Second further question is [indiscernible] question. We’ve had one year with the first licensing. Has the revenue production met your expectations, and if not, what do you think is going on?
Michael Lebby
I think it has met our expectations. There is now product there on the market. It’s not in high volume, I will admit to that. But we’re looking for that feedback of our material working in a product that’s out in the market. Yes, it’s a step in the right direction. It hasn’t been as fast as everybody would have liked, but it’s met our expectation internally.
Kevin
Oh, yes. Where do I start?
Michael Lebby
Everybody wants to go on tour so don’t be too long.
Kevin
Oh no, it’s going to be great. I’ll start high level. The timing for the 4x200 PIC chip and a completed transceiver is still ambiguous. At the Optica conference last fall, you said the 4x200 PIC to be ready in 2024. Is that still on target? Have you received prototyped 4x200 PICs yet from these wafers?
Michael Lebby
All I can say is it’s in progress.
Kevin
Okay. Recent presentations showed Vpi (phon) uniformity across four different devices. A general question is, when you receive a wafer with thousands of devices, what do you do with those thousands besides the smaller sample sizes for reliability and for performance testing?
Michael Lebby
We have a lot of statistics analysis. We do a lot of measurements. It takes a lot of time. A lot of tests. A lot of measurements. They have a lot of statistics. We used those statistics internally to make sure everything is working as we expect; in addition, to look at where you can fine-tune performance. For example, we did statistics on Vpi and we did statistics on bandwidth, and that gives us a lot of good information to hone the next set of devices that are coming through the fab. I think what we saw was that we did that measuring and we did the statistics, the performance is really good. We also saw opportunities to improve it even more. I’m not going into any detail about how much we can improve it but we’re not at the limit and we can certainly improve the bandwidth and we can certainly improve the Vpi. There’s no question about that.
Kevin
Sure. The four devices, the Vpi seemed to range from about 1 to maybe upwards of 3. Do you know why the variations exist? Are these just different device designs? For example, if someone’s willing to pay you whatever it costs to order 25 wafers of the 200G 1 volt modulators, is your foundry partner able to produce an entire wafer of those 1 volt modulators at high yield, and basically is that design locked in?
Michael Lebby
The four results you’re talking about are package modulators. I just wanted to put out to show we have four (phon) packaged modulators. I didn’t put all the statistics out. I just wanted to have a quick sample to show people. We don’t just have chips and we don’t have just chromophores. You’ll probably see a lot more statistics as we go forward in the next six to nine months. There’s no question about that. Because that’s what people want to see, they want to see a lot more statistics.
In terms of foundry and foundry capability, we’ve gone public with one foundry and they’re really excited. There’s a reason why they’re really excited. They have the confidence just like we have, but rather than answering your question in a quantifiable fashion, I will say we’re both very comfortable on what’s going on.
Kevin
For wafer scale poling, I’m going to bug you about this, is it at the point where it’s consistent, high yielding and an economical process? I know my email to you referenced if it takes a week or two to poll an entire wafer and your yields are 20%, that’s obviously not satisfactory. Besides saying you’re comfortable with that, can you give any insight into the extra time and cost and poling yields, please?
Michael Lebby
Not really? Okay. I’d just reiterate, the question keeps coming up because people have talked negatively about polling in the past. We’ve worked really hard at that, and I think I gave the same answer last year. It is not an issue for us. It is a uniform process. There’s other things we’re more concerned about. It’s not something we’re concerned about like the outside world, and I know because of negative discussions or negative talk about it, people talk about it all the time. Internally, it’s not the issues we worry about. We’re worried about other issues. We’re comfortable with that.
Kevin
A couple more if you don’t mind. For your direct PIC sales prong, are you going to be creating an entire transmit PIC with lasers and other components or are you going to be supplying the multichannel modulator array to a company who would then put together the transmit side and with the lasers and couplers, and et cetera?
Michael Lebby
That is definitely dependent on customer requirements. When you look at PICs in general, each customer has their own way of mounting lasers. As you probably are aware, every company has a different technique to do that. We want to have a platform that is generic so that if a customer wants to use their particular laser of choice, then we can accommodate it. We could take the position of going to develop lasers and work with a laser supplier and come up with our design, but what happens if a particular customer says I don’t like that supplier, I don’t like that laser type or I don’t feel the way you’ve put it down on paper is reliable enough. Lasers are one of the problems when it comes to reliability. As you probably are aware, it is an issue that gets a lot of attention. I think, for us, we don’t see it as being a critical part, but if a customer turns around and says we need you to go work with a supplier, buy (phon) lasers, make sure it’s done in a way we like, we’d certainly take a look at that.
Kevin
Okay. Last one. Slide 13, there was a quote.
Michael Lebby
I don’t remember Slide 13.
Kevin
I forget what the quote was. Slide 13, I’m just curious who said that quote.
Michael Lebby
I’m going to have to ask Atikum. Remember, Atikum?
There we go. We can’t answer it but it was an industry leader.(It was Nvidia)
Kevin
Okay. Now I think it’s my last question. Is there a possible scenario where you can name a tier one transceiver maker collaboration partner, whatever, sometime in 2024 before the 4x200 PIC chip is complete? At what stage, basically, can you reveal a serious player who was considering committing to this technology?
Michael Lebby
Our methodology is really simple. It’s got to make good business sense from both sides. I think you’ve asked this question to me last year or the year before, who’s your foundry, who’s your foundry, who’s your foundry, right? We didn’t name any foundries. We just named a foundry because it made good business sense for that particular foundry and it made good business sense for us. We will do the same, apply the same methodology for any partner, whether it’s a package partner, a transceiver partner, or even another foundry.
I think that’s the way we want to look at that, and if we see good business and they see good business, good opportunity for both, then yes, we’ll put out a press release, but I’m not going to give any guidance because I think it’s premature right now.
Kevin
Okay, that’s it. Thanks.
Male Speaker
Actually, I’m glad you brought this slide back up. My question’s related to this. You mentioned, of course, and we all saw that graphic about the face—responses that you got when you announced the performance and how excited everybody was. Then mentioning this, the Infiniband of where the industry wants to go or needs to go. Does this represent a level of confidence now at 800G but understanding that 1.6 and 3.2 is possible, or do they have to see these in demonstration before we get adoption?
Michael Lebby
The answer is we’re seeing an inflection point in the industry for this technology. This technology, as everybody is aware—there’s a lot of investors here been here a long time. It’s had a negative rap and we all know this. We’re changing that and we’re changing a lot of naysayers, and that is happening day by day. It’s not been easy, but we are doing that. When people see a demo, hopefully, just like everybody here, when you go see the demo, you’ll get excited. You’ve seen slides before but now you’re going to see something that is real, is tangible, is physical, it works. This is what the industry wants to see. They also want to see the lifetime reliability data. But I see that changing quickly now. Yes, I think this is a good time for us.
AI
