You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
Gus Richard makes no mention of IOT where ARM has been putting a major part of its investment. For me, however, the difficulty in IOT is getting any kind of handle on market information making it impossible to comment other than "wait and see".
He also says ARM will get only 5% of the server market, as against Simon Segars's prediction of 25%, which is partial of course, though ARM, by contracting with its licensees to provide it with market information, gives itself better information than anyone else.
You seem to know the server market. Do you have a view? I don't pay much attention to the analysts myself and this one seems to be basing his view on the smartphone market, which is well predicted to be maturing.
Good points Brian. Analysts are famous from making unwarranted extrapolations from domains they know to domains they don't. And thanks for your confidence in my knowledge of the server market, though I wish I deserved it.
I think ARM has potential for rapid growth in the server market thanks to in-house development by companies who own big server farms, but I don't have any sense of where the ceiling will be. Scalable ultra-fast and very low power multi-core servers are driving the growth, but ARM-based servers still have a weakness around fast I/O where Intel excels. So transaction-based computing (eg ATMs or point-of-sale <-> banks) still would seem to favor Intel over ARM architectures for example, whereas data analysis computing (big data for example) might favor ARM over Intel.
Many thanks. This is the kind of info one can only get on Semiwiki.
A second point is that ARM is almost impossible to value, other than by looking up the price, because its recent licences, its most valuable assets, generate no income for about two years. Standard P/E calculations, therefore, are no real guide. Some analysts say that the only approach is to do DCF calculations but that means you have to look 3-5 years forward to get any sort of meaningful result. I believe one has to decide whether or not you are going to stick with the company for the long term and keep abreast of the industry and ARM's market share. That means read Semiwiki!
