Quicklogic Corp (QUIK) 2011 Q1 法說會逐字稿

Good day, ladies and gentlemen, and welcome today's QuickLogic Corporation First-Quarter 2011 Earnings Call. (Operator Instructions)

I would now like to introduce your host for today's conference, Mr. Andy Pease, CEO of QuickLogic, and Ralph Marimon, CFO of QuickLogic. Mr. Marimon, you may begin.

Thank you, and good afternoon. Before we get started, let me take a moment to read our Safe Harbor Statement. During this call, we will make statements that are forward-looking. These forward-looking statements involve risks and uncertainties including, but not limited to, stated expectations relating to revenue growth from our new products; statements pertaining to our design activity and our ability to convert new design opportunities into customer activity; market acceptance of our customer products; our expected results; and our financial expectations for revenue, gross margin, operating expenses, profitability, and cash.

QuickLogic's future results could differ materially from the results described in these forward-looking statements. We refer you to the risk factors listed in our annual report on Form 10-K, quarterly reports on Form 10-Q, and prior press releases for a description of these and other risk factors. QuickLogic assumes no obligation to update any such forward-looking statements. For your information, this conference call is open to all and is being webcast live.

For the first quarter of 2011, total revenue was $5.5 million. As we previously announced, this revenue was below our guidance. It represents a sequential decline of approximately 20%.

New product revenue totaled $1.2 million, which represented a 46% sequential decrease. New product revenue was negatively impact by reduced demand from our broadband data card customers, and a non-QuickLogic technical issue that's halted shipments to our secure banking customer. Andy will address both of these situations in detail in his prepared comments.

Mature product revenue in the quarter totaled $4.3 million, which was an 8% sequential decrease and was due to slightly lower bookings during the quarter. Our non-GAAP gross profit margin for Q1 was 66%, and was above our guidance due to the mix of product shift.

Non-GAAP operating expenses for Q1 totaled $4.0 million, which was at the low end of our guidance. Non-GAAP operating expenses declined slightly compared to Q4, primarily due to the timing of expenses related to our new product development efforts.

On a non-GAAP basis, other income and expenses and taxes for Q1 totaled approximately $76,000, primarily due to income tax expenses related to our overseas operations. This resulted in a non-GAAP loss of $435,000, or $0.01 per share.

Our cash usage for the quarter was $858,000. However, due to the exercise of options, our cash position increased by $542,000 to close the quarter at $22.5 million.

While accounts receivable decreased during the quarter due to the reduced revenue level, DSOs increased, primarily due to the timing of certain collections. During the first week of April, we collected approximately $400,000, which had previously been expected to be collected in Q1.

Inventory has increased substantially over the last two quarters, primarily due to an increase in PolarPro II inventory. This increase was driven by lower shipment volumes than expected, and the need to support the very short lead times service providers are giving to our broadband data card customers. We currently expect to use this inventory as we go forward and are monitoring it carefully.

Our Q1 GAAP net loss was $878,000, or $0.02 per share. Our GAAP results included stock-based compensation charges of $443,000. Please see today's press release for a detailed reconciliation of our GAAP to non-GAAP results.

Now I'll turn it over to Andy who will update you on the status of our strategic efforts.

Thank you for joining us this afternoon. To say I am disappointed in our first quarter results is an understatement. However, I must underscore that while I am disappointed, I am not discouraged.

On the Monday following the release of our preliminary Q1 results, I left for what was literally a trip around the world. It started in China where Brian Faith, QuickLogic's new VP of Sales and Marketing, and I called on existing and potential Tier 1 customers. From there, Brian went to Taiwan and South Korea and I went to Europe.

Our goals were to audit our current position with customers, prioritize our near-term efforts, and move our best strategic opportunities forward. Today I'll share the results of our audit, what we see as our near-term priorities, and the status of our tactical and strategic objectives. Following this and Ralph's presentation of our Q2 guidance, we will welcome your questions.

In our preliminary Q1 announcement, we noticed-- noted that shortfalls in our wireless data card market were responsible for roughly 60% of our new product revenue shortfall. Our forecast for wireless data card revenue was based on forecasts from our customers and our silicon partner, Icera.

As recently at the first week of February, the alignment of these two data points gave us confidence in the guidance we presented. However, as we moved through the month of March, it became apparent that the dynamics of the wireless data card market were changing significantly.

To clarify what changed and how it impacted us, I'm going to divide the market into two pieces -- the mature markets, which were dominated by North America and Western Europe; and emerging markets.

In mature markets, custom consumer preferences have shifted away from discreet data cards. Based on our research, this shift has been driven by three events.

First, tablets have become a preferred mobile computing device. Since tablets also come with built in cellular access, their success has reduced wireless data card demand in the mature markets.

Second, due to the computing power offered by the most recent generation of smartphones, they are being looked at more often as a computing device than ever before.

Third, many of the new cell phones offer tethering. In a tethered arrangement, the smartphone performs the duty of a dedicated wireless data card.

These three changes has led to a lowering of aggregate demand. This is not to say that demand is totally dried up; it's just lower than it was. Between that and the inventory digestion we think will continue through Q2, demand for our CSSPs was negatively impacted.

In emerging markets, the story is considerably different. Demand from these markets appears to be increasing, albeit not quickly enough to offset the near-term softness our customers are experiencing in mature markets. Our customers in emerging markets face hyper-competitive conditions which result in uncertain demand, limited visibility, and greatly reduced lead times, impacting everyone in these markets. That said, we believe we are well-positioned in this market.

Since we remain the sole source supplier for wireless data cards that use the current generation Icera baseband processor, we have increased our inventory level to support the uncertain demand and shorter lead times.

In addition to managing these challenges, we are working with a Tier 1 data card manufacturer and a Tier 1 baseband supplier on a new type of data card that is highly differentiated and unique in the market it addresses.

While data cards used in the portable computing industry rely on USB connectivity, this new type of data card uses an SDIO interface and targets digital still cameras, video cameras, and mobile computing devices that have a standard SD memory slot.

Manufacturers of these cameras have recently incorporated Wi-Fi connectivity that allows users to directly upload pictures and videos to the web. One issue is that Wi-Fi hotspots are not always available. Our customer believes consumers of these devices will embrace cellular connectivity just as they have in the computing market.

Since most new portable computers include an SDIO slot, this new type of SD card can serve double duty as a wireless data connector for PCs. If this proves to be correct, the wireless SD data card will expand the aggregate data card market and replace a portion of the market currently served by USB devices.

The balance of our Q1 shortfall was attributed to issues we are facing in our secure access data card market. We and our lead customer believe these issues have been fully defined and are moving rapidly towards resolution.

In the secure access data card market, we are focused on two types of applications -- banking and business-to-business communications. In the banking application, our lead customer has near-term issues to resolve that will temporarily impact our revenue.

During 2010, revenue from our lead customer was driven primarily by our secure banking application. As recently as the first week of February, we had an order commitment from this customer covering our Q1 forecast. However, an application problem was discovered and the commitment was withdrawn.

This problem was investigated as a very high priority and traced to the design of their printed circuit board and the passive filtering it used. It was totally unrelated to QuickLogic's CSSP.

We helped our customer develop a short-term workaround for its then customer. With our support, they are in the process of fixing their design problem. Due to the time it will take to complete and test this design and gain regulatory approval, we are not forecasting any revenue for this secure banking CSSP during the second quarter.

With the release of our new ArcticLink II CX platform later this year, we will broaden our reach and value proposition in the business and banking sectors, as well as address consumer security, machine-to-machine, or M2M applications, and a wide variety of connectivity applications.

Our lead customer likes the platform so well, he has decided to design his entire business security product line around the CX platform. This is great news for the long term. However, our revenue will be impacted in the short term while our customer waits for the CX to be releasing their secure business product. Given the CX production release schedule, and the time required for our customer to test and gain industry approvals, it will probably be the end of this year before we see production begin to ramp.

While the timing here is unfortunate, it is important to understand the role our CX chip plays in this design. CX replaces four chips, including the processor, and is literally providing our customer a system on a chip, or as it is known, an SoC function.

In addition to its embedded 32-bit RISC processor, the CX platform includes OTP memory to hold the operating firmware, scratchpad memory, two separate security blocks, programmable fabric, and an interface that allows for die-stacking and multi-chip packaging.

With security becoming a huge issue across the spectrum of connected mobile devices, we think CX presents a viable SoC solution for a wide variety of machine-to-machine and consumer applications.

In mobile computing applications, the CX platform was specifically designed to support the need for side loading, virtual USB ports, offload processing, content protection, and encryption. While customers implement these important functions in our CX platform, they won't have to modify their application processor software each time they change processors. This addresses the growing time-to-market challenges we are seeing across the mobile industry.

The important takeaway is that with CX, we are the central value proposition in several of the secure data card designs we are targeting today. In these designs, CX isn't an added chip; it is the chip.

As it stands today, even in advance of the form factor reference design release we anticipate will happen late this quarter, we have very significant interest in the ArcticLink II CX from major Tier 1 manufacturers in Japan, Asia, and Europe.

During our Q1 conference call, I was proud to announce our first VEE DPO production shipments, and with our customer's position, to share the name and the model number of the end product. In Q4, we shipped production revenue of our new ArcticLink II VX chip to BenQ for their R100 tablet. In February, BenQ released the R100 for sale in selected Asian markets.

We are very excited about BenQ's decision to be the first manufacturer to use a VX. We have additional VX design opportunities, including a design in the pre-production stage from a smartphone manufacturer. While this design has progressed in line with our expectations, until we are shipping production volume, we are taking absolutely nothing for granted.

We recently shipped a pre-production order for a tablet computer that is no longer an opportunity as the OEM decided to compete on price versus performance. We are working on a number of other designs in tablets and smartphones that we hope to discuss in detail in subsequent conference calls.

While the slow start to 2011 is clearly disappointing, I hope the data I have shared with you today helps you understand why I am not at all discouraged with the bigger strategic picture for QuickLogic. Without a doubt, we face many challenges and risks. However, I believe we have the right model, the right technology, and the right team to successfully execute our strategy.

With that, I'll turn the call back over to Ralph so he can present our guidance, and following that, both Ralph and I will look forward to your questions.

We are anticipating that the reduced demand for our broadband wireless data card and secure access data card business will continue through the second quarter of 2011.

Due to this reduced demand we are forecasting that new product revenue will be flat at $1.2 million in Q2. Based on our bookings and backlog, we are estimating that our mature product revenue with be flat with the Q1 level at $4.3 million. Total revenue is forecasted to be approximately $5.5 million, plus or minus 10%.

As in prior quarters, our actual results may vary significantly due to schedule variations from our customers, which are beyond our control. Schedule changes, particularly those that may impact new product revenue, could push or pull shipments between Q2 and Q3 and impact our actual results significantly.

On a non-GAAP basis, we expect gross margin to be approximately 64%, plus or minus 3%. The gross margin percentage is driven by the forecasted production rate and the anticipated mix of products shipped during the quarter.

We are currently forecasting non-GAAP operating expenses to increase to $4.4 million, plus or minus $300,000. The increase in expenses will be driven by continuing development work on two new CSSP platform families and new hires that are forecast for our R&D department.

Non-GAAP R&D expenses are forecasted to be approximately $2.2 million, while non-GAAP SG&A expenses are also forecasted to be approximately $2.2 million during the second quarter.

Our other income and expense will be a charge of up to $60,000 during the second quarter. Our stock-based compensation expense during the second quarter is expected to be approximately $420,000.

At the midpoint of our guidance, our non-GAAP loss is expected to be approximately $0.02 per share. Excluding the impact of additional warrant and stock option exercises, we expect to use approximately $1.0 million in cash, primarily due to the lower revenue levels and higher engineering expenses.

This concludes our prepared remarks, and now we'd like to open the call for questions.

Thank you. (Operator Instructions). Our first question comes from Jason Rechel from Needham & Company. Your line is open.

Hi, guys. Thanks for taking the question. I guess my first question is on the broadband data card market. You guys touched on declining demand in mature markets. At what point can we expect to see a stabilization in demand there, or have we already kind of seen the worst decline in demand that we may expect? And at what point does the emerging markets demand start to offset that? Thanks.

Jason, I wish I could tell you that I could forecast any better than our customers or the operators can forecast. Unfortunately I can't. If we could, then we'd give you better news.

Okay, understood. Thanks. Next, you guys have talked in the past about kind of an Apple factor in the tablet market. And I think you touched on, in your prepared remarks, on one of your tablet OEMs deciding to compete on price rather than performance. Is that a trend that you guys are seeing, or are you continuing to see strong traction for your product?

As a matter of fact, we don't see that as a common thing. This is somewhat of an anomaly. Most people are trying to compete on value.

Okay, thanks. And then last one from me and then I'll hop back in the queue. Now that your mature revenue has kind of taken a step down, is this kind of the run rate that we could expect to see for the next couple of quarters?

Well, we don't have a lot of visibility into the mature, so I think you can see it certainly for the next quarter, because that was our guidance. But further out, it depends on the bookings, and from what we see right now, that's the level we're at. But we don't have a lot of visibility, and we actually don't invest a lot of time into looking that much further out on the mature business.

Okay. Very good. Thanks, guys.

Thank you. Our next question comes from Brian Coleman from Hawk Hill Asset Management. Your line is open.

Great, thank you. My first question, I just want to clarify. The tablet design that you lost, is that the tablet you've referred to in the past as going into production this summer?

That was the tablet we talked about in the last call, and I actually mentioned it as well in the pre-announcement, yes. Yes.

Okay. I'm curious, how does-- is it common for a design that's in pre-production to change like that? How does-- can you--

Well Brian, you know us, and we don't have a ton of experience in this mobile space. We've been at it for two years. So certainly in the old FPJ model it'd be highly unusual. And we've always been cautioned by people that we work with that this is, while it's not totally common, it's not unusual.

Okay. We've heard that there's been some issues with Android 3.0 and kind of stability issues and people's experience with the Motorola Xoom, and that having kind of delayed and pushed out some of these tablets coming to market. Are you seeing that at all in your designs in your pipeline?

Absolutely, absolutely. You know when-- I'll go one step further. When Android announced Honeycomb 3.0, I think the general market assumed that when it was delayed, it would go to a lot of people. And what we know is that in fact it only went to a very few people in the early times. So the-- a lot of customers are getting only their first look at Honeycomb, and that's probably because Google does not want to support a lot of different people as they try and get this right.

Okay. If you were to look out at your pipeline, and this obviously is not a hard forecast I'm asking you to make, but what would be a base case number of tablets that you could see going into production over the second half of this year?

You know, we're not in a position that we want to say that, nor do we want to put a number out there. I will say that we are encouraged by the activity, and beyond that I think that it'd be premature for me to talk about it.

Okay. And then let me try this one. The BenQ tablet, was that kind of a one-off design with them, or was that a platform design where they could be spinning off multiple devices from that?

This first tablet was a seed tablet for them, because as we reported back in Q4, BenQ has really been classically an e-reader. So they wanted to see how this tablet does, and we're still in close communication with them. As you know, they've actually highlighted VEE as a main selling feature in this tablet.

All right. With respect to VEE, I know there is a 3.0, VEE 3.0 I think coming out. Can you talk a little bit about your opportunities with respect to cost reducing VEE and potentially either expanding the market or changing the margin, profiling the [flavor] with some of the changes you can do on the design front?

Brian, we have not done any formal announcements on VEE 3.0. I am a little surprised that you bring it up. We have-- we obviously, with our very close customers, have talked under strict NDA, and I think the only thing I can really say is Ralph actually talked about the engineering costs going up.

The engineering costs going up, but that's an--

Our engineering costs going up. So we obviously do have a roadmap, and one of the things that we have to be careful about is pre-announcing products before they're ready for customers. So I think that's about all I can say on that.

All right, that's fair. Looking out I guess as an investor kind of waiting for some of these designs to come to market, and we seem to be kind of inevitably being pushed back here, can you just give a little bit of kind of more of a long arch view of how things have changed at QuickLogic in the last year? And kind of speak to your-- how you see the markets evolving and your overall value proposition in these markets, with respect to CX as well as VX.

Sure. And let me even do the timeframe maybe two years back, because I think that would be a good baseline to look at.

If you looked at our value proposition, first of all, it has changed dramatically, certainly over the 4 years I have been here, 4.5 years I've been here, where initially we were just selling low power FPJs. And then we came up with a bridge chip concept. Certainly the Icera stuff was all bridge chip. And then by introducing ArcticLink and having this customer engagement model where we would actually become part of our customer's R&D process, we really became a companionship.

To me, and I've said this before publicly, I really think CX can be a game changer for us. Because CX, when you put in an embedded processor, which by the way, we don't necessarily expose that to the customer writing their own code. We use that as a third dimension of customization we can do with CSSPs on top of putting in proven system blocks in our fabric and doing the actual drivers that sit on the application baseband processor.

And now that we're also looking at this M2M market, and certainly even the secure data card market, we see that we are not only becoming an offload engine, but for certain applications even a system on a chip.

In terms of hardware differentiation, one of the things that we've seen different over the last couple of years is I think with the introduction that Apple has had in several products, are certainly the Tier 1 guys believe that they need to differentiate via hardware. And that plays into our strong suit.

The third thing is the time-to-market factor. I think a lot of people keep trying to anticipate what Apple does and then try and quickly come out with new products. And the fact that CSSPs can give them hardware differentiation quickly also helps them with the time-to-market.

And probably the last and the biggest change I've seen in my 4 years here is keep in mind, and I think I've said this, when I started here, the customer overlap that we have now with CSSPs versus FPJs is virtually zero. So in 4 years, we've gained access to literately a collection of Tier 1 customers. Not only Tier 1 customers, but their executive managements, and probably even more importantly, Tier 1 partners. We're calling it the highest levels of all these companies.

So the story is beginning to catch hold. People believe in what we're saying, and they really see the value that CSSPs can offer them.

Okay. All right, that's all I've got. Thanks very much.

Thank you, Brian.

Thank you. Our next question comes from Hamed Khorsand from BWS Financial. Your line is open.

Hi, guys. I'm just trying to pinpoint here, it seems like the industry is moving exactly to a risk criteria that you guys faced from the get-go as far as pricing on your DPO and VEE product line. Does this force you guys to cut pricing so you guys get into more and more tablets?

I think our-- we believe our value proposition is still there. We're not going in a price-cutting mode. My experience, after being in sales for 30 years, is usually it's not a price issue. Usually. So that's-- but we are certainly have a roadmap in place, and with our top tier customers, we have communicated this roadmap, and frankly, they are pretty excited about it.

Okay. But you're also saying that you lost your Tier 1 tablet customer because of a pricing issue.

I said they decided to go with a-- they decided to keep competing on price as opposed to performance, and what I'm really talking about is really total BOM cost as opposed to that we were too expensive. That was not necessarily the issue.

Okay. And can you give how many sale trials you're in, or how many designs are you working on right now on tablets and smartphones?

No, we are not giving the details of our funnel, other than to say that we believe that we are progressing as we would expect, and we're pleased with our progress in our funnel development.

Okay. So the increase in R&D that you're seeing, is that being applied to just getting these designs out to customers, or where is this R&D being applied?

R&D is being applied to our roadmap on both the CX side, which we-- so if you look at our platform offering, we have, when we talked to both investors, customers, and the press, we have two roadmaps that we're pursuing. One is the connectivity roadmap and the other is the display and visual enhancement roadmap, and our R&D dollars are being applied to roadmap items in both.

Okay. So there's really no correlation between R&D and any increase in, as far as customer activity goes?

No.

Okay. All right.

No.

That's it for me, then. Thank you.

Thank you.

Thank you. (Operator Instructions) Our next question comes from Robert West from NI Technology Research. Your line is open.

Hi, Bob.

Please check your mute button.

Sorry about that. Good afternoon, Andy and Ralph.

Hi, Bob.

How are you doing?

Thanks for taking my call. I would like to start with one housekeeping question. During the day, preceding the Company's issuance of adjusted guidance, there was an unusually large purchase of put contracts. Is this anything the Company has looked into?

Bob, I'll give you a real brief update on that. We, and for investors that aren't aware, Friday, April 8 is when we issued our revised guidance. It was brought to Andy's and my attention over the weekend that an unusual trade had taken place in a put option. On Monday, we talked to our general council and outside council.

While the Company doesn't have a legal responsibility to report that, we thought it was in the best interest of the Company and the shareholders to report what information we had, and we turned it over to the appropriate enforcement agencies for NASDAQ and the Option Board.

We know that they are looking into it. They've made it quite clear to us that it's a confidential investigation. I'm sure they'll take the appropriate action, if any is necessary, but it's not clear that we'll ever hear back from them on it. So, again, we turned the information over to the proper authorities, and they're investigating and we'll let them do what they need to do with it.

Okay, well thank you, Ralph.

Sure.

My next question is on the one market area that I don't think you touched on which is mobile enterprise. If you could give us an update on that product area, Andy.

Mobile enterprise is progressing nicely. We go through our normal cycles where there was one mobile enterprise customer that ran its course after five years, but we are seeing new designs pop up at our existing partners. So that's actually going quite nicely.

Okay. My second question is on the Uplink conferences coming up. I noted that plans are in place for QuickLogic to have a booth at Qualcomm's second annual Uplink conference. Can you give us some color on this conference and the benefits that it accrues to QuickLogic?

Well, we think that the conference is very good for making contacts with one of our ecosystem partners and the people that they attract. It is a Qualcomm conference, as you mentioned, and in the past we've gotten to get very good dialogue not only with their partners but with Qualcomm people as well. We are frequently down there, and it's just a nice time to get together with lots of people who make themselves available.

Okay. My next question is in a similar area, dealing with smartphone platforms. In past conference calls, QuickLogic noted that VEE DPO had been included in two high profile reference design programs of a major Tier 1 OEM. The first was for a 4,000 unit reference design targeted at developers, and the second a 4,000 unit reference design targeted at OEMs. Can you give us an update on the OEM program specifically, and if it's resulting in a smartphone customer engagement for QuickLogic?

I can tell you that we have had a number of referrals from the reference platform you're discussing, and that has resulted in opportunities that are in our funnel and that are a little bit too early for us to discuss on this call. But we are very pleased with the activity we see from that particular reference platform.

Keep in mind, Bob, that that reference platform really kind of hit the market at the end of last year, beginning of this year.

Okay. Then I guess a follow up to that, which you may or may not be comfortable going into further detail on. Can you give us some customer, on the customer engagement, metrics or the timing of future revenues that might come out of this program?

Yes, we are not going to talk about that in this type of forum.

Okay. I want to turn to the CX platform. I have some questions there. Back in October of last year when that platform was first introduced with a press release, it was mentioned that the CX would sample or become available in Q1 '11. Can you give us an update on that?

Yes. The actual sampling of the platform has been pushed out to Q2. The more important part of the platform, what we decided was sampling the chip was probably not nearly as important as coming out with this what we're calling a form factor friendly reference design. So this in fact could be an end product, and it basically highlights all of the connectivity features and security features that the CX has, and we expect to be announcing that shortly.

Okay. Maybe a follow-on question. The CX platform, CX is QuickLogic's first silicon with an embedded processor. Can you give us some color on the software side of the product beyond the secure access market, at least? Do you have proven system blocks of software for the embedded processor that's going to be important to your smartphone and tablet markets?

Yes, we do. As a matter of fact, like we said, with CX, we call it firmware. We try and differentiate firmware from software, but you're talking about actual code that would be running on this RISC processor inside our CSSP. So yes, we do have applications that we are going to be announcing. I think the first thing that we did announce was side loading, and you'll be hearing more from us on this in the future quarters as we basically roll those PSBs out.

Okay. I think a final CX question. You've expended a lot of effort introducing VEE and DPO and the MX platform to the smartphone and tablet markets. Will some of these past marketing efforts and customer engagements accrue to the benefit of the CX platform and maybe allow QuickLogic to reduce time-to-market on that platform as you go forward?

That's really a great question. I'm glad you asked that, because initially, I think people could think that those two things are separate, and I think what customers have begun to see is the fact that CSSPs can help to do both sides of the equation; not only the display, but also the connectivity.

Big thing it has done for us, frankly, is given us access to Tier 1s in multiple geographies. So absolutely, it's been instrumental. And any time we've had a Tier 1 audience with VX, we also talk about CX, and there's a high degree of interest in that.

And that's across-- it's across all your markets?

Yes. I'm just mentally trying to do a checklist here, and of all the Tier 1s that we have, I can't think of anyone that was not interested in CX when we introduced it as an aside when the meeting was really more about VX.

Okay, very good. I think that that's all of my questions. And thank you for taking them and best wishes on the road ahead.

Thank you, Bob.

Thanks, Bob.

Thank you. I'd like to turn it back over to Mr. Andy Pease, CEO of QuickLogic, for closing remarks.

Well, thank you all for joining us in this call. We really appreciate your efforts and supports, and we look forward to talking to you in the near future. Thank you.

Ladies and gentlemen, thanks for participating in today's program. This concludes the program. You may all disconnect.