I haven't seen the quote, but maybe he was referring to 2+3 U parts?
They simply don't want to talk about when they're targetting for 10nm HVM.witeken said:Also note the absence of projections for their own 10nm. They seem to know better what the competition will do than themselves! They don't want influence TSMC and Samsung, do they?
They simply don't want to talk about when they're targetting for 10nm HVM.
Remember, Bay Trail-M will not be followed up by Cherry Trail, but by Braswell. Not sure if there are any differences except for the name.
Airmont. According to ZDNET, it will have Iris graphics.Braswell features Goldmont which is the 14nm tock.
Intel's Kirk Skaugen says their IGPs are already faster than 80% of the dGPUs in the market. There were some rumors of a Broadwell-K delay, but he refers to the (GT3) product that will offer a 100X increase since 2006 to be available early 2015.
That doesn't make sense - oh, he means in the market as in currently deployed and in use. In that case, Intel's IGPs are definitely faster than Intel's old 'dGPUs' embedded in their chipsets as well as older GFX cards. So, this is true, but kind of misleading.
a year ago, AMD estimated, that 70-80% of nvidia product stack were 630 and below
http://www.techradar.com/news/compu...after-nvidia-graphics-with-kaveri-apu-1199636
Witeken
Thanks for sharing with us the keynotes. FWIW I think Intel is hinting a pretty much stable outlook for the next two years from a financial POV. They will have increased CAPEX and R&D spending from their new designs and products, but their other cash cows are stable (consumer business) or growing (servers). I didn't like the lack of disclosure of their 10nm node, but since Intel is at disadvantage on the mobile market, it's better for them to not show many of their cards and surprise everyone whenever they release it.
It seems that Intel is really committed to full domination of the PC market. The contra-revenue benefits are being fed-back to the consumer market, making Bay Trail an excellent value proposition for OEMs. Airmont will definitely a hit on the PC market.
The billion dollar question hanging on their heads is whether they will succeed on the mobile market. At least in terms of talking they seem to be going with their crack teams this time, 2015 will be critical for them as the trend for them on the mobile market will be clear for everyone to see.
Did you listen something about their digital radio? It will give them a huge advantage if they can field products with this technology.
Yes, I think Intel looks very stable for the next two years. And think that both 2015 & 2016 will be critical indicators on how nimble Intel is in responding to the changing needs of the mobile SoC marketplace.
Visa vi the OPEX goal of getting spending down to 30% of revenue - I'm concerned that the MegaFabs Intel will need to build for 450nm wafers is going to delay their ability to reach that target, but that's out past the next two years.
Intel is forecasting ~6% growth for 2014 and another ~5% for 2015. What will mainly drive a lot of growth in the coming years is their DCG. In 2014 the data center will have a $14B revenue, and for fun you can multiply this by 1.15 a few times to get their 2018 revenue with a 15% CAGR. Oh, and one more thing, it's a 50% net margin business, so that will be $12B from DCG alone in 2018 to sustain Moore's Law.FWIW I think Intel is hinting a pretty much stable outlook for the next two years from a financial POV. They will have increased CAPEX and R&D spending from their new designs and products, but their other cash cows are stable (consumer business) or growing (servers).
The most important thing, in my opinion, is that they deliver big on 10nm. If they can get Broxton and SoFIA on 10nm in H2 2016 for 2017 widespread availability (e.g. SGS8) with a big materials update to provide huge improvements and with great yields, they can do a lot of damage and improve their mobile footprint and start the domination to reduce reduce ARM to no more than a footnote in computing history. I highly doubt that will happen, though (the first thing, not the domination thing).The billion dollar question hanging on their heads is whether they will succeed on the mobile market. At least in terms of talking they seem to be going with their crack teams this time, 2015 will be critical for them as the trend for them on the mobile market will be clear for everyone to see.
Nope.Did you listen something about their digital radio?
Didn't know it would be, apparently, so helpful.A double plus for Witeken's yeoman work in providing info and summaries on the investor meeting!
You can still watch it, though. Link in the FP.I wish there was a transcript of the 1st Q&A. I think there is some good info buried in the both by what was said and not said. As a whole, this management teams really seem to be function well and they are at the top of their game.
Intel is forecasting ~6% growth for 2014 and another ~5% for 2015. What will mainly drive a lot of growth in the coming years is their DCG. In 2014 the data center will have a $14B revenue, and for fun you can multiply this by 1.15 a few times to get their 2018 revenue with a 15% CAGR. Oh, and one more thing, it's a 50% net margin business, so that will be $12B from DCG alone in 2018 to sustain Moore's Law.
The most important thing, in my opinion, is that they deliver big on 10nm. If they can get Broxton and SoFIA on 10nm in H2 2016 for 2017 widespread availability (e.g. SGS8) with a big materials update to provide huge improvements and with great yields, they can do a lot of damage and improve their mobile footprint and start the domination to reduce reduce ARM to no more than a footnote in computing history. I highly doubt that will happen, though (the first thing, not the domination thing).
Nope.
The most important thing, in my opinion, is that they deliver big on 10nm. If they can get Broxton and SoFIA on 10nm in H2 2016 for 2017 widespread availability
Broxton is a 14nm Soc.
I spent most of Thursday watching the Intel (NASDAQ:INTC) Investor Meeting. Today they posted the videos and I watched it again, some sections I ran four or five times.
Bruce Burnworth said:I was stumped by a drop to 62% gross margin and 5% growth guidance. Why so low? So I emailed Russ. What's up? Russ's response: "Remember next year is the 50th anniversary of Moore's Law. 2015 will be a year of underpromising and overdelivering. No one wants to look bad on that anniversary year!"
It does not. Intel now has a Bay Trail platform with a ~5$ platform cost delta, so those costs will become progressively smaller as the high BOM Bay Trail gets replaced.I apologize if I missed it, but does the restructuring of the PC and Mobile groups shake up the contra-revenue program in any way?
I'm really interested in seeing whether Intel will achieve its contra-revenue goal of cost parity in terms of platform with ARM.
Most of the fab costs will be aggregated to COGS and thus reflected on gross margins, not on OPEX.
Thanks for the info on CAPEX being absorbed into GOGS - now I understand Intel's drive for OPEX being 30% of revenue, which really looked unrealistic at first.
Not only CAPEX, but also whatever higher operational costs the new node fabs have, but CAPEX plays a very important role on Intel's COGS: About 40% of COGS is depreciation/amortization.
Wow, that's high (by today's standards). That must be higher than heavy industries like Detroit (because Detroit doesn't turn over 'models' as quickly nor are they likely as heavily automated).