RVN! 26 Banner revised (800 x 100 px) (600 x 100 px)
Semiconductor Expert Forum

Posted on

ARM and a LEG

ARM and a LEG
by Paul McLellan on 11-01-2012 at 5:09 pm
Categories: Arm, IP

I went to Warren East’s keynote speech at ARM Techcon today. There had been some hints earlier in the week that some significant announcements would be made and, while they were not earth-shattering, I think that they will be significant in the long term.

One interesting thing that Warren pointed out is that the ARM partner alliance is over 1000 companies strong. I don’t think anyone else can come anywhere close to that. It covers semiconductor partners, EDA partners, system companies, embedded software companies, standards bodies and more. It is quite an achievement to build up that large a portfolio of support so quickly. Back in the mid-1990s when I was at VLSI Technology, ARM had to set up a fund (that VLSI contributed to) to pay the real-time-operating-system companies such as Wind River and Green Hills to port to ARM. Now they could charge for the privilege. That’s quite a change.

For me the most interesting announcement was the creation of the Lenaro Enterprise Group (or LEG, ARM and a LEG, geddit?). Lenaro is the non-profit set up by partners to share development costs of Linux implementations primarily for mobile on ARM processors. It is all open-source, collaborative, sustainable and organic. They have 150 engineers. Lenaro Enterprise Group (LEG) is a second group within the Lenaro organization to do the same for Linux for ARM-based servers. They plan to add another 150 engineers by the end of this year. If they really mean 2012 that is a fast hiring ramp

The initial companies signed up for LEG are AMD, AppliedMicro, Calxeda, Canonical, Cavium, Facebook, HP, Marvell and Red Hat join existing Linaro members ARM, HiSilicon, Samsung and ST-Ericsson. The company that immediately leaps out from that list, of course, is Facebook. The other companies are the usual suspects, semiconductor and SoC companies, and operating system companies. But Facebook is a user of this sort of technology.

There is a press release announcing LEG, full of the usual press-release sort of stuff. But echoing what I said earlier this week,”ARM servers are expected to be initially adopted in hyperscale computing environments, especially in large web farms and clusters, where flexible scaling, energy efficiency and an optimal footprint are key design requirements. The Linaro Enterprise Group will initially work on low-level Linux boot architecture and kernel software for use by SoC vendors, commercial Linux providers and OEMs in delivering the next generation of low-power ARM-based 32- and 64-bit servers. Linaro expects initial software delivery before the end of 2012 with ongoing releases thereafter.”

Oracle (not a member of LEG it seems) appeared and discussed how strategic it is to port Java to ARM’s 64 bit architecture. I was expecting them to announce a more ARM-focused server strategy in their own cloud datacenters or something, but things seemed to be restricted to Java.

There were also some announcements about the Internet of Things (IoT) but that deserves a separate post.

I was completely ready for some high-up from Facebook to come out and talk about how they are ready to deploy ARM-based datacenters but it was not to be. That had been my prediction at the start of the keynote to Dan Nenni who was sitting beside me.

Posted on

Power, Predictions and Pills: Jonathan Koomey, ARM TechCon

Power, Predictions and Pills: Jonathan Koomey, ARM TechCon
by Holly Stump on 10-31-2012 at 5:00 pm
Categories: Arm, IP


ARM TechCon Software and Systems Keynote: Why Ultra-Low Power Computing Will Change Everything Simon Segars, speaking of the importance of continuing low power initiatives, introduced Dr. Jonathan Koomey, Consulting Professor at Stanford. (First impression, our kind of guy: He wears engineer shoes, not sales shoes!)

Koomey spoke for “Revolutionary change, and that’s not just loose talk….” in energy efficiency for the future. He cited Proteus Digital Health in biomedical; testing a new device embedded in a pill with no battery! It uses digestive juices to power the device! You take the pill, it communicates to report compliance…(Orwellian?) Just one example of how low power devices will transform medicine: delivering more info about the individual. Imagine sensors which can report pH, temperatures, etc.

Koomey summarized results of a 2011 study on computing energy efficiency he was part of. He stated:

  • Energy efficiency of computation has doubled every year and a half (actually 1.6 years) from 1946 (ENIAC) to the present, which is a 100x improvement every decade…..….think Moores Law!
  • Definition of energy efficiency: Computations/kWh, based on full load, using measured power data.

My personal note: A real weakness in the usefulness of this study is that it does not reflect idle modes, which are of critical importance in really understanding the issues.

Toomey delivered several truisms this audience well understands, such as: The things we did to improve performance also benefitted power efficiency. Laptops are taking over desktop market share. Revolution is driven by confluence of trends: computing, communications, sensors, controls. Importance of idle modes.

He presented interesting data on cell phone efficiency (minutes of talk time/Wh) vs year of introduction; on this graph, the efficiency increase he is seeing is 8% per year.

Are we burning up our planet? Well, we are using more devices worldwwide, but they are more efficient. Cloud is more efficient than data centers. Data centers in the US use 2% of our power, use around 1.5% in the world. Overall electronics is maybe 5% of our worldwide energy usage.

Power management trends: from process technology, to multicore, to software optimization matters, to “approximate computing”…its not just a silicon problem, it’s a system challenge.

The most fascinating topic: Energy harvesting, from ambient sources….and the future….
• Intelligent trash cans, self-solar-powered, compact trash, send a text when full, for fewer truck trips! Economic and environmental home run!
• Wireless no-battery sensors from Univ of Washington scavenge power from radio and tv signals in metro areas….
• MicroMote, generic sensing platform with ARM core, at University of Michigan. Ultra low power: 40uWatts in active mode, an astounding 11nW in sleep mode. Devices like this could someday monitor growth of a tumour.
• Streetline Networks, promoting sensors in parking spaces, and a mesh grid, in LA. Motes use only 400uW, with Mote Technology from Dust Networks. Think about variable messages signs or subscriptions to find parking spaces.

Deeper implications Toomey summarized included:
• Move bits not atoms (things)
• Customized data collection (nanodata not big data)
• Ever more precise control of processes
• Real time analysis
• Enabling “the internet of things”

How far can these energy efficiency trends continue?

  • One physicist calculated a theoretical limit…this graph takes us to 2041, three decades!!!
  • That is if we are clever…our horizon includes exciting applications…
  • Vision of a world with low power, cheap devices distributed everywhere..for medical, for transportation, for services, for manufacturing…..

And as Toomey commented, “Our community has a high “do to talk” ratio…..”
(Yes, the shoes don’t lie…..)

Shhh…..Researchers at Purdue and the University of New South Wales recently invented a 1 electron computer…


Posted on

Beneath the Surface lies the first real test

Beneath the Surface lies the first real test
by Don Dingee on 10-31-2012 at 9:00 am
Categories: Arm, IP

At CES 2011, Steven Sinofsky of Microsoft stepped on the stage and went off the map of proven Windows territory. Announcing the next version of Windows would support the ARM Architecture, including SoCs from Qualcomm, NVIDIA, and TI, set a new course for Microsoft.

But Windows, being the battleship-sized behemoth that it is, would take nearly two years to turn onto the new heading. Getting the entire fleet of PCs, smartphones, and tablets sailing together in formation would prove no easy task. Along the way, Microsoft provided hints of what was coming. A unified user interface, leveraging social-saavy tiles previewed on the smartphone, would usher in a new interface and steer clear of patent problems Android experienced.

In parallel during those two years, ARM has become the heart of almost all smartphones and tablets. Yes, there is a Motorola Droid Razr i running around Europe, a LAVA XOLO X900 in India, and probably a few more Intel smartphone designs waiting in the wings. But there isn’t direct, line-up-for-hours demand clamoring for Intel-based phones yet, and there hasn’t been an apps crisis since most Android apps use the Dalvik VM.

On another parallel path, ARM has been preparing a push into servers with a new 64-bit architecture (via our Paul McLellan), anchored by the Cortex-A57 processor, and the addition of the largest processor partner to date in AMD. For the Linux community running server farms, this is welcome news, albeit the first hardware is still at least a full year away. But again, no lines of rabid consumers ready to ditch Intel Architecture servers exist – the buzz is confined to IT.

What the average consumer will see, right now, is the Microsoft Surface. Just turn on the TV and notice the Surface logo in the live weather shots of Times Square, or one of several very hip Windows 8 commercials in a $1B campaign. In the first juxtaposition of ARM tablet and Intel PC processing cores, Surface comes in two versions presenting a direct choice of price, weight, battery life, and apps support.

But first comes the barking. The same IT departments looking forward to ARM-based servers find themselves conflicted over Windows 8. Their rationale is the slick new tile-happy interface is too radical a departure from Windows 7. It’s going to require a lot of retraining. That will be really expensive, and consume too many support resources. Therefore, we’re not moving to Windows 8. (Stamping of feet and holding of breath goes here.)

That will last until the CFO walks in with a Windows 8 tablet, from one of several suppliers including Dell and Samsung. The same forces of bring-your-own-device, or BYOD, that created an initial backlash against Microsoft will now start to work in its favor. A few Windows 8 devices in the right places will make headway.

The real test is consumers are showing up in retail stores hoping to get their hands on a Surface for a closer look, and what they are seeing is ARM and Windows RT which was cleverly spotted a head start. That same slick tile-happy interface from the commercials is right there, but under the Surface are all-new apps. For consumers coming from smartphones or tablets, an app store isn’t a foreign idea. RT apps will come as more developers get through certification and into the Windows store, and in the short term some may be missing.

Studies like this one from Google show most tablets are used most of the time on the couch or in bed, as a social or multimedia or gaming or reading screen, and there aren’t a whole lot of apps in play. Nobody has 700,000 apps installed. This “my app store is bigger than yours” is nonsense after the initial critical mass level is reached, and Facebook and other biggies are supported.

Now entering November 2012, we have the first test where ARM and Intel sit truly side-by-side for consumer hands to pick up and try, featuring the same look-and-feel but a different user experience. Will consumers choose the Windows 8 version of Surface and other tablets, to satisfy their reflexive impulse to install legacy Windows apps? Or will they embrace the ARM platform and Windows RT, grab the essential apps, and enjoy a lighter, thinner, and less-oft-charged experience with a lower price tag?

So you can infer which side I’m on, here’s my Sammy Hagar reference:
“If I’m wrong, hey, then I’ll pay for it.
If I’m right, yeah, you’re gonna DEAL WITH IT.”

Posted on

IBM Tapes Out 14nm ARM Processor on Cadence Flow

IBM Tapes Out 14nm ARM Processor on Cadence Flow
by Paul McLellan on 10-30-2012 at 7:33 pm
Categories: Cadence, EDA

An announcement at the ARM conference was of a joint project to tape out an ARM Cortex-M0 in IBM’s 14nm FinFET process. In fact they taped out 3 different versions of the chip using different routing architectures to see the impact on yield.

This was the first 14nm ARM tapeout, it seems. I’m sure Intel has built plenty of 14nm test chips but it seems safe to assume none of them included ARM processors.

Cadence went into a lot of detail about how the flows need to be changed, especially to cope with double patterning and to do transistor extraction of the FinFETs themselves.

For me the most interesting part was that IBM actually revealed a certain amount about their process. The FinFETs need to manufactured in a complete grid and then cut to separate them into individual transistors, some of which might not actually be used. The photograph above shows the FinFETs in yellow running horizontally with the vertical metal to interconnect them into standard cells (I would have a decent picture except that the meeting had the oddly inconsistent policy that we couldn’t have a copy of the presentation but we could take pictures of the screen).

Here are some more details of the process[TABLE] align=”left” style=”width: 470px”
|-
|
| style=”text-align: center” | 32nm
| style=”text-align: center” | 28nm
| style=”text-align: center” | 20nm
| style=”text-align: center” | 14nm
|-
| Architecture
| Planar
| Planar
| Planar
| FinFET
|-
| Contacted poly pitch
| 126nm
| 114nm
| 90nm
| 80nm
|-
| Metal pitch
| 100nm
| 90nm
| 64nm
| 64nm
|-
| Local interconnect
| No
| No
| Yes
| Yes
|-
| Self-aligned contact
| No
| No
| No
| No
|-
| Strain engineering
| Yes
| Yes
| Yes
| Yes
|-
| Double patterning
| No
| No
| Yes
| Yes
|-

es at IBM:

Just like the Global Foundries 14nm announcement, this has a metal pitch unchanged from 20nm. Some cells might be smaller but in general I think this means that designs won’t really be any smaller using this technology. Lower power, faster, better leakage. But not smaller.

In the TSMC keynote earlier, the focus was on PPA (power, performance, area). We used to say power, performance, price. But the big question is how costly these processes will be compared to 28nm which looks set to be a workhorse process for a long time. The IBM/Cadence presentation was the same. Of course it is early in development and no yield optimization has been done but somehow the fact that nobody is bragging about how cheap the technology is and how everyone will want to use it immediately implies that it will not be cheaper. And worse, that designs will not get hundreds of times cheaper after several process generations.

Posted on

ARM 64-bit

ARM 64-bit
by Paul McLellan on 10-30-2012 at 6:56 pm
Categories: Arm, IP

AMD announced yesterday that they would be building 64-bit ARM-based chips intended for use in servers. What was unclear is what the processors would be like. Although ARM had announced that they would move into 64-bit processors they didn’t have any that they had actually announced as being available for licensing.

At the ARM conference today, Simon Segars announced their first two, known as the Cortex-A50 family. Not only AMD has licensed these processors, but ARM also announced Broadcom, Calxeda, HiSilicon, Samsung, and STMicroelectronics as licensees. Representatives of these companies answered questions. And unlike what it looks like in my terrible photograph, Simon’s head did not explode in the middle of the session.


The two cores announced are the Cortex-A57, which will be the most advanced high-performance processor targeted initially at servers. The Cortex-A53 (I don’t understand ARM’s numbering system either) is ARM’s most power-efficient application processor. The two processors can also be used together in big.LITTLE form to get the best of both worlds, performance when you need it, low-power when you don’t. ARM had some information on just how much smaller and more power efficient these cores are. Some of the comparisons are a bit apples to oranges, comparing 20nm versions of the new cores against 32nm versions of the old ones. But they do seem to be architecturally more efficient, not just leveraging off the process node changes. Processors using these new cores are not expected in volume production until 2014. Both processors can run both 32-bit and 64-bit code.


The interesting aspect of this is whether 64-bit ARM processors will get traction in datacenters (and another interesting question is how soon before we have 64-bit processors in our smartphones). At the press-only discussion later it was clear that people knew more than they were letting on. There was even a hint that something in this area might be announced in Warren East’s keynote on Thursday. Everyone said that the first place that ARM servers will show up are in the big fast-growing companies that are adding thousands of cores per week. There have been rumors that Facebook has been evaluating ARM. If the datacenters are the profit center for the company, not an ancillary to the main company business, and if the company is growing fast, then they are interested in ways to cost-optimize (including lower power, smaller physical footprint, cheaper to acquire). Companies like this only run a handful of applications and so it would be cost-effective to optimize them for ARM since the hardware costs are not yet sunk since they would initially go into green-field datacenters.

Posted on

Improving FPGA Prototype Debugging

Improving FPGA Prototype Debugging
by Daniel Payne on 10-30-2012 at 10:00 am
Categories: EDA

FPGA Prototyping is growing in popularity as a method to get an SoC design into hardware running at clock speeds up to 100MHz or so. One downside during traditional FPGA prototyping debug is the limited number of internal signals that you can observe while trying to chase down bugs in the hardware design in the presence of running real software and drivers. Continue reading “Improving FPGA Prototype Debugging”

Posted on

Apple and Samsung Take All the Profit

Apple and Samsung Take All the Profit
by Paul McLellan on 10-29-2012 at 4:07 pm
Categories: General

I’ve talked before about how Apple and Samsung make most of the money in the handset business (and also about how Nokia…er…doesn’t). Now there is a report from Canaccord Genuity makes it clear just how much of the profit they make: 106%. And that is down from second quarter when they made 108%.

How can they make more than all of it? Basically because the remaining players, Nokia, RIM (Blackberry), Motorola (Google), HTC, ZTE, LG and others lose money in the aggregate (I don’t think they are all unprofitable, but Nokia and RIM in particular are both large and unprofitable). Motorola was blamed for weakness in Google’s overall profits too.

Apple is estimated to capture 59% of the industry’s profits on 6.3% of unit sales (15.4% of smartphone sales which might be more relevant since Apple doesn’t make any non-smartphones). Samsung captured 47% of the profit on 25.6% of unit sales. Samsung is #1 in revenue, having taken the top spot from Nokia earlier this year. Apple is #1 on profitability. Google’s Android is the #1 operating system but Google doesn’t make money directly on it (except in its Motorola subsidiary) so it is hard to allocate a profit number to it. In fact I continue to wonder whether Android is brilliant strategy or stupid. If the aim is too hurt Microsoft and Apple it is certainly doing that (not that Apple is exactly hurting but it would have sold even more phones if Samsung/Android was not available). If the aim is to add to Google’s bottom line I’m not sure how. It’s not as if people make more use of Google on Samsung phones than on iPhones.

Posted on

Jasper Property Synthesis Apps

Jasper Property Synthesis Apps
by Paul McLellan on 10-29-2012 at 7:00 am
Categories: EDA

Jasper restructured JasperGold so that it could deliver its formal technology more flexibly by having a base system and a porfolio of apps. This would also make it easier to upgrade capabilities by creating new apps. Today, Jasper announced two new apps:

  • JasperGold Structural Property Synthesis (SPS)
  • JasperGold Behavioral Property Synthesis (BPS)


The SPS App is used to detect and eliminate common functional design errors and ensure that the code is clean before validation starts. It is used early in the validation process without the need to write a testbench or provide stimuli. The structural properties are extracted from the RTL semantics and are used in early RTL development as well as during RTL signoff. These structural properties can be configured from a wide variety of pre-defined functional checks such as dead code checks, finite state machine (FSM) checks, arithmetic overflow checks, etc. The SPS App is tightly integrated with the entire set of JasperGold Apps drastically reducing the amount of checks that go undetected, un-proven, and un-diagnosed. Properties can be ranked, pre-classified and output in standard SystemVerilog Assertions (SVA) which can then be used in any assertion-based verification (ABV) flow such as simulation, formal analysis or emulation to increase functional coverage and reduce debug time. The SPS App provides a fully automated flow to identify and generate checks without the need to annotate the RTL.


The BPS App increases productivity and reduces time-to-market by generating assertions, constraints, and covers using the RTL and the existing simulation results obtained from batch simulations (FSDB/VCD) or interactive simulation (PLI). The BPS App is unique in its ability to create ‘white-box’ and ‘black-box’ properties as well as temporal multi-cycle properties. In addition,the BPS App can synthesize properties for signals from different modules across different levels of hierarchy. As with the SPS App, BPS provides an automated method for ranking and pre-classifying properties. The BPS App ranks synthesized properties according to their added functional verification value compared to design and manually written assertions, to reduce the number of property candidates the engineer must review. Moreover, the Apps methodology provides a unique flow that allows engineers to combine the BPS App and other JasperGold Apps to speed formal verification that significantly increases formal proof convergence. BPS supports both VCD and FSDB/VCD file formats.

SPS allows common bugs to be found early (and so cheaply) and improves the quality of verification beyond that achieved just with simulation. BPS can then be used in conjunction with simulation results to increase coverage, find coverage holes and improve the verification process.

Posted on

Apple v. Samsung: Mixed Phone Marriages End in Divorce?

Apple v. Samsung: Mixed Phone Marriages End in Divorce?
by Daniel Nenni on 10-28-2012 at 8:10 pm
Categories: General

A funny thing happened at dinner the other night. The SemiWiki blog “8 Things I Hate about My iPhone5” caused quite a discussion. Half the table had Samsung phones and the other half iPhones. It really was more of a religious or political debate versus a rational consumer electronic discussion. An interesting side note, it seems mixed marriages, where one spouse has a Samsung phone and the other an iPhone, were much more likely to end in divorce. Seriously, my position is that the majority of ex-spouses or ex significant others have opposing phones.

To test this hypothesis you can participate in a highly scientific poll HERE. Anybody can vote so please do.

My marriage used to be mixed. In the past I let my wife and children pick whatever phones they liked. Moving to smartphones however I insisted on one phone type for all (crowd sourcing) and the vote was 2 for Android (oldest son and I) versus 4 for Apple. Now we all happily belong to the Church of Apple and shun Samsungites.

The legal action between Apple and Samsung probably increased the mixed phone divorce rate. I spend quite a bit of my professional time on the legal side of things so I have been following the different Apple v. Samsung cases around the world. You can see the full Apple U.S. complaint HERE which resulted in a $1B+ jury verdict. Samsung internal emails about copying Apple products really sunk them. Generally, in most case law nowadays, incriminating email plays a significant role so be warned.

Apple lost a case in the UK with the judge ruling that “Samsung tablets are not as cool as iPads” with the judge imposing an interesting penalty. Apple was required to post an apology with a font size of at least Arial 14.

Well, the Apple “apology” is out now and it is indicative of the sometimes juvenile Apple versus Samsung discussions I have heard. Here it is in Arial 14:

Samsung / Apple UK judgment
On 9th July 2012 the High Court of Justice of England and Wales ruled that Samsung Electronic (UK) Limited’s Galaxy Tablet Computer, namely the Galaxy Tab 10.1, Tab 8.9 and Tab 7.7 do not infringe Apple’s registered design No. 0000181607-0001. A copy of the full judgment of the High court is available on the following link www.bailii.org/ew/cases/EWHC/Patents/2012/1882.html.

That is the apology part. Here is the not so apologetic part:

In the ruling, the judge made several important points comparing the designs of the Apple and Samsung products:

“The extreme simplicity of the Apple design is striking. Overall it has undecorated flat surfaces with a plate of glass on the front all the way out to a very thin rim and a blank back. There is a crisp edge around the rim and a combination of curves, both at the corners and the sides. The design looks like an object the informed user would want to pick up and hold. It is an understated, smooth and simple product. It is a cool design.”

“The informed user’s overall impression of each of the Samsung Galaxy Tablets is the following. From the front they belong to the family which includes the Apple design; but the Samsung products are very thin, almost insubstantial members of that family with unusual details on the back. They do not have the same understated and extreme simplicity which is possessed by the Apple design. They are not as cool.”

That Judgment has effect throughout the European Union and was upheld by the Court of Appeal on 18 October 2012. A copy of the Court of Appeal’s judgment is available on the following link www.bailii.org/ew/cases/EWCA/Civ/2012/1339.html. There is no injunction in respect of the registered design in force anywhere in Europe.

However, in a case tried in Germany regarding the same patent, the court found that Samsung engaged in unfair competition by copying the iPad design. A U.S. jury also found Samsung guilty of infringing on Apple’s design and utility patents, awarding over one billion U.S. dollars in damages to Apple Inc. So while the U.K. court did not find Samsung guilty of infringement, other courts have recognized that in the course of creating its Galaxy tablet, Samsung willfully copied Apple’s far more popular iPad.

Clearly, like the iProducts, Apple has taken a simple apology to a whole new level with an added focus on the Apple user experience! 😎

Also see: A Brief History of Mobile: Gen 1 and 2

Posted on

An AMS Reference Flow for Power Management Designs

An AMS Reference Flow for Power Management Designs
by Daniel Payne on 10-26-2012 at 5:42 pm
Categories: Cadence, EDA

At DAC in June I visited and blogged about 30+ EDA and Semi IP companies, however I didn’t have time to watch the TowerJazz presentation in the Cadence Theater entitled: AMS Flow for Power Management Designs. Today I watched the 26 minute video and have summarized what I learned in this blog post. Continue reading “An AMS Reference Flow for Power Management Designs”