I wonder if we will hit the $39M target (the middle of the range of this quarter's guidance). If we do, then we're OK. If we don't, it could signal a problem and the share price will suffer.
Although, Gopi is right about one thing (the article states):
"The merchant solutions in ARM servers that are in position to matter in 2016 include Applied Micro (X-Gene processor) and Cavium (ThunderX processor). The cache coherency issue in 2-socket configurations we brought up this past summer should be fixed in 4Q15 for ThunderX, which we expect to “duke it out” with Applied Micro’s X-Gene. We did not hear anything about AMD in ARM servers on Day 1 of our trip."
CAVM doesn't even have a product yet! And still, they compete! AMD is the 'little engine that couldn't'. No one should be paying attention to AMD. AMCC needs to commercialize X-Gene 3 at around the same time as ThunderX.
Post by christopher on Sept 14, 2015 16:23:27 GMT -5
phobos,
could the above sentence: "CAVM doesn't even have a product yet!"
be all about ONE financial analyst being the first to respond to a situation that has yet to begin, signaling investors to a beginning
of a possible "run" and or "sell off" regarding the products built by the two or three competing companies.
The above story involving Raymond James states that this competition between products will be happening during the year 2016. AMCC is already in the 2Q of 2016.
" AMCC needs to commercialize X-Gene 3 at around the same time as Thunder-X."
I noticed that too. A major press conference with no reason for being. I think (I hope!) you must be right. If AMCC offers a SoC with superior core-count and superior quality cores before CAVM, I think that AMCC walks away with the ARM data center server market. No question.
Post by christopher on Sept 14, 2015 19:14:06 GMT -5
Thanks Phobos,
Awesome Video!! very informative and interesting
For a group of extreme technology experts to begin designing and building these devices over six years ago and to have Intel sitting idly by doing nothing in comparison at all to compete with this. Holy Shit !!
WoW! I do not want to begin thinking too much before I should be, But I do have a feeling...
Looking forward to October 14
Last Edit: Sept 14, 2015 19:14:49 GMT -5 by christopher
INTC did nothing for a very good reason. The market that AMCC is seeking to address is (now) pocket change to INTC. INTC needs to address the broadest markets possible, those markets that can make a difference to INTC's top line. This is a similar reason to why VTSS wasn't that worried about BRCM. Perhaps INTC's purchase of FPGA-maker Altera is one attempt to address this (?)
I watched the following presentation by Mr. Gopi, and got some insight as to why he may not be as worried by CAVM:
The X-Weave product is designed to interconnect many racks of servers in what Mr. Gopi calls "scale-out". To the best of my knowledge, CAVM has nothing like this, and the scale-out will be up to the companies who make servers to tie everything together. AMCC has not only paid attention to the CPU, but it has also paid attention to how the servers communicate, memory latency, etc. I think this is a more complete package than what CAVM is offering (even if CAVM's chip turns out to be better --> and I'm starting to doubt that it will be).
I reiterate, if the mysterious press conference in October is to announce the commercialization of X-Gene 3, I think AMCC wins the race.