Comms chips

SiGe Semi Scores $42.8M

Chipmaker SiGe Semiconductor Inc. secured its second round of funding, to the tune of CDN$66.4 million (US$42.8 million), the company announced today.

The round included SiGe Semi's initial investors and brought in three new investors: Prism Venture Partners, TD Capital, and 3i Group plc (see SiGe Semi Scores $42.8M). Previous financing included a CDN$36.1 million (US$23.3 million) round in 2000 and an initial CDN $6.3 million (US$4.4 million) in angel funding.

As the name implies, SiGe Semi specializes in silicon germanium (SiGe), which is used for making high-speed semiconductors that can outperform silicon chips but are built using complementary metal-oxide semiconductor (CMOS) processes. Other high-performance options include indium phosphide (InP) and gallium arsenide (GaAs).

The advantage to CMOS is easy integration; as design sophistication increases, multiple CMOS chips can usually be combined into one. But high-speed analog and RF parts made in CMOS with silicon have traditionally needed the most expensive manufacturing techniques to get their performance high enough. Making these parts in another material such as SiGe allows an older and cheaper manufacturing process to be used and still ensures the parts will have the necessary performance characteristics.

"The idea that CMOS is always cheaper is not the case. It is, for some big integrated designs, but not in the field we're in," says Jim Derbyshire, SiGe Semi CEO.

The company's strategy is to partner with CMOS chip vendors, using SiGe to fill particular holes in those vendors chipsets. In the 802.11 market, for example, SiGe Semi provides power amplifiers alone, and the combined set of CMOS and SiGe chips is offered to manufacturers as a whole.

Bluetooth is arguably the company's most successful market so far, and it's dabbled in cable as well. And last year, SiGe Semi made the jump into optical, announcing parts for 2.5- and 10-Gbit/s networks (see SiGe Semi Enters Optical Ring).

Optical won't be a huge seller at first, but SiGe Semi officials want to stake their claim for the time when the market reemerges. "We know the market is seen in the press as dead, and there are a lot of companies pulling out," Derbyshire says. ”But there is still a volume of parts being purchased, and SiGe provides a decidedly improved way of providing those functions."

A similar philosophy is coming from InP specialists, such as Inphi Corp. and Velocium, which also are providing high-speed optical components (see Inphi Moves Beyond InP).

For optical, SiGe Semi plans to build laser drivers, laser modulator drivers, post-limiting amplifiers, and transimpedence amplifiers (TIAs). The company has TIAs shipping at 2.5- and 10-Gbit/s, and within the next month it plans to announce several more 2.5-Gbit/s components as well as an improved 10-Gbit/s TIA.

SiGe Semi was spun out of the National Research Council of Canada in 1996 to provide SiGe manufacturing consulting, but by 2000 it had begun to sell its own chips, outsourcing the manufacturing. The company is based in Ottawa and has 95 employees.

— Craig Matsumoto, Senior Editor, Light Reading

whyiswhy 12/5/2012 | 12:52:26 AM
re: SiGe Semi Scores $42.8M Let's see, AMCC, Vitesse, and every other ESTABLISHED GaAs foundry out there is doing SiGe...and these guys raised NEW VENTURE money?

It just proves (once again) VCs have absolutely no clue...
Physical_Layer 12/5/2012 | 12:52:12 AM
re: SiGe Semi Scores $42.8M No offense intended, but I think your comment is silly. SiGe is just a material. Would you say any company that is using steel as a construction material isn't worth investing in because everyone uses steel? No! It all depends in what you are doing differently. The material is only dimension.

I know nothing about the company, for the record.
whyiswhy 12/5/2012 | 12:51:54 AM
re: SiGe Semi Scores $42.8M ...and steel is just a material: but would you invest in a steel mill?

More to the point, would you invest in a Silicon wafer fab and some nifty products from it? How about a GaAs foundry and some super nifty products from it?

If so, I got a Golden Bridge I can sell you...hey, you can collect tolls from the cars that cross!

The big boys have had multiple SiGe products out in the market for several years now. The VC boondoggle for SiGe-let's-build-a-big-foundry-and-take-over-the-world was over two years ago when the bubble burst. They should go fabless, and not waste all that money money. Now my insurance rates are going to go up again.

Some day, we will figure out how to let the limiteds sue the crap out of the VCs that waste their money just to get management fees, and still be limiteds.


Physical_Layer 12/5/2012 | 12:51:50 AM
re: SiGe Semi Scores $42.8M WhyisWhy - SiGe is a fabless company according to their website. I stick by my analogy. Just because a company USES steel (not MAKES it), doesn't mean they don't have some creative value and IP to add to the market. Look at silicon - I don't care how long people have been making chips out of sand, there will continue to be innovations worthy of investment.

If for some reason I'm wrong here, (ie, let's say SiGe raised the cash to build a fab) then there is merit to your argument. Am I wrong?

whyiswhy 12/5/2012 | 12:51:35 AM
re: SiGe Semi Scores $42.8M Better read their history:

"During the period 1996 to 1998 SiGe developed its silicon germanium expertise and started a GPS module product business activity. Both grew but had different business drivers. Under pressure from the investors, the Company was split into SiGe Microsystems driven by Derek Houghton and SiGEM driven by John Roberts to enable each part to focus. SiGEM is currently on the Toronto Stock Exchange under the symbol GPS. In November 1998, Maxim, one of SiGe's early customers, began marketing its first silicon germanium products. SiGe carried out the deposition step of the manufacturing process."

They threw their VCs money down the drain building a SiGe fab, exactly as I said. They wasted either somewhere between $30 or $70M (US) in the process. Then in 2000 they gave up on that strategy and are now fabless.

I heard of the company several years ago. I was not aware of their strategy shift; I take back my earlier comments. This round may make some money for the investors. The founders are probably screwed.



Sign In