Archive for November, 2007

Caltech Scientists Set Data Transfer Record

Thursday, November 29th, 2007

The California Institute of Technology just announced today that its scientists have set a new record for sustained data transfer, showing how long distance, wide area links can be used for clustered supercomputing. The record 80+ Gbps sustained data transfer was demonstrated at the SuperComputing 2007 conference. According to Caltech:

Multigigabit/s end-to-end network performance will empower scientists to form “virtual organizations” on a planetary scale, sharing their collective computing and data resources in a flexible way. In particular, this is vital for projects on the frontiers of science and engineering, in data-intensive fields such as particle physics, astronomy, bioinformatics, global climate modeling, geosciences, fusion, and neutron science.

Essentially, this kind of technology is useful for distributing difficult computer applications across the Internet. It’s interesting to note the heavy influence of technology firms here in Southern California in contributing to the effort. According to Caltech:

The record-setting demonstration was made possible through the use of seven 10-Gbps links to SC07 provided by SCinet, CENIC, National Lambda Rail, and Internet2, together with a fully populated Cisco 6500E series switch-router, 10-gigabit Ethernet network interfaces provided by Intel and Myricom, and a fiber channel disk array provided by Data Direct Networks equipped with 4-Gbps host bus adapters from QLogic. The server equipment consisted of 36 widely available Supermicro systems using dual quad-core Intel Xeon processors, and Western Digital SATA disks.

 

The setup includes hardware from Arcadia-based Myricom; Chatsworth-based Data Direct Networks; Aliso Viejo-based QLogic; and Lake Forest-based Western Digital. The whole demonstration was run on the FAST TCP protocol, which is from Caltech and is the basis of Pasadena startup FastSoft.

When you don’t want to be found: the problem with social networking sites

Wednesday, November 28th, 2007

There’s a great article by Cory Doctorow (of BoingBoing fame) in Informationweek, talking about how annoying/creepy co-workers (and the like) are the reason why social networking sites are so volatile.

That’s the problem with Facebook, LinkedIn, etc. — there are lots of people, and reasons, you don’t want to be connected into a social network. I notice — with some exceptions — that here in Southern California very, very few venture capitalists, investment bankers, or other “power players” are connected into any social networks. Not to mention if you’re a celebrity (aside from the “official” fan profiles on MySpace). Why? Simply, you don’t want people knowing what you are doing at every moment of the day; you don’t want everyone and their next door neighbor pitching you for deals; you don’t want people taking cycles out of your already overloaded schedule.

Everyone knows who I am talking about: we’re not talking about college buddies and good friends, we’re talking about the sort of scary guy who keeps coming up to you at industry events trying to get any venture investor to buy into his scheme to use ESP to contact aliens; the would-be entrepreneur who (violently) won’t take “no” for an answer; that wierdo you met back in college and thought you’d be rid of.

It might not even be a good idea for folks you do know, and perhaps even trust, to know exactly what you’re doing all the time. Do you want to let everyone know you spent last week in New York interviewing investment bankers? Or that, as a VC, advertise to the world that you are about to make an investment in a firm which you think is the “next Microsoft” — before the deal is done? Or, do you want to know prospective investors that you spent the weekend getting drunk with your college buddies?

Shopzilla for sale?

Tuesday, November 27th, 2007

Here’s an interesting item: Silicon Alley Insider is reporting that Scripps is trying to sell Shopzilla, which it purchased in June of 2005 for $525M. According to the Insider, Scripps is hoping to get somewhere between $500M and $600M for the company.

If I don’t return your email…

Tuesday, November 27th, 2007

This might be why:

I recently began running my email through Google’s Gmail (in addition to at least two other layers of spam filtering), in an attempt to better filter the stream of email I am getting. You’ll note those 14,155 pieces of spam all are from the last two weeks of email, and don’t include all the spam email already filtered out before hitting Gmail.

The venture/angel gap

Tuesday, November 27th, 2007

There’s been an interesting trend over the last year or two in Southern California, which is the emergence of a new class of companies focused on helping companies bridge the angel investment to venture capital gap. The gap — the sometimes difficulty hurdle of going from informal seed stage investors to the professionally  managed venture world — is a particularly acute issue for some of the angel backed firms here looking to gain additional capital. The issue? Although these firms are often able to gain some basic seed capital from small angels, the companies are often not in the shape where a venture firm is willing to chance it on them. They might not have a management team in place, they may be tackling a profitable–but smaller–market, or they might just need to get their company or product a bit farther than individuals angels can fund them.

Into that gap has stepped at least three (if not more) companies I am aware of — all here in Southern California — including Momentum Ventures, Groundwork Equity, and Venture Farm. All three are run by very operationally-focused folks, and all provide some combination of equity and operational experience to their companies. Venture Farm just announced an investment in XSCapacity -  Momentum has made investments in Thermark, Sendio, and other firms, holding exec positions at all of their investments. Groundwork recently put some dollars into Interneer, also coupled with some exec placements.

It will be interesting to see how the model plays out, but so far, the firms seem to be putting their money where their mouth is, and staking equity — not fees — on the success of their companies.

Cyber Monday

Monday, November 26th, 2007

It’s Cyber Monday, which has traditionally been one of the busiest shopping day of the year for online retailers. comScore Media Metrix is predicting more than $700M in sales today, as people get back to the office, and apparently spend Monday shopping instead of working. Last year, Cyber Monday sales totaled $608M. Southern California companies which might benefit from the flurry of online shopping include Shopzilla and Pricegrabber, with their comparison shopping services; the many online retailers here (Buy.com, NewEgg.com, Cooking.com, etc.); and the many electronics firms (too many to list) who benefit from this shopping season.

20 Worst VC investments of all time?

Tuesday, November 20th, 2007

Amy Quinn from the publication InsideCRM sent over this link of their recent list of the 20 Worst VC Investments of All Time, a ranking of venture funded companies which have crashed and burned. Topping the list is Amp’d Mobile; the only other So-Cal firm I can spot on the list is #11, eToys.

Most of these are Boom 1.0 firms — ie Webvan, Pets.com, Kozmo.com. It would be interesting to compare the types of numbers raised on those firms (ie $100’s of millions) versus the money being invested into firms today.

Good post on “finders”: or, how not to find venture capital

Monday, November 19th, 2007

Marc Averitt has a good post on “finders” — folks who help companies find capital — and how useful they are (or aren’t) in the venture capital area. Well worth a read. The majority of VCs I talk to will not fund deals from finders; and, most of the well regarded investment bankers will not engage early stage venture deals to help them find investment, because they understand that the VCs do not like funding deals with third party finders.

VMware, or why is infrastructure suddenly sexy again?

Thursday, November 15th, 2007

I’ve been rather amused by the amount of coverage given recently to VMware, a Palo Alto-based company which makes system virtualization software. Their software is used by enterprise IT datacenters to consolidate the amount of hardware they are using for applications. For example, instead of buying six different PCs to run six different applications, you can instead buy one big box and run all six “virtual” PCs inside that one box.

It’s a great solution, which for many, many years has been gaining traction in the enterprise space.

All of the sudden, however — after the firm’s blockbuster IPO — everyone seems to be paying attention to everything the firm is doing. The company was actually purchased by EMC back in 2003 - EMC is still a majority shareholder — and the whole growth story is much older than the IPO. I remember attending VMworld 2004, the firm’s annual conference, as an exhibitor, and thinking how smart of a move EMC made in buying the firm. (They paid $635M for VMware in 2003; the market cap today is right around $32.5 billion). I worked very closely with people at VMware in my position (at that time, socalTECH was still just something I was running as a hobby) and it was incredible how much growth they were seeing–but how no one in the technology or business press noticed.

I actually think the recent interest is a result of an overall — but subtle — shift in interest, and some funding, away from Internet and Web 2.0 investments, towards investments in infrastructure, software, hardware, and chips.  It’s probably the nature of the cyclical cycles in the technology industry, with some counter-trend investors starting to look at the more neglected industries as valuation (and hype) gets a bit too lofty in the Internet/web areas.

In many ways, the interest in the area is a repeat of the post-bubble syndrome seen with Internet companies. Immediately after the bubble, Internet companies were shunned–everyone removed the dot com from their names, no one could get funding, much less a meeting from a VC, and you wouldn’t want to mention you worked for an Internet company at a cocktail party. But, because of that, no competitors were funded, companies who had real business models could quietly build their businesses, and–at least here in Southern California–we saw a number of huge, successful, and very profitable firms built.

I think there is the potential for that to happen in the infrastructure, software, and hardware firms of today. Left to their own devices, they have not been attacked by dozens of startups, if they’ve figured out their business model should be quietly building their business, and will be the next big winners. Time will tell.

Being outside Silicon Valley, and costs of development

Wednesday, November 14th, 2007

Here’s an interesting blog post I ran across today:  Why Being Outside Silicon Valley is Good For Startups. In the post, Suzanne Dingwall Williams, a corporate lawyer and former VC in Canada, says:

One of my clients has been on an acquisition spree in Silicon Valley this fall, gathering interesting perspectives between wheatgrass and soy shakes. His notes: development cost and employee retention are two of the most significant barriers to growth for Silicon Valley startups. For these reasons many SV VCs, he says, will not fund startups unless at least 50% of their development work is done outside of California.

It’s interesting looking at the global perspective, a few weeks ago I noted a post which complained about the difficulty of employee retention–of people leaving for Silicon Valley — this from soneone in North Carolina. I guess it’s all a matter of perspective.

What I have heard here from some Southern California firms, is that some VC’s won’t fund companies unless at least 50% of their development work is being done offshore (India, China, Romania, Russia, etc.).