Samueli:Yes. You need to upgrade your connectivity to HDMI 2.0.
The spec designed to handle the higher frame rates is already done and
available. You will see HDMI 2.0-compliant chips available within a
year. I see these as startup issues that will get sorted out over time.
EE Times: What’s the second hottest thing you saw at the show?
Samueli:Gigabit wireless . . . [pauses] although this is more about us at Broadcom.
EE Times:Why is that important?
With 5G [Wi-Fi 802.11ac operating on the 5-GHz band], wireless media
sharing and the whole home video distribution have finally become a
For the first time, I think that carriers – cable
operators and DSL guys – are convinced that they can deliver reliable
HDTV programming throughout the home. Carrier-grade video content comes
to media gateway, distributed to set tops, notebook computers and
Further, with the Miracast protocol [peer-to-peer
wireless screencast standard created by the Wi-Fi Alliance] in place,
people can also wirelessly display what they see on their handheld [tablet or smartphone] on their big-screen TV.
EE Times:Do you think that the second-screen scenario you just described may take away the significance of smart TV?
Samueli: Exactly. Why spend a lot of money on TV, when the technology in your hand keeps upgrading more often?
Samsung has shown a plan for “user-upgradable smart TV.” Their idea is
to attach the “Evolution Kit” device to the back of a Samsung smart
Samueli: Well ... it’s not so simple. Look at
the evolution from iPhone 1 to iPhone 5. The upgrade is not just in
software, but substantial changes are also made in new hardware. I think
TV is a wrong platform for constant upgrades.
EE Times: What’s your third pick among the hottest trends at CES?
Proliferation of Wi-Fi in everything. We are finding literally
hundreds of gadgets on the show floor with Wi-Fi in them. They range
from home appliances to thermostats to alarm clocks. I just saw a Wi-Fi
connected talking alarm clock developed by a startup called ivee [that responds to voice commands to tell time, date, temperature].
EE Times: Beyond Wi-Fi in everything, how do you see the Internet of Things (IoT) taking off?
There are two types of IoT. One is those plugged into the wall, like
washing machines and refrigerators. Another is based on sensors running
on very low power, enabled by something like Bluetooth low energy. Standards
are already here, but I think this will be a gradual process. We need
really low cost, and really low power devices communicating among
themselves to enable IoT.
EE Times: How far are you
at Broadcom in terms of developing ultra low cost and ultra low power
devices? Are you acquiring companies to enable this?
Samueli: No. This is based on our internal R&D.
EE Times: How low power does a device have to be?
year of battery life. There are a lot of things we can do [to bring
down the power] by creating a deep sleep mode or making a device shut
EE Times: Voice recognition seems to have gained substantial momentum at this year’s CES.
Yes. It’s gotten to the point that people are feeling comfortable with
it, I think. Siri may not understand what you are saying but it can
transcribe your commands. Voice will be important as a user
interface. Navigating a TV guide, you may just want to talk to your TV
or set top, for example, “Which NFL games are on today?” Remotes will
EE Times: Any other near term trends?
I’d add connected car. The car is becoming a mobile living room. There
will be LTE and Wi-Fi within a car to download data, music and video.
Of course they do. They won't give us any faster internet until they are ready for it. I'm sure it is already possible to provide much faster connection. But they wait. Who can tell what they are waiting for? As for the standards, they set it, for sure. All together. travel insurance for EU residents
Frank, while what you describe is theoretically true, I don't think it's the wave of the future. The idea of people setting up home servers to send around massive files between computers in the home was never a big hit, and is becoming less so.
This kind of file transfer is going to "the cloud," as current lingo calls it. So in fact, your last mile connection is where the action is, rather than the home network. And any IoT appliances in the home, that may actually remain within the home network, are likely to be noise level in terms of bit rate.
As TV migrates to the Internet, this will be even more true. Multiple people watching different TV programs will result in multiple high bit rate streams from "prime time anytime" web servers, through the broadband link, then through the home network, rather than from any sort of centralized in-home PVRs.
I remember when Kodak first came out with its flat grain Ectar 25 negative film. Even though theoretically "no one" would appreciate the improvement, in 4 X 6 prints, over previous negative films, the improvement was obvious. An image so smooth that you'd swear the film's emulsion was liquid.
Same deal here. People were saying that HDTV was unnecessary. And yet, even on not-so-large screens, it's gorgeous. With UHD, you can get a little closer to your normal size TV set, and still see a beautifully smooth image.
The eye/brain system is complicated. Even if first order approximations imply that "no one" will notice the difference, your brain will inform otherwise. Just look at the way people gush over "retina displays," on very small screens. These have a greater pixel count than 1080p.
What especially appeals to me about UHD is that codec improvements since the introduction of HDTV should make UHD require little or no more channel capacity than HDTV. So it should actually be practical and doable. As opposed to being a bandwidth hog.
"unless you have an extremely large TV..." is part of the answer. The other part is, as tb1 says below, the "immersive experience." I expect average screen size as well as resolution to continue to increase, subject to the usual caveats about the economic sweet spots. Not many consumers will pay $20k for a huge UHDTV display, but millions will pay $2K for such a display. The same applies to your comment about bandwidth. As long as there is insatiable demand for more bandwidth, there are profits to be made in finding ways to provide more bandwidth.
It is the nature of our business to make things better & cheaper over time, and the main issue is rarely "if," but rather "when."
One of the things I find interesting about this article is the number of times he said something to the effect of "the standard is already here." It wasn't that long ago when the products more often preceded the final standard. 802.11n was that way. If I remember correctly a was also. Are companies getting smarter about getting together on standards early on? Or are some of these future standards wishful thinking that will need a lot of modifications when the implementation technology is available?
Dear Junko - thank you for another informative interview. Dr. Samueli's comment that a war on number of compute cores (processor compute power)may be overblown is interesting -- from a company that is a master on very large (MIPS) cores in networking....
Perhaps you could have asked two more pertinent questions:
1. MIPS versus ARM cores - BRCM just purchased architectural licenses for 32- and 64-bit cores ("real men fine-tune standard ARM cores") -- any trends?
2. BRCM is the king of wireless connectivity Combos. Top-2 is Qualcomm and Qualcomm, while still have a Combo IC (RF radio portions, integrates digital portion of connectivity into processor.... for a variety of well detailed specific reasons. Qualcomm is the king of baseband processors, of course. Any thoughts on that ftom BRCM's CTO?
Many thanks in advance if you could follow-up on the above - as you did with your Rockchip interview.
And -- HNY Junko!