Sylvia and tread: Implications of Lattice/Vantis deal on Altera and Xilinx
Salomon Smith Barney
Thursday, April 22, 1999
--SUMMARY:----
* Yesterday after the market close, Lattice Semiconductor announced an
agreement to purchase Vantis, a wholly-owned programmable logic device
(PLD) subsidiary of Advanced Micro Devices.
* If consummated, the deal combines the number 3 and 4 suppliers in the PLD industry. Three companies, Altera (roughly 32% market share), Xilinx (31%), and Lattice/Vantis (21%) will effectively dominate the market.
* Vantis has had a weak track record for many years, and we believe it will take considerable time for Lattice to improve the fortunes of the unit.
* The deal furthers the concentration of power within the PLD industry into the top tier. As such, we believe the business impact upon Altera and Xilinx will be relatively benign.
Mirror, mirror, on the wall...
Lattice and Vantis are almost identical is size (roughly $200 million
each in sales last year) and product mix. Both have been strong
suppliers of product-term complex programmable logic devices (CPLDs),
ranking #2 and #3 behind Altera. Both have been the largest suppliers of commodity simple PLDs (SPLDs, or GALs), and neither has any meaningful field programmable gate array (FPGA) or look-up table (LUT) PLD exposure, the PLD segment dominated by Xilinx and Altera. In short, Vantis has been Lattice's most comparable competitor for many years.
Implications likely limited.
Lattice will likely face a number of integration challenges (e.g., sales force and channel management, product line assimilation, and
manufacturing rationalization) over the coming year that may make it more vulnerable at certain customer accounts to competitive inroads by Altera and Xilinx. On the other hand, the increased bulk of Lattice/Vantis may reassure some customers about the staying power of the combined entity.
There is also the possibility that the PLD pricing environment may
improve modestly due to the removal of a competitor who has been known to use price as a primary weapon. On balance, however, we believe the
market share ambitions of each of the remaining players will probably
keep the price environment as competitive as ever.
The merger is a milestone, but only in the sense of being a marker along a road the PLD industry was already travelling. In sum, we believe the industry will continue to coalesce into the hands of fewer players over the next several years.
ANALYSIS OF THE DEAL
Background
The programmable logic device market is a $2 billion segment of the
semiconductor industry, and has grown 20% on average since 1989. The
high density segment (i.e., CPLDs and FPGAs, which now represent 86% of the market) has grown 25-30% on average over that same time frame. As such, it is one of the fastest growing segments of the semiconductor industry.
AMD/Vantis was once the largest supplier in the market, but has
essentially failed to grow as AMD's focus remained locked on the
microprocessor market. The company also committed a number of strategic missteps over the last several years, including out-sourcing its software development tools and missing the rise of the FPGA segment of the PLD industry. The company has recognized these mistakes, and created Vantis a few years ago as a quasi-independent, wholly-owned subsidiary in an attempt to give the unit greater independence.
Why Lattice bought Vantis
Why buy a business that hasn't grown for the better part of a decade?
We see several positives from Lattice's perspective:
1 Certainly, removing a competitor from the market has potential benefits in terms of pricing power and customer account control.
2 Since we believe AMD was a determined seller, the purchase also keeps Vantis out of the hands of competitors. Had Vantis fallen into the hands of Xilinx or Cypress, Lattice's competitive position may have been seriously compromised.
3 Vantis may also control a certain number of patents or other
intellectual property that Lattice may find attractive.
4 Perhaps most importantly, the combination brings Lattice closer in size to the two industry leaders. The industry was rapidly bifurcating into a three-tiered system, with Altera and Xilinx in the first tier, Lattice, Vantis and Actel in the second tier, and Lucent, Cypress, Atmel and everybody else in the third tier. While the merger may not legitimately move Lattice firmly into the first tier, it certainly delays any deterioration in the gap.
Why Altera and Xilinx Won't Worry Too Much
The negatives about the transaction that Altera and Xilinx will use
against the combined entity include:
1 The product lines are redundant and not very synergistic. Indeed, the sales channel will be burdened with two sets of incompatible software tools with little hope of integration in the next year or more.
2 Last week, Lattice salespeople were fluent in describing to customers the shortcomings of Vantis' products. What will they say next week?
3 There will be inevitable disruptions in sales relationships and product plans, with consequent confusion and frustration in the customer base.
4 Vantis has been a fading factor in the market for the better part of a decade. Even though Lattice management is well regarded, it is unclear whether they can reverse the slide. |
