WEBVTT 1 00:00:00.000 --> 00:00:00.730 To the 2 00:00:00.970 --> 00:00:04.450 STFC-RAL-CR03: we don't have a pointer. Right? So I use this phone. Yes. 3 00:00:05.750 --> 00:00:07.420 STFC-RAL-CR03: yeah. Good. So 4 00:00:09.780 --> 00:00:19.650 STFC-RAL-CR03: supposed to be a summary of this conference. which had a hundred 25 talks. They were very well organized in categories, all these categories. 5 00:00:19.840 --> 00:00:28.000 STFC-RAL-CR03: 16 of them. Everyone had a keynote speaker, so it was extremely a lot of people, and 6 00:00:28.800 --> 00:00:48.179 STFC-RAL-CR03: I have to confess that 7 00:00:48.270 --> 00:00:59.549 STFC-RAL-CR03: we're living in hard times already. And II feel I have to just make a remark, which is that you know I've I've been in this place for 50 57 years. I had 1,967 8 00:00:59.880 --> 00:01:01.579 worked it all that time. 9 00:01:01.890 --> 00:01:12.199 STFC-RAL-CR03: and it's been absolutely wonderful. First on Nimrod, which is a long accelerator, and then it's served on Ps. And the Sps. Then it snapped, and 10 00:01:12.340 --> 00:01:16.919 STFC-RAL-CR03: on Sld and and then on the linear colliding which might have happened. 11 00:01:17.110 --> 00:01:43.610 STFC-RAL-CR03: So that's not sort of done. But the other thing that I've done here is to meet people, and by the 100 of you actually covered up all the way through absolutely wonderful education for me, having come as a kid from South Africa, and and having a very narrow view of the culture in the world, and meeting so many people, international people from all over. And I mean, just in in terms of our current 12 00:01:43.740 --> 00:01:50.240 STFC-RAL-CR03: prices in the world. I mean the thing I've learned so much from 13 00:01:50.400 --> 00:01:59.989 STFC-RAL-CR03: from Saturday and from Andre, and even like yesterday on the stairs I had a short chat with Andre. And you know, I said, 14 00:02:00.040 --> 00:02:03.700 STFC-RAL-CR03: what do you think? A day? He says, Why, why is there all this terrible? 15 00:02:03.900 --> 00:02:32.170 STFC-RAL-CR03: He says broken promises. And then he started talking about something that I had no knowledge of, and it's it's one of my other thing I was thinking of before you. None of you will remember him, but we had a fantastic Boulos, Palestinian, and he worked in Jeff Manning's group. Jeff is Jewish, and they were the best of friends you can imagine, and you know said to me, the 2 State Solution work. I'm convinced that. And I've seen so many things in my career here. People are very diverse 16 00:02:32.170 --> 00:02:46.480 STFC-RAL-CR03: cultural backgrounds. And what have you and able to work so well together. So I mean, I'm I'm still optimistic hope that's justified. But I just want to make that remark as as the introduction, because these are terrible times that we see now. 17 00:02:46.600 --> 00:03:11.159 STFC-RAL-CR03: my apology is the fact that I'm gonna cover only a tiny fraction of what's this and and the other thing is that normally, when I get talks, I know what I'm talking about here in many areas where I'm not sure talking about, because other people's work. And it's such a diverse range of topics, and so I would invite anybody in the audience. Please 18 00:03:11.160 --> 00:03:27.180 STFC-RAL-CR03: don't, or or on zoom don't hesitate to to speak up if you would like to correct me, because I'm sure there are many corrections that I would love to be still in about, and the slides themselves you can find here. And 19 00:03:27.480 --> 00:03:28.830 STFC-RAL-CR03: so 20 00:03:31.910 --> 00:03:36.129 STFC-RAL-CR03: so my selection is, yeah, 21 00:03:36.950 --> 00:04:02.910 STFC-RAL-CR03: those categories. In any case, sort of obscure the fact that there are some areas which have really moved forward very rapidly, and which I think are fantastic growth areas. The one which is most exciting to me is the fact of moving from monolithic or hybrid pixel detectors into stacked Cmos image sensors which are being led by Sony until now. But others journey in. 22 00:04:02.910 --> 00:04:17.180 STFC-RAL-CR03: and that to me is reminiscent of the evolution of monolithic detectors 50 years ago, and that lead to a Nobel Prize, and I wouldn't be surprised if this development does as well. 23 00:04:17.300 --> 00:04:31.539 STFC-RAL-CR03: so I'll talk a bit about that, and then really moving forward rapidly, and then I have some thoughts which might be a little bit 24 00:04:31.740 --> 00:04:42.050 STFC-RAL-CR03: contentious about risk management, and I may be off off the the not accurate on that, and if so, please speak up. 25 00:04:42.280 --> 00:04:49.299 STFC-RAL-CR03: I'd like to say some words about LCD. Stewart, slack and an absolutely brilliant startup just before this workshop. Eighteenth of September. 26 00:04:49.340 --> 00:05:10.579 STFC-RAL-CR03: And it's turned on incredibly well. And it's it's created a huge challenge for position sensitive detectors, in fact, there was hardly any presentations at the workshop people so busy working at home trying to get their detectors up to speed for this. It is a huge challenge. And the other area which is related to that is ultra fast. 27 00:05:10.580 --> 00:05:23.449 STFC-RAL-CR03: Any beef scale electron microscopically higher energy. Typically electronics have been about 300 kb, and now up to like 3 Mev. And the reason for that and the physics and and science 28 00:05:23.450 --> 00:05:33.709 STFC-RAL-CR03: payoff. From that we'll say a few words about, and that's where CD. To add second laser pulses for which this year the physics Nobel Prize has connected it to that. 29 00:05:33.880 --> 00:05:47.179 STFC-RAL-CR03: I'll say some words about very high energy cabaret, strong units which certainly caught my eye. And we're going on. It's not my area of expertise, and there may well be others who come perhaps on Zoom. Who can say so? Some words about that 30 00:05:47.410 --> 00:05:48.840 STFC-RAL-CR03: dog matter 31 00:05:48.860 --> 00:06:09.489 STFC-RAL-CR03: always been interested in the possibility of that calls, and they've been, you know, theoretical arguments about this go back and forth for many years. Martha is almost dead, but it turns out that's not the case. I'll show you this. Talk about that and then. 32 00:06:09.750 --> 00:06:29.460 STFC-RAL-CR03: there's a slide which I don't even know. It's a joke. Actually, I mean, it's a so called, supposedly the dark method distribution in in the Milky Way. I mean, I've seen simulations of that from from way back, and it's generally it's just a highly precise spherical distribution, right? And with 33 00:06:29.460 --> 00:06:44.579 STFC-RAL-CR03: that serious. 34 00:06:45.240 --> 00:06:47.230 STFC-RAL-CR03: oh, sorry. 35 00:06:48.420 --> 00:06:53.489 STFC-RAL-CR03: So what about the stacked census? So the idea is that you scissor? 36 00:06:54.060 --> 00:07:14.429 STFC-RAL-CR03: I'm sorry showing an exploded view. Right? So basically what you what you have, a, a, a base layer which has to be sufficiently thick at the beginning to provide mechanical stability. And then so in each case the red line here represents the active, seamless circuit. So these are all standard 37 00:07:14.470 --> 00:07:39.469 STFC-RAL-CR03: cbos, you know unit layers seamos chips, but which are then at the wafer level stack together. So this is showing, exploding new. And so, for example, this will always have the below. And so, as you stack them, you get this, this possible interference, for example. 38 00:07:39.550 --> 00:08:05.610 STFC-RAL-CR03: And the and if you're talking about an image sensor. Many agencies are just memory chips. I mean, you know, you can get to very, very large memory chips and stack in this way. But for for imaging systems, then obviously the top one, it has to be the one where the activity is. And so typically in a camera you'll have micro lenses, and you'll have the color filters. 39 00:08:05.610 --> 00:08:23.790 STFC-RAL-CR03: and we'll decide negotiation. 40 00:08:23.790 --> 00:08:44.860 STFC-RAL-CR03: And so this shows you just one example of how that can be done, and the way you make connections through it via Tsv's which, of course, that that's an art in itself, because these have to be electrically isolated, and the silicon, and so on. So there's a huge amount of technology in those, but it's all come good 41 00:08:44.860 --> 00:09:01.139 STFC-RAL-CR03: to a very great extent, and that is very much thanks to a group that certainly and thanks one individual in that group. So of course, if you're talking about imaging devices, the visible light, and also X-rays. If you make this layer thick enough 42 00:09:01.140 --> 00:09:24.370 STFC-RAL-CR03: function, you got 43 00:09:24.440 --> 00:09:27.759 STFC-RAL-CR03: so the simplest tag is one where you've just got 44 00:09:27.770 --> 00:09:32.950 STFC-RAL-CR03: 2 ways that's face to face, and and on you go. 45 00:09:32.950 --> 00:09:54.379 STFC-RAL-CR03: Now, when you look at that, you're probably thinking, who what's that? What's that gonna do in terms of. And there's an important detail which I'll I'll mention on this sort of in the text. Here, let me just see what is so here? Yeah. So so the thing which is is the fact that by this, by using this method, you can have a 46 00:09:54.380 --> 00:10:17.899 STFC-RAL-CR03: huge amount of processing power within the area of the just just like the retina. Right? And so you can have in the case of X-ray diffraction, you could do all sorts of of imaging image extraction from the data that you see. So you don't have to ever think about taking the raw data off. That's all process in the stack. 47 00:10:17.990 --> 00:10:30.310 STFC-RAL-CR03: And, for example, for imaging systems, for vehicles and things like emergency braking. And so you get faster response than if you have to start reading things out to some extent also. 48 00:10:30.310 --> 00:10:54.719 STFC-RAL-CR03: So normally, we don't need stitching. I don't know if if you're not stitching it, but we come to that if you don't but what you do usually need is these should be 4 side possible, so that you can make a raise, say for X-ray detection, and so on. That are the size that you need by having 4 side. But that is easily done, because there's no Wi-fi. You always go with Tsv's in the department. You can then make all your 49 00:10:54.720 --> 00:11:02.200 STFC-RAL-CR03: connections within the area of the of the sensor. So that's a. That's a very nice geometrical simplification. 50 00:11:02.220 --> 00:11:04.339 STFC-RAL-CR03: So 51 00:11:04.480 --> 00:11:22.290 STFC-RAL-CR03: the the other thing to say, because I said about crosswalk problems in seagos circuitry, you frequently get difficulties. Due to current flowing in the substrate, and things of that kind, because you've got analog and digit. So all all on the same complex chip 52 00:11:22.290 --> 00:11:41.209 STFC-RAL-CR03: whereas whereas in this case, another different players. And what you can do is what show here as well as active secretary. And then, for example, have a whole metal screen layer on top there so and that. How can you do all that, and then maintain our attachments to each other. 53 00:11:41.210 --> 00:12:05.839 STFC-RAL-CR03: and the method in which you which you do, that I have to be extremely accurate and flamerized, and you use a method called that machine to achieve that which was has been in the Cboc business for quite a number of years now. And so finally, the thing is, once you make a complete assembly, then it may be sufficiently robust that you could also take material out of here that you could send that one from the back. 54 00:12:05.840 --> 00:12:21.009 STFC-RAL-CR03: So back, then, in either, before assembly of these guys and of this one after, you can then get an appropriate overall status that if that's important. It's okay, not important. You can leave a thing. 55 00:12:22.120 --> 00:12:23.440 STFC-RAL-CR03: Sorry. 56 00:12:25.200 --> 00:12:51.090 STFC-RAL-CR03: So here you see the real thing. So this is, I think this is from tech insights that were just sliced through the imaging area of one of the Sony. And this is from 2 years ago. This is, I wrote a certain career article, and this is one that's shown there. And so here's the way for the wafer bond. Now, these are first thing to observe is that these are copper copper bonds, and you you cannot see any imperfection in bloodline. It looks absolutely perfect. 57 00:12:51.100 --> 00:13:12.900 STFC-RAL-CR03: And it took the stone of Ayashi. Who is the guy behind all this. It took him 2 years of failed tests before he got that. That's just one of the long. He's been on this seat for for probably 15 years working on this with his group. And so that was one of the features getting the alignment, the way for the way for alignment. The scale is shown here. So you work about 8 microns year above 58 00:13:12.900 --> 00:13:32.480 STFC-RAL-CR03: microbes. And what you're seeing on microlenses, color filters. These are deep trench isolation to keep the the signals within the pixel, and then the channel stops metal layers and and the first level 59 00:13:32.480 --> 00:13:35.330 STFC-RAL-CR03: frontier. 60 00:13:35.600 --> 00:14:01.239 STFC-RAL-CR03: amplifier on that that stack. And then these connections are all outside the imaging area. You notice that there are. There are things here layers of of metal for shielding, and actually, I haven't yet found out. Anybody who knows what this is at. The point is, you can see that these are quite ropey. I mean, you could never make 61 00:14:01.240 --> 00:14:19.180 STFC-RAL-CR03: a a connection. If you have this side of service, so it's their machine which does that. So we'll we'll have a look at that feature. And so this is a 2 tier. So you see what's used in the art. 2 years ago. This is what that machine is. 62 00:14:19.690 --> 00:14:40.240 STFC-RAL-CR03: The name is comes from Damascus, because this was developed very much in Damascus, in. I can't remember eleventh century something like that, but it's actually far older than that here. You see that the Chinese had this 300 BC. And so what it is is. This is a bronze vessel with gold and silver inlay. Silas obviously told it. 63 00:14:40.320 --> 00:15:04.639 STFC-RAL-CR03: And here you see a more modern version. But what it is is that how you make this is that you engrave the pattern that you wish to to have, and then you cover the whole. In this case you cover the whole thing with gold, and then repeated, and this is on oxidized steel. And so in this case, you get the pattern that appearing after. And that's exactly what's done 64 00:15:04.640 --> 00:15:29.160 STFC-RAL-CR03: with the metal layers. These multi metal layers of where you have the where you have one layer after the other is is ground and polished flat, and so the circuitry appears. You first go to middle of the whole thing, and then the circuitry appears by by the process of of, and in the same in the same time. You then get a complete 65 00:15:29.160 --> 00:15:42.099 STFC-RAL-CR03: planar service. So when you do your next metal there, it's sitting on a planer surface. So you planarize all the way through, maybe 10 layers of metals technique. right? So these 66 00:15:42.160 --> 00:16:00.680 STFC-RAL-CR03: so so Japanese award these days, started 2,008 and so he's still working on developing next generation memory technology for these devices. 67 00:16:00.820 --> 00:16:20.530 STFC-RAL-CR03: In 2,022, they announced the first stack, Cis for the 2 layer imaging cheetah, which I'll show you in a minute. And so there you get more of the electronics away from the pixel, and therefore you get a more. You know, a larger area of pixel for the for the light collection. 68 00:16:20.530 --> 00:16:31.739 STFC-RAL-CR03: They're currently 19 camera models in the world using these frontier systems logged in by these people, and they have the fastest frame rate visible likely to buy 69 00:16:31.740 --> 00:16:42.469 STFC-RAL-CR03: and many frontier applications. Of course, and and 90% of the 70 camera output now uses that senses, and there's a quotation 70 00:16:42.890 --> 00:16:47.520 STFC-RAL-CR03: from the school of Archie about that work that he did. 71 00:16:47.870 --> 00:16:51.990 STFC-RAL-CR03: He says there's a a long standing golden rule at. So. 72 00:16:52.100 --> 00:17:10.350 STFC-RAL-CR03: if you have a good idea, pursue it without waiting for the boss's approval. If it works, make it official, if not, you can leave it there. This rule epitomizes the workplace environment that allowed Kumabayashi to take up and pursue new ideas until he succeeded, and it culminated in the Statsimos image sensor. 73 00:17:10.470 --> 00:17:11.710 STFC-RAL-CR03: and 74 00:17:11.839 --> 00:17:16.999 STFC-RAL-CR03: there are many hurdles. It could be that one of the 75 00:17:17.079 --> 00:17:46.229 STFC-RAL-CR03: now working towards that in our own field, like Campbell reported for side multiple time. Mix for chips, and that with the Csv from Institute and a lot of sloes. Who's working on stage detectors with with Alice is now, thinking of getting into the the the second feature as well. There's one foundry at least, which does offer both. 76 00:17:46.230 --> 00:17:50.550 STFC-RAL-CR03: so that should open many doors for our field. 77 00:17:50.570 --> 00:18:12.890 STFC-RAL-CR03: There wasn't any use at this workshop on the work of Greg Depp, of Brookhaven, and the Harvey Hutsifer Rican were both. He's working with Sony in terms of their work. But it is advancing. And there's obviously potentially a very large one. I think that Lcms, too, would be really tremendously opened up if they stopped 78 00:18:13.160 --> 00:18:22.580 STFC-RAL-CR03: so here's an example. So this is the conventional. That's the one I showed first the 2 layer one. So there you've got bio, and which contains these some 79 00:18:22.740 --> 00:18:45.090 STFC-RAL-CR03: pent. 4 PIN photon notes in the pixel, and this in the same. This is in the pixel. This is all one pixel, right with the the front end circuitry. And so then, when they went for this, for the 2 layer, which is the most recent development. Then the pixel transistors go below and you've got more area, more efficiency for the light connection. 80 00:18:45.120 --> 00:19:13.900 STFC-RAL-CR03: But of course, even there you're not at the same level with a Ccd. If you want the the highest photo detection efficiency, then you use a Ccd, but you then got the very slow readout. So it's still a bit of astronomical and other applications where the optical qualities paramount. But readup speed isn't an issue. So this is this is an indication that this in real life. So what I showed before you got the lenses, the the filters, defense, isolation, and then 2 layers 81 00:19:13.900 --> 00:19:20.230 STFC-RAL-CR03: so that that we minimize the electronics in the in the Pixel app 82 00:19:22.730 --> 00:19:30.139 STFC-RAL-CR03: there was a wonderful discussion by synopsis, a panel discussion which you can find on on 83 00:19:30.600 --> 00:19:51.080 STFC-RAL-CR03: Youtube and find the thing accelerating mainstream adoption of multi-dive systems, which is what we're talking about. There were representatives of the needless in the business, and you see the application areas. Automotive is there? We want 84 00:19:51.360 --> 00:20:05.659 STFC-RAL-CR03: the the scale of the activity, according to the chairman, is that there's 3 times more going on this year than last year. Sonopasters are currently tracking 100 of these multi dive projects worldwide. So it's really taken off 85 00:20:05.660 --> 00:20:30.310 STFC-RAL-CR03: in in the industry and and what are the main challenges that they see now? The big one that kept coming up in the discussions and need to drive standards, because at the moment. We have certain foundries such as Samsung, who do have standards available. So it's just like design rules when you're doing design and and so you can go in there, and you can work to those, and then 86 00:20:30.310 --> 00:20:44.189 STFC-RAL-CR03: your layer, whatever you design, can then be compatible with the people who are working on other layers. So it's much more opportunity for collaboration. 87 00:20:44.190 --> 00:21:09.960 STFC-RAL-CR03: And for example, there are certain universal chip that Internet express which is a wider standard and but day to day interconnects. And that's now, you see, a large number of companies are doing that. So the whole thing is becoming more standardized and and that people can come in. Small companies are coming in and working in that. So it may very well be an opportunity for our field and everybody yet doing it 88 00:21:09.960 --> 00:21:27.920 STFC-RAL-CR03: and altogether working. 89 00:21:27.990 --> 00:21:29.010 STFC-RAL-CR03: I said. 90 00:21:29.040 --> 00:21:53.719 STFC-RAL-CR03: so there's very enthusias moving chip centric to system centric system. And it's all just beginning so here's couple of things that very recent news. There's a a Cfx complementary field transistors the future transistor type. You see it here with the with the one another, and that is 2 reports. Intel and Tsmc. 91 00:21:53.720 --> 00:22:01.450 STFC-RAL-CR03: Are going to give a presentations at the Ied and the annual meeting which is taking place in in December. 92 00:22:01.520 --> 00:22:03.440 STFC-RAL-CR03: And 93 00:22:03.500 --> 00:22:06.970 STFC-RAL-CR03: look looking very good. 94 00:22:09.360 --> 00:22:34.099 STFC-RAL-CR03: Here's a couple more. This is a company I never heard of in by these 3 guys spring of this year started up, and it's developing civic and note the lasers to attach to a single chip bit to be in information between them. So this is allowing for coming more from copper to to optical, and obviously the huge advantages there. 95 00:22:34.100 --> 00:22:43.300 STFC-RAL-CR03: And another interesting thing is way, have so far been using sony chips for their camera sensors. And it's not a camera 96 00:22:43.300 --> 00:23:08.090 STFC-RAL-CR03: and and recently, because problems with the Us. Suppliers, and so on, working on independently. So there's a lot going on in the world. Cameras and and this one could have a 2, 432 megabit. So 430. Megabix. Okay? 97 00:23:08.090 --> 00:23:11.579 Well, there's there's there's a there's reason behind it. 98 00:23:11.610 --> 00:23:30.389 STFC-RAL-CR03: According to report, the sense will be one is cross feature 36 to one pixel Binning, the unimaginably good. No live performance. 36 people do that has great potential. So this this is not really 99 00:23:30.490 --> 00:23:35.110 STFC-RAL-CR03: okay. This is this is like Campbell's recent achievement is 100 00:23:35.170 --> 00:23:52.779 STFC-RAL-CR03: just probably very proud of it. So this is. This is his chip is so. This is a hybrid chip. But you see, it's 4 side possible, because of the fact that it connects with Tsp, so that's that's a big step. 101 00:23:53.560 --> 00:24:22.569 STFC-RAL-CR03: That's usually a wonderful talk. So he's very much linked into the at least stuff. And so the it s 2 is what we got here. That's currently taking data with longer the pixels of the Alpine type. And so it's 10 m² of silicon doing wonderfully well running and what they're doing now is that replacing the innermost 102 00:24:22.670 --> 00:24:43.620 STFC-RAL-CR03: engineering model, silicon. 103 00:24:43.620 --> 00:24:54.939 STFC-RAL-CR03: 28 cent means in length and 3 layers and and that is now in really well, that's done with the 104 00:24:55.570 --> 00:24:59.839 STFC-RAL-CR03: Tower gas panasonic, 65 nanometer process. 105 00:24:59.910 --> 00:25:09.509 STFC-RAL-CR03: and that is, that is 106 00:25:09.670 --> 00:25:23.560 STFC-RAL-CR03: stitch design. And it's pretty much way for style. These are these are, and that that unfold that that pretty well fills the way. 107 00:25:23.690 --> 00:25:44.550 STFC-RAL-CR03: Of course. It's 65 nanometers in the in the industry that push it down to 2 nanometer effective size measured by the transistor density and so we are. Do the job extremely well, and the phone is expensive. 108 00:25:44.570 --> 00:25:46.720 STFC-RAL-CR03: and if you went to 2 now. 109 00:25:46.900 --> 00:26:03.849 STFC-RAL-CR03: this is some. There's a very interesting Youtube on this top of your Asml out both of Philips and you probably many, many of you will will know that. But basically they've now gone to the extreme UV 110 00:26:03.850 --> 00:26:25.710 STFC-RAL-CR03: and previous and that is to get down to it's a large truck. 111 00:26:25.710 --> 00:26:50.980 STFC-RAL-CR03: but I hadn't realized is that it's not. It's in the Netherlands, but it's sound for the Phillips. But there are 30 proponent machines assembled here, which are gathered from the subsidiaries around the world, and and then and then put together, and commissioning is done, and then disassembled, and ships the customer 20 trucks. 112 00:26:51.060 --> 00:27:19.399 STFC-RAL-CR03: 3, 7, or 7 transport base, and a cost of 200 years, the custom. So and and they? And they said, I think that the market slowed down a little bit, but they expected to deliver 30 units during this year. Incredible! A company! As as you probably know, that the export to China is currently got. And so what are the Chinese? Well, they're not. They're not sitting on their hands. I mean, this is a bit longer term. But basically. 113 00:27:19.400 --> 00:27:31.299 STFC-RAL-CR03: what's going on right now is that they're talking about and this is idea that I'm actually from slack. This is and Daniel 114 00:27:31.300 --> 00:28:01.200 STFC-RAL-CR03: accelerating by, and they they micropunching storage ring. And so this is a storage ring in which they box and emulators which feed various types of activity, one of which is, let's say, many lithography machines. They don't. Actually, this is a substitute for that. Some euv but it's generated as a as a central life, and so 115 00:28:01.230 --> 00:28:02.530 STFC-RAL-CR03: hope it, then 116 00:28:02.750 --> 00:28:27.710 STFC-RAL-CR03: this is in the future I look, but it's only been programmed compared with Current Assembly, SSD would be a more ideal life source, higher average power and higher chip production output with lower unit costs. They say they can get 100 watts to a kilowatt on the wafer compared with 5 watts from Asnl. So I mean, the whole area is is moving forward in a very interesting way. 117 00:28:28.800 --> 00:28:34.310 STFC-RAL-CR03: So Petra Merkel gave a very nice keynote. Talk on particle physics. 118 00:28:34.560 --> 00:28:59.520 STFC-RAL-CR03: she shows here link to a present and and so these are smart pixels, and the idea is, is is properties. Angle Pt, instead of raw data, uses data, rates, manageable levels, uses AI to perform physics, motivated data reduction on Asic. So then. 119 00:28:59.520 --> 00:29:05.759 STFC-RAL-CR03: many of the ideas that some that we were talking about in stack. But this is not in a set environment 120 00:29:05.760 --> 00:29:26.710 STFC-RAL-CR03: and so th they are going to do this. And of course, I think it's true that if you went to a stack architecture should be much easier to move forward, for example, for the Fcc. So again, it's one of those things where you see that moving forward, another example from 121 00:29:26.820 --> 00:29:30.300 STFC-RAL-CR03: from her talk was that some are at this 122 00:29:30.430 --> 00:29:46.140 STFC-RAL-CR03: the which is, of course, used to get pilot suppression in the highly lost operation, 30 to 50 pixel, but 30 to 50, 50 even 123 00:29:46.800 --> 00:30:01.250 STFC-RAL-CR03: They found that they needed to work on the radiation hardness and something that I had to go about before carbon enrichment. You could read about it in the in the slides. Which is helped and they have a nasty problem 124 00:30:01.250 --> 00:30:24.500 STFC-RAL-CR03: single event in the in the in the chip and so that was holding them back for a while, but they found that they could reduce the operating voltages and get around that and so they went up to 10, and there was a U paid company and 125 00:30:24.630 --> 00:30:42.609 STFC-RAL-CR03: but it all went to 3 Chinese companies who could produce them on the Timescale at very fast production. And so they got the contract and starting for production next year. So I don't know what, so less from that. But that's certainly interesting to me that all all 3 126 00:30:42.610 --> 00:31:00.250 STFC-RAL-CR03: Americans also in, and I think Japanese and so these are th. This is, of course, just the the censorship, and they read up with Asics from Tsn in the in the usual way for hybrids. So there! There are 127 00:31:00.250 --> 00:31:22.500 STFC-RAL-CR03: 2 double layers of detectors. And then, all of a sudden, this is all coming from the the the Chinese part of the app collaboration, which is very, very compact system, because, of course, it has to squeeze in between the track and the Eco and space available. So that's a very challenging engineering piece of work. 128 00:31:22.830 --> 00:31:41.470 STFC-RAL-CR03: This is where accidents. But often 129 00:31:41.470 --> 00:32:05.569 STFC-RAL-CR03: people don't talk much about them in public. And so lessons may not be shared, and some projects just get abandoned. So I do a few examples here, 2 of which are my own experience. So I can. I don't know who worry about attending myself. The the first. When Pete first tried in the 19 130 00:32:05.620 --> 00:32:31.099 STFC-RAL-CR03: 79, I think it was to get. There was skepticism, people work at all, and so on, and we had space in the to try out the telescope with 3 Ccds. And we thought you know, we'd done optical tests before we promised. And then, just before we use startup 131 00:32:31.250 --> 00:32:38.040 STFC-RAL-CR03: we had some mechanical work that needed to be done inside the price that on on the system. 132 00:32:38.050 --> 00:32:45.480 STFC-RAL-CR03: and whereas the people working with the Ccds which was herself Fred Wicken's 133 00:32:45.560 --> 00:33:07.899 STFC-RAL-CR03: Steve Watts, and that was probably end of that time. Knew about the careful handling, that extremely static sensitive. There was no static protection on those devices. How careful we had to be. We hadn't conveyed that information to our mechanical technicians who were then going in to do this work, and so 134 00:33:08.290 --> 00:33:15.479 STFC-RAL-CR03: what we were thinking I don't know, but they just went in, and they unplugged these things, did the work, put them back together, and of course they're all dead. 135 00:33:15.480 --> 00:33:40.200 STFC-RAL-CR03: and that was so, and Steve was furious about that, and sorry how open they haven't been told. And so from then on what we did, we changed our procedures, and we and we made the procedures that no work would be managed to get a few Ccd. For each to be in a hurry, and they work for improve. But the fact is that we then change that procedure so that we would 136 00:33:40.200 --> 00:33:43.989 STFC-RAL-CR03: always then that any work without detectives. This went on for 137 00:33:44.130 --> 00:33:48.180 STFC-RAL-CR03: for 2020 years after that 138 00:33:48.370 --> 00:34:13.059 STFC-RAL-CR03: and then work. And that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's that's 139 00:34:13.060 --> 00:34:28.709 STFC-RAL-CR03: to that type. So you know, that was that was a lesson learned. And I often have chatted with people, and they aren't quite that careful. Not that it's not a bad thing to do, because, having rigorous procedures and making sure that followed and signed off is really important. 140 00:34:28.710 --> 00:34:40.810 STFC-RAL-CR03: The second example, of course, is simple fact is, that didn't 141 00:34:40.940 --> 00:35:10.619 STFC-RAL-CR03: protect base. The fact that one might just be forgotten from time to time, because there was always just be glad to leave the tunnel because of some sort of along and so and the first production project, which is not yet working. But it's a smack. And being assembled. 142 00:35:10.620 --> 00:35:28.999 STFC-RAL-CR03: And what happened there was that that contract? They produce prototype Ccds, which were absolutely, beautifully. And then the Do. We put whole thing on hold for more than a year because funding issues. And then they started up. Okay, finally, go back to production. They couldn't make them anymore. 143 00:35:29.000 --> 00:35:53.410 STFC-RAL-CR03: did not work. And so I was at that time retired to me on because they were part of that project, and ages, 75 and I was in starting to attend these weekly meetings, and it was a bit strange to me, because, the meetings were always focused on tiny design details to improve the design. And I was just sitting there thinking. 144 00:35:53.410 --> 00:36:08.499 STFC-RAL-CR03: this can't be. Hey? That work before, right? You didn't change the design. It was quite clear it was the same design. And so oh, no! Something was being on edge, and so if they be tweaking the design, and they brush through another batch. And, sure enough, exactly the same problem. And so 145 00:36:08.680 --> 00:36:20.599 STFC-RAL-CR03: The answer to that, I will. I will explain on next slide. It was you shouldn't focus on your design. Anything, change there. And when 146 00:36:20.970 --> 00:36:34.439 STFC-RAL-CR03: actually, the guy in charge of it went on holiday. And so I dashed over to and talk to the production guys and said, Look, think of anything in the production change. And there were several things, and we rapidly managed to solve the problem 147 00:36:34.440 --> 00:36:54.179 STFC-RAL-CR03: and and then, of course, there was a simple back incident. The Lxcd. Already was reported that conference and since replace the box. But I presume that they will have much more robust procedures for that problem, because it was. 148 00:36:54.180 --> 00:37:01.910 STFC-RAL-CR03: It was 1 one of the more frankness. It was just just a screw where they they came. A remote. 149 00:37:01.910 --> 00:37:19.319 STFC-RAL-CR03: and so this is the nsst thing. So what happened there was that what the reason they didn't work was that when you put these Ccds and and tested the room temperature. They were okay. 150 00:37:19.620 --> 00:37:25.789 STFC-RAL-CR03: The more you cool it down when you expect the noise to improve. That's why you'd run them cold. The noise got worse. 151 00:37:25.840 --> 00:37:54.369 STFC-RAL-CR03: And what was that due to? Well, so what happened was that in the processing? Say before? You say you. But so if you want. If you want to minimize some some some part of the circuitry. So you will have, and you will then deposit the metal over the whole thing, and then etch that away, and now you've got your your pattern that you want so 152 00:37:54.370 --> 00:38:00.529 STFC-RAL-CR03: like we were talking about previously. Now. What was that? So when I went to talk to 153 00:38:00.530 --> 00:38:11.590 STFC-RAL-CR03: Production Manager and I said, Have you changed anything? He said, well, we've got beautiful new lamb etcher cost millions and very superior using. 154 00:38:11.590 --> 00:38:30.620 STFC-RAL-CR03: And so I said, how are we sure doing the job right? And he said, well, he said. We've got the other one lying in the corner. We could just try that on a on a so what was going on was that here when you thought you had 155 00:38:30.650 --> 00:38:59.310 STFC-RAL-CR03: I completely a a a completely good contact between the metal and the and the and the and the silicon in practice it wasn't. And so this is this, they made a little resistor, and so here they were measuring the resistance. When you put applied voltage across that. And so you see that for for you know reasonable voltages, you don't see any problem at all. It's basically a short circuit. 156 00:38:59.310 --> 00:39:09.779 STFC-RAL-CR03: But as you go to lower voltages and low temperatures, because here it's room temperature that's that's down here. And so I need 5 degree center plus 5 view 157 00:39:10.380 --> 00:39:17.279 STFC-RAL-CR03: plus 5 cent. Right is this one here? So it's only when you go. The resistance goes through the roof. 158 00:39:17.280 --> 00:39:40.659 STFC-RAL-CR03: And so what was happening there in the end was that there was some this, this this way, this extra work was, it's thought intelligently. When it came to the interface of the silicon. It was a bit too intelligent. So it's figured it stopped a few nanometers short, or put down some sort of extra deposit or something like that, and that was responsible. And so and in the end, when 159 00:39:40.660 --> 00:39:58.369 STFC-RAL-CR03: and ask how that actually solved the guy just said, Oh, we just tweaked a few parameters. So I think they talk to the manufacturer, but that project was almost being abandoned. They were going to give the production to an American company because each of you couldn't make the device. So that was a very close to the edge. 160 00:39:58.950 --> 00:40:17.579 STFC-RAL-CR03: right? And so now this vacuum incident of Lxcp was this was discussed at the at the conference, and you know there was. There was a an excess pressure between the inner vacuum and the outer vacuum, and the and the rf boxes like quote, deformed 161 00:40:17.580 --> 00:40:39.010 STFC-RAL-CR03: and in a way that some it was something like 80 of of distortion, so that meant that they had to operate with the thing open. Obviously couldn't do. Physics they would have been if they'd been in place. Shut down they will. They will replace that 162 00:40:39.840 --> 00:40:54.550 STFC-RAL-CR03: right? So celebrating the start of Lcns 2 that has gone extremely well, as I said, it's a big set of breaking down Lcms. The the 163 00:40:54.780 --> 00:41:12.110 STFC-RAL-CR03: light source that they've had since the start up of that work operates at the standard bunch frequency that you get out of the original Stat machine of 120 Hertz. And now they've got a a super connecting system 164 00:41:12.110 --> 00:41:35.549 STFC-RAL-CR03: replacement in the same tunnel, but actually not replacement, both work in parallel, and that's pushing up per Megahertz. So that's absolutely wonderful for science. And and opening up all sorts of windows in chemistry and all sorts of areas of science. Which? Which require that but of course it's an enormous challenge for the detectives, because, you know, got these images coming out at a 165 00:41:35.550 --> 00:41:54.830 STFC-RAL-CR03: and I can't resist showing you this. This is 1962, was trying to convince the the trustees of Stanford University to to let them build this 2 mile long machine, which was you know. considered outrageous at that time. 166 00:41:54.830 --> 00:42:10.999 STFC-RAL-CR03: And one of them asked the obvious question, what's the expected lifetime of this machine? And it's answers. Panofsky was the most honest guy you could meet, and he never tried to pull all over anybody's eyes, he wouldn't do straight on. So he said, well, I think the lifetime this machine be about 10 years. 167 00:42:11.000 --> 00:42:27.940 STFC-RAL-CR03: unless someone has a good idea, and that provides us. What's happened ever since. Almost every you can count, you know the sphere and and and and all the the activities that have gone on there decade by decade. And this is very prophetic. 168 00:42:28.250 --> 00:42:43.319 STFC-RAL-CR03: right? So the sort of thing that they will be able to do. They really can get this simulation? Obviously, on what they call molecular movies, which can actually see the the movement of molecules. Typically. 169 00:42:43.500 --> 00:42:51.820 STFC-RAL-CR03: for example related to to light photosynthesis, retinal functions in which you have a stimulation with a very short 170 00:42:52.030 --> 00:43:15.949 STFC-RAL-CR03: I should like. So that's really this is no price comes in and pro by a short burst of X-rays following, and then you can vary the delay, and as you vary the latency, evolution, and these molecules with most extraordinary things as a as a function of time after they've been stimulated. And and and the eventual AI is to do. They do this now on 171 00:43:16.010 --> 00:43:17.280 STFC-RAL-CR03: crystals. 172 00:43:17.330 --> 00:43:26.880 STFC-RAL-CR03: tiny crystals, nanocrystals that call and and the aim. So it's never shorted in terms of ambition. 173 00:43:26.880 --> 00:43:50.350 STFC-RAL-CR03: The aim is eventually to do it for Svr single particle imaging. So these would be molecules that are floating around in a in a, in a, in an atmosphere helium, probably, and from time to time this very intense punch comes through and hits this poor molecule, which, of course, immediately destroys it. I mean, it's Coulomb Coulomb explosion, and that strips out all the electrons 174 00:43:50.350 --> 00:44:19.749 STFC-RAL-CR03: and but if the bunch is short enough, then you get the diffraction image before any movement takes place. And so this is destruction, diffraction, destruction, and that is a goal which I have no idea. So on my way. So there is a very good group developing the Texas all this stuff, and through the Lcns 120 notes. They had these various 175 00:44:19.750 --> 00:44:31.300 STFC-RAL-CR03: generations, and so on, with various some capabilities, and I've never got a new family detective which are much more 176 00:44:31.300 --> 00:44:52.599 STFC-RAL-CR03: adventures and and reach towards the requirements. But II can't go into detail, but some my own feeling is that this is not yet near what's actually needed, and that's where I think the stack. The Texas will come in and and really help some hoping that that may well be. And and there there are people working on that 177 00:44:52.600 --> 00:45:22.649 STFC-RAL-CR03: which is working on that. And there's a conference annual conference workshop which is coming up, and that will probably report there on their progress. That sort of get the funding from that. It was tried more than a decade ago by a group at at Zoom. 178 00:45:22.650 --> 00:45:43.340 STFC-RAL-CR03: Talk about how close they couldn't quite get it going and so they've already had one attempt, and with all the work that's going on in in industry, it looks like this is a very good time to be pushing again. And of course the Us. Have got this Chips and Science Act funding 179 00:45:43.350 --> 00:46:02.110 STFC-RAL-CR03: nominally 280 billion over the next 5 years, and that they should have had 24 billion this year, and I've heard it's already probably cut down to 19. But that's not. That's not 0. And so, you know, there's a lot of activity there. And there, I think various universities are talking about training 180 00:46:02.110 --> 00:46:16.489 STFC-RAL-CR03: 50,000 electronic engineers we're able to to to take up. This is on to huge huge. You know optimism about these things on there under the Biden administration. Don't worry. 6. Okay, this is a 181 00:46:16.860 --> 00:46:37.549 STFC-RAL-CR03: getting near the end. This is a why did people want this high energy? Electron being for for microscopy? The reason is that if you look at complete cells, here's a eukaryotic cell. Then they are quite thick in the middle. So the 300 kin 182 00:46:37.550 --> 00:46:59.839 STFC-RAL-CR03: electron energy is sufficient for doing the the imaging in this area. But once you get into the nuclear area, not so this you can do if you can get up to 3 Mev and so that is now underway in the various groups around the world. 183 00:46:59.840 --> 00:47:24.799 STFC-RAL-CR03: so that that's where you want to get the higher energy capability. And there was a there was a workshop on this recently, and the Uk are involved. I think this is a dog speaker. So this is a relative diffraction and imaging. And this is the machine which they are putting together. Proposed machine. I think it's 184 00:47:24.800 --> 00:47:28.900 STFC-RAL-CR03: and this is 185 00:47:28.900 --> 00:47:55.620 STFC-RAL-CR03: talked about in a context. So this is, this is sort of being and and but Cq acceleration, much higher gradients that could be reduced to a single cell of a Cq structure. So this, which is very promising from the point of view of of pushing this forward. 186 00:47:56.290 --> 00:48:14.999 STFC-RAL-CR03: Right? I can. I go for 10? Yeah. Okay, so Gamma ray very high energy, gamma, ray astronomy. Why is that interesting? There's a lovely talk from John Mappington, from Leicester, and various topics which you can, which you can see here. It's all built on the 187 00:48:15.060 --> 00:48:40.069 STFC-RAL-CR03: glass detective, which was constructed at slack while I was working there, which then got renamed phone detector. And that is the thing which, for example, found these these galactic center, and so on. And it's had a very a very good program, but it's limited to something like 300 gev. 188 00:48:40.090 --> 00:48:57.780 STFC-RAL-CR03: And now you want to go much higher. Why is it so interesting? Because gamma rays, of course, unlike other cosmic rays. Don't get deflected. So if you can get good pointing accuracy can see see the source was, and so things like these, but there are many which are 189 00:48:57.780 --> 00:49:11.209 STFC-RAL-CR03: a much more aggressive. So there's there's an international activity going on. Take this forward. And he described 2 big projects. One is this southern wide field, Gabarus and the SW. 190 00:49:11.330 --> 00:49:35.830 STFC-RAL-CR03: In South America, and one of them another before new version of that which is in zoom he was a bit chees up. SCSI haven't yet decided to join either those with his arguments 191 00:49:35.830 --> 00:49:50.330 STFC-RAL-CR03: really interesting. I was confused by his presentation, because at first I thought he was saying that all these things have to be an extremely high altitude, and he mentioned 4 kilometers. I think that is applying only. 192 00:49:50.360 --> 00:50:03.630 STFC-RAL-CR03: and so the 193 00:50:03.630 --> 00:50:28.410 STFC-RAL-CR03: the 194 00:50:28.570 --> 00:50:49.389 STFC-RAL-CR03: telescope system that can be lower down, and that, indeed, I think, is the case. This is at about a 2 kilometer altitude and and only the the water show copies the needs to be higher up. And that's another view of the of the workshop, got sites thought to be a sites still to be selected. 195 00:50:50.160 --> 00:51:06.430 STFC-RAL-CR03: I put this up because this is these were people that that attended the conference, and I'm kind of mentally apologizing so, more because I hardly talk about any of their work, because it was just. It was mind blowing the amount of some detail and detailed usual work. But it's all there in response. 196 00:51:06.630 --> 00:51:27.110 STFC-RAL-CR03: you know, areas of your then the following week, I couldn't resist to just put in a couple more slides because there was a fest or severe soccer. Now, you know, of course, everybody has these fests, and some of them but university people, and usually it's, you know, half a day of talk. It's a 3 day meeting. 197 00:51:27.110 --> 00:51:51.910 STFC-RAL-CR03: Souvia is so you know, so famous and a and loved by people all over the world, and so they had easily filled 3 days of talks on this, which is excellent. So, and of course he noted for me highly skeptical. So Dan Hooper, you know, on this line of skeptic sumer. Oxford is one of the, and of course the thing which he's particularly been skeptical about is dark energy. 198 00:51:51.980 --> 00:52:07.149 STFC-RAL-CR03: He believes that dark energy in the Nobel Prize given for dark energy discovery may be incorrect, that there isn't any such single stock energy, and so on, and he will explain that as much detail as as you want, and he hasn't been through wrong. Obviously. No. 199 00:52:07.180 --> 00:52:29.579 STFC-RAL-CR03: the Physics department tried to prevent him publishing his paper because it was so offensive to the people at Berkeley. I mean that that censorship. Can you imagine that university, anyway. So you have. This is the thing about primordial black holes as a source about matter. What fraction could they be up here would be if it's 100% without matter. 200 00:52:29.580 --> 00:52:49.550 STFC-RAL-CR03: Now in in this mass of the primordial that called, if that too light, then they would have lost since the big back so. But then there's this window here where they could be the whole thing. And then, as you go further up, micro, I think the micro 201 00:52:49.550 --> 00:52:55.240 STFC-RAL-CR03: that if you look through our galaxy at south beyond 202 00:52:55.310 --> 00:53:23.129 STFC-RAL-CR03: then if you've got heavy black holes that you would see the gravitational lensing to to to obviously. And so that rules that I don't know why. So these are. See the master. 203 00:53:23.740 --> 00:53:40.389 STFC-RAL-CR03: Then there was this, finding there was this talk by William Edmonton, Cambridge, and he was talking about the Gaia satellite, which is again one of these telephone E. 2 v. Beautiful telescope system in space at the Lagrange Point, and they they measured 204 00:53:41.360 --> 00:54:08.510 STFC-RAL-CR03: 10 to the 9 Milky Way stars. That's a good percentage of the Milky Way stars, and they've observed each of them 70 times over. The 5 years they've been running the thing has exceeded its life. It's still operating. It was supposed not to be supposed to be dysfunctional by now, but it's still operating, and they only keep it going until 20 55 but so we showed that. And then we showed this slide. Which 205 00:54:08.620 --> 00:54:13.669 STFC-RAL-CR03: what's this? So here's the business. 206 00:54:13.690 --> 00:54:14.880 STFC-RAL-CR03: and this 207 00:54:19.420 --> 00:54:23.129 STFC-RAL-CR03: hasn't that accepted. To me it looks like he took a photo of his. 208 00:54:23.380 --> 00:54:42.650 STFC-RAL-CR03: Oh, nothing. Okay. 209 00:54:42.650 --> 00:55:05.300 STFC-RAL-CR03: Test. And so it was a bit light hearted, but nobody queried it. So so I'm I'm really. I personally have an immediately got no reply. So where we really know is that silent and all that is is extremely beautiful. 210 00:55:05.510 --> 00:55:19.319 STFC-RAL-CR03: You know. How fast is it? 2, 25 million years period. And so the last time system was weird is how the dinosaurs were only starting to arise, takes a long time to make a circuit of that 211 00:55:19.460 --> 00:55:24.589 STFC-RAL-CR03: on the so that's pretty well. Oh, oh, yeah. 212 00:55:24.600 --> 00:55:41.920 STFC-RAL-CR03: Sue Mei's punchline after the so he said very little, and there were all these wonderful talks, and he did say some words, and and more or less what he said. He said, well, it's wonderful. I have a job where every day to get out of bed, hoping to prove that everything that went before is completely wrong. 213 00:55:41.920 --> 00:56:07.839 STFC-RAL-CR03: Yeah, it's a lovely attitude, but I thought, I, my grandson, when he was an 11 month old baby. I remember he he was trying to walk, and he kept falling over, and so on, and one day I could almost see the glint in his eye, and he he thought he got it from whatever he tried the day before he did, and he started walking, and he was so chucked about it. And and there's another example here which is is wonderful. In my opinion. This is up for my former boss 214 00:56:07.890 --> 00:56:31.280 STFC-RAL-CR03: and after about 6 months he says, I think I'd like to try walking. 215 00:56:31.440 --> 00:56:37.539 STFC-RAL-CR03: Chris Icon. I have no authority really thinking with his broken back. And so I said, Well, if you like, have a go. 216 00:56:37.540 --> 00:57:02.290 STFC-RAL-CR03: Ambition allowed a year ago. 217 00:57:02.470 --> 00:57:12.880 STFC-RAL-CR03: but 90 92, 93 but so II thought that was that was nice punchline from Suvir, and we can all take 218 00:57:12.970 --> 00:57:16.759 STFC-RAL-CR03: That's it. Okay, thank you, Ben.