looking for. There is a careful balance
in any startup between innovation and getting
the product done. And if you let things that
don't exist (liquid routing, content caching
in the core) drive the design, you can easily
make the wrong design decisions.

I do think that innovation is a wonderful thing
and that designers should push the envelope
in new products, but the art is in picking
the right number and type of innovations.

lob 12/4/2012 | 10:07:34 PM
re: Pluris Shutdown Confirmed skeptic -

I think flexibility never hurts if it doesn't cost arm and leg :) Existing routers are too rigid in what they can do, and being able to change forwarding path with software upgrades is rather attractive (Pluris marketing was heavy on long useful life of the boxes, so that'd fit). Heat dissipation would be a killer, though; but you cannot beat simplicity/TTM of software routing.

You are, of course, right regarding the art of the design.
xoip 12/4/2012 | 10:07:31 PM
re: Pluris Shutdown Confirmed can somebody elaborate on the failures observed during trials - a post mortem will be great. was it bad software, bad hardware or surprises.
XoIP
fishbone 12/4/2012 | 10:07:31 PM
re: Pluris Shutdown Confirmed I agree when you talk about Joe being a great CEO
but can you explain why such a great CEO would choose to run a company staffed with only "NICE" engineers? Joe must have been a genius to convince VC's to dump millions of money into a really "NICE" company. How would you know how many offers were rejected by potential candidates unless you were a part of this "nice" but untalented group of people?
<<   <   Page 2 / 5   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE