Intel Corp (INTC)

[This Causes an Error]

The big advantage Intel has had thus far is transistor performance which is what really counts for high performance at low power.

I have believed this for years, but seeing Merrifield at 2.13GHz, Moorefield at 2.33GHz, and BYT-T at 2.4GHz, when Krait 400 (roughly the same perf/MHz as SLM) can do 2.5GHz max in a 5" smartphone has me raising an eyebrow.

For Merrifield, I had expected Intel would go all-out on the turbo. With a dual core, pushing to 2.6GHz+ on SLM would have been a real winner (+22% performance). The low clock, even with a density optimized process, is surprising.

Your buddy Denni talked about relative density over a year ago.

I thought he was FOS over a year ago and legitimately believed Intel had a real density advantage at 22nm over 28nm. It was only later that I dug into the metal pitches of the foundries' 28nm nodes that I realized, "oh drat, he's right". Intel has the highest SRAM density, though, but of course that doesn't mean much for a design that's mostly logic.

You blame Intel about mobile yet you have known practically everything. You are the one that needs to own up to being wrong not Intel.

I take it you haven't been in Intel very long. Go back to Investor Meeting 2011 and 2012 and tell me that my assumptions (based on Intel's goals/expectations) were unrealistic. Intel has a high quality management team and is generally spot on with guidance/expectations. The recent misses from them are decidedly un-Intel.

I know you want to attack me, but I think that you (and others) do so because I tend to offer inconvenient truths and perspectives. Everybody loved me when I was an Intel perma-bull, but the minute I start going, "wait a minute, Intel said this,this, and that and didn't deliver", some of you want to crucify me.