Post by Admin on Aug 5, 2015 8:54:44 GMT -5
I'm counting around 11 competitors in a currently very competitive ARM server/connectivity space:
amd, amcc, Avgo (Broadcom), cavm, ezch, Intc (Altera), iphi, mrvl, Nxp (Freescale), pmcs, qcom.
The above list is not in any particular order (alphabetical), but that's 11 competitors.
The addressable market of relevant ARM chips was given on page 3 of the presentation above (Roth Presentation). I gave the probability of the market getting to $3B (15% of the $20B total server market by 2018) a 10% chance ($300M). I gave the ARM server market a 20% chance of getting to $2B ($400M). I gave it a 30% of getting to $1B ($300M), and a 40% chance of it getting to 1% of the overall server market ($80M). That leaves my back of the envelope calculation of the total market being something like $300M + $400M + $300M + $80M = $1080M.
If everyone stays in the business, which would be impossible, then they would likely stratify the market into the following market shares: 33%, 17%, and 11%, and everyone else would get the remaining 39%. The respective quarterly revenues for the top three would be: $89M, $45M, and $30M. The $30M number probably isn't even enough to cover AMCC's modest costs (BE is revenue of $40M~$45M for AMCC). I should also say that I am not figuring-in revenues from additional non-ARM non-processor products (Gopi was a little cagey about numbers and mix in the conference call).
Right now in my mind, the competition is CAVM and INTC (Altera). CAVM because they are game competitors and INTC because they have piles of cash that they can throw at the competition for years. Whoever falls to position number three in the market loses. This is a market over-ripe for consolidation.
This year AMCC turns it around, replacing the lost PowerPC products with growing ARM revenue, and next year revenues take off. If AMCC is not having $50M~$90M quarters by 2018, I'm afraid that it would indicate that things have not gone as planned.
amd, amcc, Avgo (Broadcom), cavm, ezch, Intc (Altera), iphi, mrvl, Nxp (Freescale), pmcs, qcom.
The above list is not in any particular order (alphabetical), but that's 11 competitors.
The addressable market of relevant ARM chips was given on page 3 of the presentation above (Roth Presentation). I gave the probability of the market getting to $3B (15% of the $20B total server market by 2018) a 10% chance ($300M). I gave the ARM server market a 20% chance of getting to $2B ($400M). I gave it a 30% of getting to $1B ($300M), and a 40% chance of it getting to 1% of the overall server market ($80M). That leaves my back of the envelope calculation of the total market being something like $300M + $400M + $300M + $80M = $1080M.
If everyone stays in the business, which would be impossible, then they would likely stratify the market into the following market shares: 33%, 17%, and 11%, and everyone else would get the remaining 39%. The respective quarterly revenues for the top three would be: $89M, $45M, and $30M. The $30M number probably isn't even enough to cover AMCC's modest costs (BE is revenue of $40M~$45M for AMCC). I should also say that I am not figuring-in revenues from additional non-ARM non-processor products (Gopi was a little cagey about numbers and mix in the conference call).
Right now in my mind, the competition is CAVM and INTC (Altera). CAVM because they are game competitors and INTC because they have piles of cash that they can throw at the competition for years. Whoever falls to position number three in the market loses. This is a market over-ripe for consolidation.
This year AMCC turns it around, replacing the lost PowerPC products with growing ARM revenue, and next year revenues take off. If AMCC is not having $50M~$90M quarters by 2018, I'm afraid that it would indicate that things have not gone as planned.