1 00:00:22,000 --> 00:00:25,360 Well, hello and welcome back to Impact Quantum, the podcast for the quantum 2 00:00:25,360 --> 00:00:29,090 curious. We. We firmly 3 00:00:29,090 --> 00:00:32,570 believe you don't need to be a PhD, although it certainly helps 4 00:00:33,050 --> 00:00:36,450 to participate in this emerging field. And with me is the most quantum 5 00:00:36,450 --> 00:00:39,770 curious person I know, Candice Gahooley. How's it going, Candice? 6 00:00:40,010 --> 00:00:43,130 It's great. Thank you so much, Frank. You know, today 7 00:00:44,250 --> 00:00:47,930 it's crazy. This week actually we got a massive, massive snowstorm. 8 00:00:48,570 --> 00:00:52,290 And although where I am in Montreal, Quebec, we always see the 9 00:00:52,290 --> 00:00:56,130 first flakes for Halloween, we normally don't have such 10 00:00:56,130 --> 00:00:59,290 a deluge. I probably have nine inches outside. 11 00:00:59,830 --> 00:01:03,310 It's, it's just crazy. So I'm all looking at the beautiful 12 00:01:03,310 --> 00:01:06,870 snow, but I'm getting my, my head all in gear to have a great conversation 13 00:01:06,870 --> 00:01:10,390 today. I'm really excited about our guest. Yeah, awesome. 14 00:01:10,870 --> 00:01:14,670 It is chilly down here. We didn't get any snow, although I think in Western 15 00:01:14,670 --> 00:01:18,270 Maryland they did get some snow, but it is chilly here 16 00:01:18,270 --> 00:01:22,030 and I may have to fire up the GGX to do some fine tuning 17 00:01:22,030 --> 00:01:25,510 to heat up my office as well as turn on some of these monitors. 18 00:01:26,750 --> 00:01:30,030 So who do we have talking to us today? Candace. Right, so today we're talking 19 00:01:30,030 --> 00:01:33,630 to Mahmoud Sabuni. He is the lead 20 00:01:33,790 --> 00:01:36,590 quantum processor engineer at 21 00:01:37,470 --> 00:01:41,310 oqd. Very cool. Hi, how are you today? 22 00:01:41,390 --> 00:01:45,230 Okay. Yeah, thanks for having me here. 23 00:01:45,630 --> 00:01:49,430 Yeah. Here also we have a little bit of snow, like last couple 24 00:01:49,430 --> 00:01:53,230 of days still like leaves on the trees, but you could 25 00:01:53,230 --> 00:01:56,810 see the snow and then snow and then after that you see the 26 00:01:56,810 --> 00:02:00,570 leaves on top of the snow. That's kind of interesting feature that 27 00:02:00,570 --> 00:02:04,010 you could see. So you're also in Canada, right? Yeah, yeah, in 28 00:02:04,010 --> 00:02:07,850 Waterloo, Canada. Okay, very cool. For the people who know, don't know, 29 00:02:07,850 --> 00:02:11,650 like Waterloo, where is it? Like close to the Toronto, like 70 or 80 30 00:02:11,650 --> 00:02:15,490 kilometers south. Yes, that's. 31 00:02:15,890 --> 00:02:19,170 I'm here from probably almost like 10 years. 32 00:02:20,130 --> 00:02:23,930 Start by like some postdoc and some other activity at 33 00:02:23,930 --> 00:02:27,600 Google as the optical engineer and then back 34 00:02:27,600 --> 00:02:31,120 to this open quantum design startup 35 00:02:31,600 --> 00:02:35,280 in Waterloo. Oh, very cool, very cool. Your 36 00:02:35,280 --> 00:02:38,840 LinkedIn is very impressive. So I have plenty of questions around 37 00:02:38,840 --> 00:02:42,080 that. Open quantum design. 38 00:02:44,160 --> 00:02:46,400 Tell me about that. That's an interesting. 39 00:02:47,840 --> 00:02:51,120 What does open quantum design do? Or. Yes, that's. 40 00:02:51,440 --> 00:02:54,720 Yeah, that's probably the core of the Discussion that we can go through. 41 00:02:55,680 --> 00:02:59,480 First of all I'm more like a hardware person. Like I worked more on the 42 00:02:59,480 --> 00:03:03,120 hardware side on like quantum information since my 43 00:03:03,120 --> 00:03:06,320 PhD which was in in Sweden in Europe 44 00:03:06,960 --> 00:03:10,800 and also my master there around the quantum 45 00:03:10,800 --> 00:03:14,320 information technology and like storage 46 00:03:14,560 --> 00:03:18,320 basically quantum memory during my PhD and after after 47 00:03:18,320 --> 00:03:22,160 that I came to the like quantum computing parts 48 00:03:22,640 --> 00:03:26,320 and as a postdoc here at iqc. 49 00:03:26,960 --> 00:03:30,480 And yeah I just put some gap for the 50 00:03:30,640 --> 00:03:34,320 Google time which was more classical optics and hardware. But later 51 00:03:34,640 --> 00:03:37,920 like there is some three PI here in Waterloo. 52 00:03:38,240 --> 00:03:41,880 Two of them Crystal Senko and Raj Wolselm working on the 53 00:03:41,880 --> 00:03:45,650 Ion Trap machine and one Roger Molko 54 00:03:45,650 --> 00:03:48,810 working more on on more AI and software side. 55 00:03:49,210 --> 00:03:53,010 They decided to start making like an 56 00:03:53,010 --> 00:03:56,330 iron trap based full stack quantum computer. 57 00:03:57,690 --> 00:04:01,290 This company start from like a February 24th and 58 00:04:01,370 --> 00:04:04,890 I joined them at the at the same time then we 59 00:04:05,050 --> 00:04:08,410 the idea is to have a full stack quantum 60 00:04:08,800 --> 00:04:12,360 computer based on iron trap available for anyone who 61 00:04:12,360 --> 00:04:15,360 wants to rebuild it. Like the 62 00:04:16,000 --> 00:04:19,720 way that we do is that we have like some software team 63 00:04:19,720 --> 00:04:23,560 and some hardware team. I'm more on the hardware side and we are 64 00:04:23,560 --> 00:04:27,360 developing like some prototype and at the same time we 65 00:04:27,360 --> 00:04:31,200 put all our designs on GitHub available 66 00:04:31,360 --> 00:04:35,200 for like anyone who wants to rebuild or use 67 00:04:35,200 --> 00:04:38,540 that kind of module for for his setup in future. 68 00:04:39,260 --> 00:04:43,100 And the the big picture is to like do the same thing 69 00:04:43,100 --> 00:04:46,860 that happens for either software or some even hardware 70 00:04:46,860 --> 00:04:50,620 in the classical computers in the quantum 71 00:04:50,620 --> 00:04:54,300 computers. Like there are some like you can explain more details like 72 00:04:54,300 --> 00:04:58,060 what kind of reasonings is behind this will be successful or not 73 00:04:58,060 --> 00:05:01,660 that. Yeah we don't know but we're pushing that. And during 74 00:05:01,660 --> 00:05:05,230 last year we could reach to some good milestones. 75 00:05:05,390 --> 00:05:08,950 We could get some collaborators like 76 00:05:08,950 --> 00:05:12,510 partners. We have like five different companies already 77 00:05:13,070 --> 00:05:16,550 putting money or full time employee on this open source 78 00:05:16,550 --> 00:05:19,550 activity and we are pushing towards 79 00:05:20,430 --> 00:05:23,790 getting our first aspect of the machine out soon. 80 00:05:24,590 --> 00:05:28,390 Hope like January 26th that shows okay this machine 81 00:05:28,390 --> 00:05:32,220 is already live and available and yeah 82 00:05:32,220 --> 00:05:35,820 looking for more people to come and contribute to see 83 00:05:35,820 --> 00:05:39,420 how we can push it forward. That's the more like 84 00:05:39,420 --> 00:05:42,380 a central activity of the open quantum design. 85 00:05:43,900 --> 00:05:45,340 So in practice 86 00:05:48,540 --> 00:05:52,300 what does open quantum hardware design mean 87 00:05:52,540 --> 00:05:56,380 in practice and how is it different from traditional 88 00:05:56,380 --> 00:06:00,220 closed or proprietary approaches in physics 89 00:06:00,300 --> 00:06:03,180 research? Yes like 90 00:06:04,300 --> 00:06:07,740 if we want to make very similar example in the 91 00:06:07,740 --> 00:06:11,260 classical world I could bring example of the 92 00:06:11,500 --> 00:06:14,300 RISC V company that's a company 93 00:06:15,340 --> 00:06:19,060 that actually we have some people from there also like 94 00:06:19,060 --> 00:06:22,700 with the same idea came to the one quantum design as a board member 95 00:06:23,260 --> 00:06:26,920 to Push the idea here also in 80s and 96 00:06:26,920 --> 00:06:30,560 90s there was like discussion about the 97 00:06:30,560 --> 00:06:34,200 designing of the CPUs. Like different 98 00:06:34,200 --> 00:06:37,760 companies like closed or open just start 99 00:06:37,760 --> 00:06:41,160 to work on that kind of architectures. And then 100 00:06:42,760 --> 00:06:46,360 at some point RISC V as an open source company 101 00:06:46,440 --> 00:06:50,040 start building the standards or 102 00:06:50,120 --> 00:06:53,720 designing chips. And these days like that 103 00:06:54,120 --> 00:06:56,840 the RISC V it has already 104 00:06:57,880 --> 00:07:01,560 70 plus members. And then most 105 00:07:01,560 --> 00:07:05,360 of the CPU design in Intel AMD anywhere else 106 00:07:05,360 --> 00:07:09,080 it's using those standards that's built on the open 107 00:07:09,080 --> 00:07:12,680 source activities. The same idea 108 00:07:12,680 --> 00:07:16,200 here also could be in principle 109 00:07:16,200 --> 00:07:19,550 implemented. Like for example 110 00:07:21,310 --> 00:07:24,990 benchmarking of the quantum computer is something 111 00:07:25,150 --> 00:07:28,910 which is very tricky. Like different company. Like 112 00:07:29,310 --> 00:07:32,830 +10 different companies already announcing different 113 00:07:34,030 --> 00:07:37,870 like machine. And then we don't have still 114 00:07:37,870 --> 00:07:41,590 fully standard benchmark system to do 115 00:07:41,590 --> 00:07:44,670 the benchmarking on different machine. And 116 00:07:45,230 --> 00:07:48,910 one of the reason is that those machines are closed and that 117 00:07:50,830 --> 00:07:54,230 no one have access to those machines to test and 118 00:07:54,230 --> 00:07:57,910 benchmark them. Then if they have some machine at 119 00:07:57,910 --> 00:08:01,710 least in some of the architectures like Ion Trap 120 00:08:01,950 --> 00:08:05,750 or in future could be other others also you 121 00:08:05,750 --> 00:08:09,470 can run your algorithm on that machine and 122 00:08:09,820 --> 00:08:13,500 tested machine and then test your standards and then write the standards 123 00:08:13,500 --> 00:08:17,340 based on those open access machine for anyone around the world. 124 00:08:17,340 --> 00:08:21,060 And then the rest of the people can also use that. This 125 00:08:21,060 --> 00:08:24,140 is not against any like a 126 00:08:24,700 --> 00:08:28,460 IP based company also like they could also get benefit of this 127 00:08:28,780 --> 00:08:32,380 open source and build the standard around that. 128 00:08:32,380 --> 00:08:36,140 Because around this like a point people can come 129 00:08:36,890 --> 00:08:40,170 and exchange idea and also develop 130 00:08:40,410 --> 00:08:43,290 whatever that they have done so far. And then 131 00:08:43,930 --> 00:08:47,130 integration of those like multi 132 00:08:48,410 --> 00:08:50,570 idea people can 133 00:08:52,409 --> 00:08:56,010 in principle maybe outperform 134 00:08:56,010 --> 00:08:58,410 the flow system. 135 00:09:00,490 --> 00:09:03,780 That's the idea behind this. But there is some other 136 00:09:03,940 --> 00:09:07,340 feature also would be interesting from hardware point of view. Because from 137 00:09:07,340 --> 00:09:11,020 software point of view you can see a lot of comparison 138 00:09:11,020 --> 00:09:14,260 around the board. For example Linux versus Windows 139 00:09:14,900 --> 00:09:18,740 like maybe the top 140 00:09:19,140 --> 00:09:21,540 like the most important project around the world 141 00:09:22,980 --> 00:09:26,780 running Linux, not Windows like Linux is the open source 142 00:09:26,780 --> 00:09:30,560 that's in the software. It's very clear that 143 00:09:30,640 --> 00:09:33,920 that's a path that can be very successful 144 00:09:34,400 --> 00:09:37,760 in the hardware regime. Especially in 145 00:09:37,840 --> 00:09:39,680 quantum computation. 146 00:09:41,840 --> 00:09:45,680 Very expensive hardware you need to develop. And then the open 147 00:09:46,480 --> 00:09:49,520 source community will have difficulty to 148 00:09:50,480 --> 00:09:54,320 gather those together and then test whatever they 149 00:09:54,320 --> 00:09:58,000 want to do on that hardware. The idea on OQD 150 00:09:58,000 --> 00:10:01,040 is to make that kind of test beds for 151 00:10:01,600 --> 00:10:05,120 anyone who want to work to single 152 00:10:05,120 --> 00:10:08,880 details of the machine. Like 153 00:10:08,880 --> 00:10:12,640 other machine that is available already mainly is cloud based. 154 00:10:13,120 --> 00:10:16,640 You can run some algorithm on those but 155 00:10:18,080 --> 00:10:21,840 you don't have access to full ingredient of that 156 00:10:21,840 --> 00:10:25,450 machine. Yeah, that's simply because of the IP reason. 157 00:10:26,090 --> 00:10:29,610 But here everything is transparent. You can see 158 00:10:29,850 --> 00:10:31,210 each single module. 159 00:10:33,530 --> 00:10:36,090 For that reason it will be easier to 160 00:10:37,130 --> 00:10:40,570 first benchmark it second also learn it 161 00:10:41,770 --> 00:10:45,370 from this hardware available. A lot of 162 00:10:45,370 --> 00:10:49,180 people like a software oriented that they are interested 163 00:10:49,180 --> 00:10:52,580 to run some calibration on some real quantum machine. 164 00:10:52,820 --> 00:10:55,380 They can do that and 165 00:10:56,500 --> 00:11:00,220 that's a big benefit on the learning curve. 166 00:11:00,220 --> 00:11:03,780 Like build some better workforce development. 167 00:11:04,260 --> 00:11:07,940 Or even for closed system they can use this kind of people 168 00:11:08,020 --> 00:11:11,700 who know it's capable of doing 169 00:11:11,700 --> 00:11:15,300 something which was not possible without access to the hardware. 170 00:11:16,680 --> 00:11:20,280 That's one big plus. And also 171 00:11:20,440 --> 00:11:24,160 like for example colleges that they 172 00:11:24,160 --> 00:11:28,000 want to train some quantum engineer at the bachelor level that 173 00:11:28,000 --> 00:11:31,320 they want to handle quantum machine. 174 00:11:31,640 --> 00:11:35,240 They can use this kind of system. Like they cannot 175 00:11:35,240 --> 00:11:39,040 go and buy some quantum computer from like big IP based 176 00:11:39,040 --> 00:11:42,890 company and then have access to all full details because 177 00:11:42,890 --> 00:11:46,090 it's IP based. But here they can have 178 00:11:47,050 --> 00:11:50,570 machine and then run it and learn it 179 00:11:50,570 --> 00:11:54,410 and improve it for whatever 180 00:11:54,410 --> 00:11:58,170 purpose that they have. This machine is not necessary to be the top best 181 00:11:58,170 --> 00:12:01,930 machine in the world. It's just need to run some simple 182 00:12:02,090 --> 00:12:04,010 functionality. And 183 00:12:05,850 --> 00:12:08,650 because of the reason that I mentioned like benchmarking, you can do it on a 184 00:12:08,650 --> 00:12:12,320 simple machine learning, you can do it on the simple machine. And 185 00:12:12,320 --> 00:12:15,920 also workforce development also you can. Do 186 00:12:16,160 --> 00:12:19,840 on this machine with the open design 187 00:12:20,560 --> 00:12:23,920 transparency, open source. Do you find that this 188 00:12:23,920 --> 00:12:26,880 accelerates innovation? That sharing 189 00:12:27,680 --> 00:12:31,320 these designs actually helps solve complex engineering 190 00:12:31,320 --> 00:12:34,960 problems, perhaps faster? Yeah, that's the idea. 191 00:12:35,520 --> 00:12:39,220 If you want to compare it in physics person. And then I 192 00:12:39,220 --> 00:12:42,540 was comparing to some phenomena in. In nature, like 193 00:12:42,700 --> 00:12:46,220 comparing the light bulb to the laser. Like the laser 194 00:12:46,220 --> 00:12:50,020 happened whenever that you have phase coherences of the 195 00:12:50,020 --> 00:12:53,460 photons on top of each other and they will 196 00:12:53,460 --> 00:12:57,180 amplify the the final result here also 197 00:12:57,740 --> 00:13:01,580 that could be a. Yeah, it's. It's difficult to prove 198 00:13:01,580 --> 00:13:04,780 it, but there are some example as I mentioned, like 199 00:13:05,360 --> 00:13:08,960 software developmentally and also hardware like this 200 00:13:08,960 --> 00:13:11,760 RISC V versus ARM or Intel 201 00:13:13,600 --> 00:13:17,400 that shows that was successful here also there's a good 202 00:13:17,400 --> 00:13:21,240 chance that can be successful and this comes 203 00:13:21,240 --> 00:13:24,440 from like University of 204 00:13:24,440 --> 00:13:27,920 Waterloo. Anyway they spend time on 205 00:13:28,000 --> 00:13:31,040 development different sections and different modules. 206 00:13:31,740 --> 00:13:35,580 And that would be good to share it with the academia first for 207 00:13:35,580 --> 00:13:39,340 sure. And then why not share it with everyone 208 00:13:39,980 --> 00:13:43,500 and then try to. Also we 209 00:13:43,580 --> 00:13:47,020 also could benefit a lot. Like for example we have 210 00:13:47,099 --> 00:13:50,860 some locking mechanism under our laser 211 00:13:51,180 --> 00:13:54,700 that we don't have much and bandwidth of the people 212 00:13:54,780 --> 00:13:58,420 to develop that. But someone else could come and say 213 00:13:58,420 --> 00:14:02,220 oh, I did this with all single details. Not 214 00:14:02,220 --> 00:14:05,980 just a paper, not the published paper or archive version, just with the 215 00:14:05,980 --> 00:14:09,700 details of the electronics, mechanics, optics, 216 00:14:10,340 --> 00:14:14,180 diagrams and share those information and then 217 00:14:14,180 --> 00:14:17,900 we can build that system in here. On the other hand, 218 00:14:17,900 --> 00:14:21,380 the other partner also can benefit from our whatever 219 00:14:21,540 --> 00:14:25,260 for example our optical circuit board design which is 220 00:14:25,260 --> 00:14:28,940 transparent, everything available and use that as 221 00:14:29,180 --> 00:14:33,020 like resources to build its own system 222 00:14:33,020 --> 00:14:36,220 in different country completely. 223 00:14:37,100 --> 00:14:40,380 Yeah, we have some collaborators like from India, 224 00:14:40,860 --> 00:14:44,380 from like Brazil, from a lot of countries 225 00:14:44,540 --> 00:14:48,380 in the south hemisphere like coming 226 00:14:48,380 --> 00:14:50,620 for this kind of options. 227 00:14:52,340 --> 00:14:56,100 And of course there will be also some difficulty in terms 228 00:14:56,100 --> 00:14:59,780 of politics that will be challenging to 229 00:14:59,780 --> 00:15:03,340 how, when and where you want to distinguish between 230 00:15:03,340 --> 00:15:07,060 these open and close and IP base. That's. That's the 231 00:15:07,060 --> 00:15:10,900 still open question that yeah, there is some resistance 232 00:15:10,900 --> 00:15:14,540 sometimes. So there's a lot to unpack 233 00:15:14,540 --> 00:15:17,630 there. But one of the things I saw on 234 00:15:18,190 --> 00:15:21,870 your website for OQD was 235 00:15:24,270 --> 00:15:27,710 you referenced the term full stack quantum computer. Yes. 236 00:15:28,190 --> 00:15:30,430 What is a full stack quantum computer? 237 00:15:31,870 --> 00:15:34,270 Full stack means like you can. 238 00:15:35,630 --> 00:15:38,510 The whole quantum computer is several 239 00:15:39,150 --> 00:15:42,520 layer of different information 240 00:15:42,920 --> 00:15:46,200 from your classical computer code 241 00:15:46,760 --> 00:15:50,440 and then that will convert to some mid layer and 242 00:15:50,440 --> 00:15:54,120 then the mid layer to very low level in the. We call it like a 243 00:15:54,120 --> 00:15:57,680 meta layer that you want to talk to the. In our case to 244 00:15:57,680 --> 00:16:01,400 atoms and there will be a lot of like exchange 245 00:16:01,400 --> 00:16:05,120 in between. In at open quantum 246 00:16:05,120 --> 00:16:08,770 design we have some partner that works on the 247 00:16:08,770 --> 00:16:12,370 software like for example Xanadu as a IP 248 00:16:12,370 --> 00:16:15,770 based company they provide some agreement to 249 00:16:16,090 --> 00:16:19,850 OQD and provide some full time employee they can 250 00:16:20,010 --> 00:16:23,450 adopt their like a software which is open source 251 00:16:24,570 --> 00:16:28,290 to our hardware which also is open source. 252 00:16:28,290 --> 00:16:31,770 Like we are working on OQD Aintrap using 253 00:16:31,850 --> 00:16:35,700 the Arctic software which is the open source hardware 254 00:16:36,500 --> 00:16:39,980 developed by. I think 255 00:16:39,980 --> 00:16:43,300 it's from Maryland actually. And 256 00:16:44,900 --> 00:16:48,660 these two combination could be as a full stack. 257 00:16:48,660 --> 00:16:52,260 Like you can run your code high level and then it will convert everything 258 00:16:52,340 --> 00:16:56,140 down to the. To the pulse level. Talking to atoms. Take the 259 00:16:56,140 --> 00:16:59,700 data and then plot the data, extract the information 260 00:17:00,650 --> 00:17:03,770 and use it. That's the meaning of the full stack. Like 261 00:17:04,250 --> 00:17:07,770 from high level to bare metal layer 262 00:17:08,170 --> 00:17:09,930 and vice versa, 263 00:17:12,410 --> 00:17:15,050 open to everyone. The software 264 00:17:16,169 --> 00:17:19,290 in most cases are available 265 00:17:21,370 --> 00:17:24,570 like in superconducting qubit in 266 00:17:25,290 --> 00:17:29,130 photonics and ion trap. It's a 267 00:17:29,130 --> 00:17:32,770 little bit lagging. Like we need some software at the high 268 00:17:32,770 --> 00:17:36,250 level. That's this kind of project trying to fill out that 269 00:17:36,250 --> 00:17:38,090 gap. And 270 00:17:40,170 --> 00:17:42,970 yeah, that's the meaning of the full stack. 271 00:17:44,010 --> 00:17:47,610 Interesting. Full stack is one of those terms you hear a lot in technology 272 00:17:48,090 --> 00:17:50,570 and it means different things to different people and 273 00:17:51,770 --> 00:17:55,050 it Means different things in different contexts too. So that's why when I saw that 274 00:17:55,850 --> 00:17:59,410 usually my reaction to the the term full stack it triggers me to 275 00:17:59,410 --> 00:18:02,650 flashbacks of conversations I had with recruiters throughout the years. 276 00:18:03,450 --> 00:18:07,010 So I'm like no, so 277 00:18:07,010 --> 00:18:10,770 so. So I like your definition better. So okay. Yeah, it is here is 278 00:18:10,770 --> 00:18:13,450 like a. Like a simpler version that. 279 00:18:15,290 --> 00:18:18,090 Yes, that's more reasonable. Version two. 280 00:18:20,730 --> 00:18:24,370 Yeah, yeah. Of course for like using the 281 00:18:24,370 --> 00:18:28,010 quantum computer you cannot directly go to the 282 00:18:28,010 --> 00:18:31,450 bare metal. You need some middle layer which is very 283 00:18:31,450 --> 00:18:35,050 crucial and still needs a lot of development 284 00:18:35,690 --> 00:18:39,410 connecting those hardware in terms of. Because one of 285 00:18:39,410 --> 00:18:43,050 the enemies here that we have in like. Okay. Whatever 286 00:18:43,290 --> 00:18:47,130 properties that it will make a lot of difficulty for 287 00:18:47,130 --> 00:18:50,780 us is the phase. Phase means 288 00:18:50,780 --> 00:18:54,340 like whenever that you have a wave you need to 289 00:18:54,340 --> 00:18:58,100 predict exactly at. In future time what is 290 00:18:58,100 --> 00:19:01,860 your amplitude. And this could jitter a little bit. 291 00:19:02,500 --> 00:19:05,460 And in our hardware side 292 00:19:06,260 --> 00:19:09,780 this will be crucial to control this phase. 293 00:19:10,580 --> 00:19:13,860 And in the radio frequency 294 00:19:14,100 --> 00:19:17,630 hardware that's It's a challenging. 295 00:19:18,030 --> 00:19:21,670 That's not in our expertise like 296 00:19:21,670 --> 00:19:25,390 a bandwidth in open quantum design. And we are trying to 297 00:19:25,390 --> 00:19:28,670 get help from like some expert in the rf. 298 00:19:28,990 --> 00:19:32,750 There is nothing available I could say 299 00:19:33,230 --> 00:19:36,990 reliable for special 300 00:19:36,990 --> 00:19:39,390 for ion trap that you can use it 301 00:19:41,010 --> 00:19:44,770 easily and control your. Your ions or your atoms. 302 00:19:45,170 --> 00:19:48,930 And even the. In 303 00:19:48,930 --> 00:19:52,610 the. In the closed system also which is very expensive. 304 00:19:52,850 --> 00:19:56,290 It's not fully functional to control the ions yet. 305 00:19:56,930 --> 00:20:00,210 And yeah we have some gap there that we need to 306 00:20:00,930 --> 00:20:04,210 fill out. Controlling the phase reliably 307 00:20:04,610 --> 00:20:08,330 and in. In an open source community which is as I said is 308 00:20:08,330 --> 00:20:11,730 based on the Arctic and Cyanura 309 00:20:11,730 --> 00:20:15,250 blocks. It's a company for ion trap. The same like a 310 00:20:15,570 --> 00:20:19,010 company like other names coming to the superconducting 311 00:20:19,010 --> 00:20:22,610 qubits. Also they're a little bit ahead than 312 00:20:22,610 --> 00:20:26,290 ion traps if you want to compare it in that area. 313 00:20:26,930 --> 00:20:30,210 Okay. Ion traps you don't really hear. 314 00:20:30,850 --> 00:20:34,470 I haven't heard a lot about that lately. What exactly is is an 315 00:20:34,470 --> 00:20:35,950 ion trap versus. 316 00:20:37,710 --> 00:20:41,470 Superconducting for superconducting? Like what. How does that relate. I 317 00:20:41,470 --> 00:20:44,750 know it's one of the hardware kind of families. Yes. Like 318 00:20:45,470 --> 00:20:49,270 for any quantum computation you need some 319 00:20:49,270 --> 00:20:52,670 species. Some people using photon 320 00:20:53,470 --> 00:20:57,270 which is like a photonic based quantum computing like a 321 00:20:57,270 --> 00:20:59,870 Xanadu or Psi quantum. They are using photons. 322 00:21:01,390 --> 00:21:04,430 Some people using a atom 323 00:21:05,390 --> 00:21:08,910 directly single atoms. That's like a neutral atoms 324 00:21:10,030 --> 00:21:12,830 like a Q era for example. Using this kind of 325 00:21:13,390 --> 00:21:15,710 species to encode the information. 326 00:21:17,150 --> 00:21:20,190 And another type of 327 00:21:20,670 --> 00:21:24,430 quantum computer that using species of ion 328 00:21:24,590 --> 00:21:28,090 means like you have atoms and then you 329 00:21:28,090 --> 00:21:31,090 shoot out one of the electrons ionized. 330 00:21:31,890 --> 00:21:35,450 Why you make it ionized? Because it will 331 00:21:35,450 --> 00:21:38,370 be easier and deeper to trap it. 332 00:21:39,090 --> 00:21:42,930 What the trap means like you need to have like 333 00:21:43,090 --> 00:21:46,810 atoms and these atoms, single 334 00:21:46,810 --> 00:21:50,450 atoms, no interaction with the environment and then 335 00:21:50,450 --> 00:21:54,270 levitate in whatever area that you have. We have a 336 00:21:54,270 --> 00:21:57,790 vacuum system. We push atoms inside like 337 00:21:57,790 --> 00:22:01,310 evaporate atoms to the vacuum system and 338 00:22:01,310 --> 00:22:05,030 then shoot the laser to the atom. A bunch of Atom 10 339 00:22:05,030 --> 00:22:08,790 to 2010 to power 20 like and then ionize 340 00:22:09,030 --> 00:22:12,630 atoms send electron out will be ion. And the 341 00:22:12,630 --> 00:22:16,230 ion, because it has a charge, you 342 00:22:16,230 --> 00:22:20,040 can trap it with the DC and RF voltage. And 343 00:22:20,040 --> 00:22:23,560 then that trap is very deep. Compared to the neutral 344 00:22:23,560 --> 00:22:27,400 atoms. Neutral atoms is shallower. Like if 345 00:22:27,400 --> 00:22:31,200 you some other hydrogen atom inside your vacuum 346 00:22:31,200 --> 00:22:34,960 system come and hit the atom can kick 347 00:22:34,960 --> 00:22:38,680 it out. But in ion trap is deeper. Like you 348 00:22:38,680 --> 00:22:42,320 need more energy to kick it out than if you trap 349 00:22:42,480 --> 00:22:46,320 like one ion in your vacuum system. You can have it 350 00:22:46,890 --> 00:22:50,570 quite long for a day or two and then do the experiment with it. 351 00:22:50,730 --> 00:22:54,290 Compared to neutral atom which is like quickly will 352 00:22:54,290 --> 00:22:58,050 escape and you need to trap it again and again. But 353 00:22:58,050 --> 00:23:01,810 in terms of the physics they are quite similar. Like the way that we 354 00:23:01,810 --> 00:23:05,530 trap is different. But in terms of physics you can find a lot of 355 00:23:05,530 --> 00:23:09,170 similarity. Yeah. Compared to the 356 00:23:09,170 --> 00:23:12,570 superconducting cube. Superconducting qubits are circuits. 357 00:23:13,010 --> 00:23:16,210 Like in principles are electronic circuits 358 00:23:16,850 --> 00:23:20,530 in the quantum level. Like they have some non linear component. 359 00:23:20,530 --> 00:23:23,330 Like in the normal electric circuit we have 360 00:23:23,330 --> 00:23:26,850 rcl. But they have some component which is 361 00:23:27,330 --> 00:23:31,090 non linear. And they can work as the source of 362 00:23:31,090 --> 00:23:34,890 making the energy differences and make a qubit. They 363 00:23:34,890 --> 00:23:38,530 are handmade compared to the atoms which is 364 00:23:39,120 --> 00:23:42,760 natural. Like it's very 365 00:23:42,760 --> 00:23:45,880 difficult to make 10 quantum bit in 366 00:23:45,880 --> 00:23:49,520 superconducting qubits similar. Exactly. Because 367 00:23:49,520 --> 00:23:53,280 it's handling. But in in atomic word they 368 00:23:53,280 --> 00:23:57,120 are similar in nature. In. In that sense it's easier 369 00:23:57,200 --> 00:24:00,880 to to start with. But the 370 00:24:00,880 --> 00:24:04,080 difficulty will come afterwards when the engineering comes. Like 371 00:24:05,120 --> 00:24:08,400 the electronics supports a lot the 372 00:24:08,400 --> 00:24:12,240 superconducting cubits back in AD 19. Then 373 00:24:12,400 --> 00:24:16,080 there's a huge engineer development there which is 374 00:24:16,080 --> 00:24:19,680 missing in atoms and ion community. 375 00:24:20,559 --> 00:24:23,760 If we could fill that gap, they could 376 00:24:23,920 --> 00:24:27,280 outperform the superconducting qubit because 377 00:24:27,360 --> 00:24:30,000 naturally their properties are better. 378 00:24:31,310 --> 00:24:33,790 That's the differences between these two. Like a 379 00:24:34,990 --> 00:24:38,510 way of or three photons, superconducting 380 00:24:38,510 --> 00:24:41,710 qubits and atom or ion base. 381 00:24:44,030 --> 00:24:47,630 Okay. So there's a lot of precision that goes behind 382 00:24:47,790 --> 00:24:51,310 the ion traps. It's very exciting. You get lasers, 383 00:24:51,390 --> 00:24:54,750 vacuum chambers, electromagnetic fields, 384 00:24:55,150 --> 00:24:58,870 all these mind blowing ideas all working 385 00:24:58,870 --> 00:25:01,870 in Harmony. What's the most challenging part 386 00:25:02,750 --> 00:25:05,950 of building or maintaining an ion trap system 387 00:25:07,630 --> 00:25:11,470 at the moment? Scalability. That's the, 388 00:25:11,710 --> 00:25:15,510 that's the, that's the challenging part. The 389 00:25:15,510 --> 00:25:19,230 rest is already is manageable. Like you can have 390 00:25:20,030 --> 00:25:23,550 recently Quantinium published like a 98 qubit. 391 00:25:24,310 --> 00:25:27,750 And with the benchmarking, when I say 392 00:25:27,990 --> 00:25:31,670 there is some spec of your machine that will define 393 00:25:32,310 --> 00:25:36,150 how it's working in terms 394 00:25:36,150 --> 00:25:39,670 of the element level and in terms of system level. 395 00:25:41,990 --> 00:25:45,670 There's a huge people that working on this benchmarking 396 00:25:45,670 --> 00:25:48,470 different machine. And 397 00:25:51,100 --> 00:25:54,580 what was, what was the question? I forgot I was saying what was the most 398 00:25:54,580 --> 00:25:58,380 difficult. The difficult like the scalability. Then you want to increase 399 00:25:58,380 --> 00:26:02,140 this like a 98 qubit to 500 or thousand 400 00:26:02,140 --> 00:26:05,500 or 10,000 or 30,000. How you want to do that? 401 00:26:05,820 --> 00:26:08,700 That's challenging. Like different 402 00:26:08,940 --> 00:26:12,620 architectures already like is 403 00:26:12,620 --> 00:26:16,150 under investigation to see how we can do that. 404 00:26:18,470 --> 00:26:22,150 Yeah, two main way to do for the ion trap 405 00:26:22,550 --> 00:26:25,990 is either do it like a 406 00:26:26,150 --> 00:26:29,510 node base, like have 100 qubit here, 407 00:26:29,750 --> 00:26:33,070 100 qubit somewhere else like within like 408 00:26:33,070 --> 00:26:36,710 acceptable range and then connect them with photons. Then that 409 00:26:36,710 --> 00:26:39,750 way you can extend and escape. 410 00:26:41,200 --> 00:26:45,000 That's one approach that IonQ and IX or 411 00:26:45,000 --> 00:26:48,840 Linux recently starting that kind of approach. And 412 00:26:48,840 --> 00:26:51,840 also another approach is it's called qccd. 413 00:26:52,720 --> 00:26:56,480 You have core center and then you transfer 414 00:26:56,960 --> 00:27:00,600 physically or ion to somewhere else and do something and bring it 415 00:27:00,600 --> 00:27:04,440 back like transferring between different nodes 416 00:27:04,440 --> 00:27:08,080 that you have. That's also another approach that 417 00:27:09,040 --> 00:27:11,920 some company is using that approach to 418 00:27:12,940 --> 00:27:16,740 reach to scalability. But still it's very difficult challenge 419 00:27:16,740 --> 00:27:20,060 to overcome. And like a company put 420 00:27:20,460 --> 00:27:24,060 like a 32,000, 2030 or 421 00:27:24,060 --> 00:27:26,940 2035 to reach to some level of 30K 422 00:27:27,900 --> 00:27:31,700 level of scalability like number 423 00:27:31,700 --> 00:27:35,300 of the qubit that you have plus the rate of error that you 424 00:27:35,300 --> 00:27:38,780 have each single qubit. So do you think the 425 00:27:38,780 --> 00:27:41,500 trapped ion systems will eventually power 426 00:27:42,220 --> 00:27:45,960 commercial quantum computers or are they more likely to 427 00:27:45,960 --> 00:27:49,160 remain like the gold standard for scientific 428 00:27:49,240 --> 00:27:52,760 precision and benchmarking? Yeah, it's a 429 00:27:52,760 --> 00:27:55,960 difficult question, but the paper publication that 430 00:27:56,360 --> 00:27:59,960 you could already see in 431 00:27:59,960 --> 00:28:03,640 terms of system benchmarking. 432 00:28:03,880 --> 00:28:07,600 Like system benchmarking means that you have some specific 433 00:28:07,600 --> 00:28:11,290 algorithm and you give it to me, I will run it on 434 00:28:11,290 --> 00:28:15,130 ion trap machine. You give it to the second ion trap 435 00:28:15,130 --> 00:28:18,690 machine, you give it to the neutral atoms machine, you give it to the 436 00:28:18,690 --> 00:28:21,570 photonics machine, superconducting machine and compare the result 437 00:28:22,610 --> 00:28:26,370 in terms of accuracy and the time that it takes 438 00:28:26,370 --> 00:28:28,770 to get back the data. 439 00:28:31,410 --> 00:28:34,850 Like iron trap, it's top now like with the 440 00:28:35,090 --> 00:28:38,610 error rate and Even with the lower number of the 441 00:28:38,610 --> 00:28:42,250 qubit compared to, for example, superconducting. But 442 00:28:42,890 --> 00:28:46,610 in terms of the overall, like a benchmarking is still. Yeah, it's 443 00:28:46,610 --> 00:28:48,890 better than superconducting. 444 00:28:50,090 --> 00:28:53,570 But this question. Yeah, it's very difficult to 445 00:28:53,570 --> 00:28:57,050 answer. There's like some other competitor, like a photonic base. 446 00:28:57,290 --> 00:29:01,050 They're claiming for a million qubit, but it's not out there 447 00:29:01,050 --> 00:29:04,660 yet. Right, but it could come, like, who knows? 448 00:29:05,700 --> 00:29:09,540 Yeah, so. So are sheer number of 449 00:29:09,540 --> 00:29:13,300 qubits going to matter or is it logical qubits and there's physical qubits. 450 00:29:13,300 --> 00:29:16,900 Like is. Is there? I think what I really want to know is like, what 451 00:29:17,539 --> 00:29:21,140 trapped ions, you know, the dealing is good for one type of problem 452 00:29:21,620 --> 00:29:25,220 photonics are good for. Where does trapped ion really, like, 453 00:29:25,220 --> 00:29:29,070 shine? Like, the original question that 454 00:29:29,070 --> 00:29:32,750 you mentioned was about the logical and physical qubit. Well, yeah, I know. 455 00:29:32,750 --> 00:29:36,070 I. Sorry, I had way too much coffee, so I dumped a couple of questions 456 00:29:36,070 --> 00:29:38,150 on you. Sorry about that. 457 00:29:40,150 --> 00:29:43,190 But the first question I want to ask, 458 00:29:44,550 --> 00:29:46,630 let's hit the undo button on that. 459 00:29:48,710 --> 00:29:52,510 Trapped ions, where. What problems are 460 00:29:52,510 --> 00:29:56,250 they really perfect for? Is really the question I want to know. Like, 461 00:29:56,250 --> 00:29:59,690 where. You know, if I'm looking at a whole suite of problems, 462 00:30:00,010 --> 00:30:03,690 kneeling is good for one type of thing, photonics. And where does. 463 00:30:03,690 --> 00:30:07,210 Where does trapped ion really excel? Because, yeah, I think the target 464 00:30:07,290 --> 00:30:11,090 for trap ion is the universal quantum computation. It's not just 465 00:30:11,090 --> 00:30:14,410 any gotcha. Okay. It's like, yeah, in any 466 00:30:14,410 --> 00:30:18,010 algorithm that you can give it and get the answer, it's a matter 467 00:30:18,090 --> 00:30:21,710 of like quantum volume, like a number of a qubit, error 468 00:30:21,710 --> 00:30:24,910 rate, those kind of things. But the target is to solve 469 00:30:25,870 --> 00:30:29,510 like a hard problem that the classical computer cannot. That. That's the 470 00:30:29,510 --> 00:30:30,910 target. Gotcha. 471 00:30:33,230 --> 00:30:36,630 Yeah. Because I can easily see like, you know, kind of. I'm old enough to 472 00:30:36,630 --> 00:30:40,270 remember the, the early, like WINTEL days where like this is 100 473 00:30:40,270 --> 00:30:43,950 megahertz, this is 150 megahertz, this is 166 megahertz. 474 00:30:44,110 --> 00:30:47,790 Right. You know, like that, like that became like a marketing scheme. 475 00:30:47,950 --> 00:30:51,790 Like, and I know that there was a hard. There was a speed boost attached 476 00:30:51,790 --> 00:30:55,390 to that, but yeah, beyond a certain point, it was not really a 477 00:30:55,390 --> 00:30:58,670 meaningful measure of how fast the machine is. 478 00:30:59,390 --> 00:31:03,070 And it seems to me that quantum, like number of qubits and all the. 479 00:31:03,070 --> 00:31:06,430 That's even more complicated. And number of qubits does not 480 00:31:06,430 --> 00:31:10,150 necessarily mean number of logical qubits. 481 00:31:10,150 --> 00:31:13,750 So it seems like how you know, at some 482 00:31:13,750 --> 00:31:17,570 point when this becomes real, real, real and not that it's 483 00:31:17,570 --> 00:31:21,330 not real. Today. But when it becomes something that you know you'll see ads for, 484 00:31:24,610 --> 00:31:28,130 how are they going to be measured? Like how do you compare one quantum computer 485 00:31:28,130 --> 00:31:31,649 to another? Like you know, that was, that was the question. Actually Kendence 486 00:31:31,649 --> 00:31:35,410 mentioned that benchmarking has a several different 487 00:31:35,650 --> 00:31:39,490 like layer, right? Benchmark different like 488 00:31:40,130 --> 00:31:43,770 component level and say okay, I do one 489 00:31:43,770 --> 00:31:47,210 gate, two gate and the spam detection with this much 490 00:31:47,210 --> 00:31:50,810 error. That's very good. But it's not the whole 491 00:31:50,810 --> 00:31:54,650 story. The whole story is that you give me a like 492 00:31:54,650 --> 00:31:58,010 a sample algorithm, that it's a standard algorithm and then 493 00:31:58,250 --> 00:32:01,930 you can run it on classical machine and see, okay, takes 10 to 494 00:32:02,010 --> 00:32:05,730 25 years to be solved and then run it on like a quantum 495 00:32:05,730 --> 00:32:09,060 computer and then see how long it takes 496 00:32:09,220 --> 00:32:13,060 and how accurate the result is and then compared 497 00:32:13,060 --> 00:32:16,740 with the other machine. Like this type of benchmarking is already 498 00:32:16,820 --> 00:32:19,460 is ongoing. Like a lot of 499 00:32:20,820 --> 00:32:24,180 quantum machine that is out there. They are 500 00:32:24,260 --> 00:32:27,700 publishing based on that kind of like they're trying to 501 00:32:27,700 --> 00:32:31,500 publish several algorithms. 1, 2, 3 and the paper and saying okay, 502 00:32:31,500 --> 00:32:35,310 we did this on this machine and then this is the result. 503 00:32:35,790 --> 00:32:39,590 And then even they compare it with some other machine back to back then 504 00:32:39,590 --> 00:32:43,110 show the result how like how much 505 00:32:43,110 --> 00:32:46,630 error with how much uncertainty you can give this 506 00:32:46,630 --> 00:32:50,270 answer those kind of measure is a more 507 00:32:50,270 --> 00:32:53,870 system level benchmarking. That's more important 508 00:32:54,030 --> 00:32:57,670 at the end of story. I gotcha. All right, that makes 509 00:32:57,670 --> 00:33:00,950 sense. So for students or engineers who are 510 00:33:00,950 --> 00:33:03,070 fascinated by trapped ions, 511 00:33:04,590 --> 00:33:07,830 what skills or areas of study would you 512 00:33:07,830 --> 00:33:10,590 recommend they start exploring now? 513 00:33:12,910 --> 00:33:16,350 Definitely start with physics background and AML physics, 514 00:33:16,350 --> 00:33:19,830 atomic and molecular physics. That's the 515 00:33:19,830 --> 00:33:23,110 core for 516 00:33:23,110 --> 00:33:26,470 understanding the single ingredient 517 00:33:26,470 --> 00:33:30,220 inside. But at the moment we need a lot of 518 00:33:30,220 --> 00:33:33,900 other skills need to be developed also like optical 519 00:33:34,140 --> 00:33:36,620 engineer. Like we need to 520 00:33:38,060 --> 00:33:41,660 take the data from atoms or ions to 521 00:33:42,460 --> 00:33:46,060 our detector and then we need some collection 522 00:33:46,060 --> 00:33:49,100 system and imaging system for example. That's very important. 523 00:33:50,540 --> 00:33:54,220 Understanding laser, how we can use the laser, 524 00:33:54,990 --> 00:33:57,950 how like we can control the different spec of the light. 525 00:33:58,430 --> 00:34:01,750 That's very important. And either 526 00:34:01,750 --> 00:34:05,470 mechanical engineer comes with the game because you need 527 00:34:05,470 --> 00:34:09,150 to make like here at oqd 528 00:34:10,270 --> 00:34:13,990 every time we hire several co op students that we have 529 00:34:13,990 --> 00:34:17,830 in here in Canada that the students in the bachelor level in 530 00:34:17,830 --> 00:34:21,670 mechanical engineer should come to some company and learn for 531 00:34:21,670 --> 00:34:25,339 something for for like a four months and they have to be 532 00:34:25,339 --> 00:34:28,859 paid also they come to this machine and then they helps us 533 00:34:28,859 --> 00:34:32,499 to build some optomechanical modules 534 00:34:33,219 --> 00:34:36,899 that we need to make it with specific spec 535 00:34:37,779 --> 00:34:41,059 to build our whole machine. That's also 536 00:34:41,699 --> 00:34:45,539 important. It's not to the core of the like iron 537 00:34:45,539 --> 00:34:48,899 trap, but it's very super critical 538 00:34:50,360 --> 00:34:53,640 to, to, to show the criticality of this 539 00:34:54,440 --> 00:34:58,200 like making mechanical stable system. Like we 540 00:34:58,200 --> 00:35:01,560 have 30 ions sitting 4 micrometer 541 00:35:01,880 --> 00:35:05,640 beside each other, like 120 micron. And 542 00:35:05,640 --> 00:35:09,320 then this is the, like a line of the ions that we start 543 00:35:09,320 --> 00:35:12,600 working and, and doing quantum computation with them. 544 00:35:13,800 --> 00:35:17,650 Like the thickness of the hair, human hair, is 100 micron. 545 00:35:18,050 --> 00:35:21,890 Like this, like a string is sitting at the cut of the human 546 00:35:21,890 --> 00:35:24,690 hair. And then you need to control each 547 00:35:24,930 --> 00:35:28,290 individual atoms with lasers. 548 00:35:28,850 --> 00:35:32,210 And then the laser that you're talking with atom number one 549 00:35:32,370 --> 00:35:35,930 shouldn't talk to, with atom number two. You shouldn't have crosstalk 550 00:35:35,930 --> 00:35:39,530 like four micrometer away and 551 00:35:39,530 --> 00:35:43,210 four micrometer away. And then you can imagine 552 00:35:43,210 --> 00:35:46,650 how stable your mechanical system needs to be 553 00:35:47,130 --> 00:35:50,330 to not make these two mixing each other. 554 00:35:52,250 --> 00:35:55,450 And it become very important indirectly to 555 00:35:56,410 --> 00:35:59,450 the main problem. But for people who are interested 556 00:36:00,250 --> 00:36:03,930 as a builder or as a user, two different categories. 557 00:36:04,570 --> 00:36:07,850 Builders are the people who should know about the 558 00:36:08,360 --> 00:36:11,560 atomic physics, this laser, optical engineering, 559 00:36:11,560 --> 00:36:15,000 mechanical electrical users. 560 00:36:15,160 --> 00:36:18,840 They should know more about the software because like whenever that 561 00:36:18,840 --> 00:36:22,680 you use your classical computer, you don't ask about like how the CPU 562 00:36:22,680 --> 00:36:26,200 is built. The software level, 563 00:36:26,280 --> 00:36:29,960 like more theoretical physics will be the user. Like 564 00:36:31,320 --> 00:36:35,160 the comparison will be builder of telescope and user of the 565 00:36:35,160 --> 00:36:38,950 telescope. Like there's like a two different category and we are in the builder 566 00:36:38,950 --> 00:36:42,110 side of open quantum design. We are building the, 567 00:36:42,510 --> 00:36:45,870 the machine. And then some people coming from 568 00:36:45,870 --> 00:36:49,070 theoretical physics can come and use it afterwards. 569 00:36:51,230 --> 00:36:52,110 That makes sense. 570 00:36:55,310 --> 00:36:56,190 That makes sense. 571 00:36:58,830 --> 00:37:02,520 What's the most surprising or beautiful thing 572 00:37:02,520 --> 00:37:06,360 you've ever seen happen when working in a quantum 573 00:37:06,360 --> 00:37:07,000 experiment? 574 00:37:09,720 --> 00:37:13,080 Yeah, like you see some unexpected effect 575 00:37:13,720 --> 00:37:17,199 and you see that kind of noise and then after some 576 00:37:17,199 --> 00:37:20,920 investigation you see that, oh, that's another phenomena that it's 577 00:37:20,920 --> 00:37:24,520 kind of coming to your game and showing and it, 578 00:37:24,680 --> 00:37:28,000 yeah. Then go to the theory and see, oh, that's, that's kind of interesting things 579 00:37:28,000 --> 00:37:31,410 to develop. Even sometimes you will be 580 00:37:31,410 --> 00:37:34,930 sidetracked to that kind of problem and see some 581 00:37:34,930 --> 00:37:38,770 achievement happen at that kind of noise that you saw in the system. 582 00:37:39,410 --> 00:37:42,130 I've seen several of these examples on my 583 00:37:43,090 --> 00:37:46,130 own research. It's kind of very super 584 00:37:46,530 --> 00:37:50,370 interesting and exciting. Yeah, I also wonder too, 585 00:37:50,370 --> 00:37:54,090 like, you know, how much, how much of it is you get this, like, is 586 00:37:54,090 --> 00:37:57,810 that, wait, is my equipment like messed up or is my seeing a new 587 00:37:57,970 --> 00:38:01,390 undocumented phenomena? Right. Like, you know. Yeah, yeah, yeah. There's probably a bit of 588 00:38:01,390 --> 00:38:04,910 excitement, a little bit of skepticism and a little bit of like you know, 589 00:38:05,070 --> 00:38:08,790 not sure which it is. Yeah, that's very important in 590 00:38:08,790 --> 00:38:12,430 the AMO physics lab. One of the differences that I would 591 00:38:12,590 --> 00:38:16,190 I should highlight here is that in. If 592 00:38:16,190 --> 00:38:19,910 have you ever been in like a experimental lab, like AMO 593 00:38:19,910 --> 00:38:23,510 lab? I have not, no. No, it's. It's kind 594 00:38:23,510 --> 00:38:27,070 of very super messy. Everything like the 595 00:38:27,070 --> 00:38:30,670 cables coming, lasers and the mechanics. And then 596 00:38:31,260 --> 00:38:34,820 in principle that system works just once and then paper 597 00:38:34,820 --> 00:38:38,380 published and then second time you don't know that it's working or not. You need 598 00:38:38,380 --> 00:38:41,980 to spend a little time to like in principle students 599 00:38:42,220 --> 00:38:45,940 spend like a 90% of the time to fix the problems 600 00:38:45,940 --> 00:38:49,300 ongoing and then probably 10% doing a real 601 00:38:49,300 --> 00:38:52,580 experiment on a normal like 602 00:38:52,580 --> 00:38:56,300 AMLABS. One of the goal of the OQD 603 00:38:56,300 --> 00:39:00,040 is that change this ratio to lower value. 604 00:39:00,520 --> 00:39:04,200 Try to make the system more stable and more modular 605 00:39:04,600 --> 00:39:08,200 in the way that we can monitor different sections 606 00:39:08,360 --> 00:39:11,880 and also make it stable 607 00:39:11,880 --> 00:39:15,640 overall working for a long time. More 608 00:39:15,640 --> 00:39:19,400 work as a commercial product compared to the R and D product. 609 00:39:20,600 --> 00:39:24,030 The comparison will be like a breadboard in electronics and then the 610 00:39:24,030 --> 00:39:27,310 PCB version, like breadboard is the time that you do 611 00:39:27,310 --> 00:39:30,550 testing and then PCB is like a solid. 612 00:39:30,950 --> 00:39:34,590 And yeah, in OQD we try to go from 613 00:39:34,590 --> 00:39:38,270 like a breadboard version, messy version to more solid and 614 00:39:38,270 --> 00:39:41,990 PCB type version and try to change its 615 00:39:41,990 --> 00:39:45,830 balance. Like spend less time on fixing problems, spend more time 616 00:39:45,830 --> 00:39:49,680 on doing the experiment. Actually that's 617 00:39:49,680 --> 00:39:53,200 one of the main goal here. And we could see like we have another lab 618 00:39:53,200 --> 00:39:56,920 design, the OQD that is R and D based in 619 00:39:56,920 --> 00:40:00,400 the grandpa of our lab. And we could see 620 00:40:00,720 --> 00:40:04,440 the thing that they spend like a year to achieve. We 621 00:40:04,440 --> 00:40:06,000 could do it in like two weeks. 622 00:40:07,840 --> 00:40:11,680 That's very, very distinct value that 623 00:40:11,680 --> 00:40:15,440 you could see. We have a lot of camera photodiode in our system, probably plus 624 00:40:15,760 --> 00:40:19,600 hundred that monitor different level of the system 625 00:40:19,840 --> 00:40:23,440 and report the errors. And that will 626 00:40:23,440 --> 00:40:27,240 help you to maintain the system commercially. 627 00:40:27,240 --> 00:40:30,080 Not. Not as a physicist, it's a more engineer. 628 00:40:30,880 --> 00:40:34,680 Okay, interesting. So there's a 629 00:40:34,680 --> 00:40:38,480 lot of kind of interdisciplinary collaboration that's 630 00:40:38,480 --> 00:40:42,120 happening exactly in quantum development. Right. You've got your physicists, 631 00:40:42,120 --> 00:40:45,880 you've got your engineers, computer scientists. So how 632 00:40:45,880 --> 00:40:49,290 do you, how do you find the shared 633 00:40:49,290 --> 00:40:52,810 language so that you're able to kind of bridge those 634 00:40:52,810 --> 00:40:56,570 disciplines effectively? Yeah, Someone 635 00:40:56,650 --> 00:41:00,010 who has like a little bit of each one of those needs 636 00:41:00,250 --> 00:41:03,530 should be on top of the project. Like should lead everything. 637 00:41:03,770 --> 00:41:06,810 Like a person who knows a little bit software, mechanical 638 00:41:07,450 --> 00:41:09,610 optics should be on top of that too. 639 00:41:10,890 --> 00:41:14,330 Whenever that you hire someone from a Specific vision 640 00:41:14,810 --> 00:41:18,570 can get benefit of his experts knowledge out. 641 00:41:20,010 --> 00:41:23,610 Yeah, that's very crucial to have someone who has 642 00:41:24,650 --> 00:41:27,930 done a little bit on some of those kind of activities. 643 00:41:29,210 --> 00:41:29,850 Okay. 644 00:41:32,810 --> 00:41:36,370 Yeah. For people who probably trust that you can later 645 00:41:36,370 --> 00:41:39,450 put like open Quantum design Link and also GitHub 646 00:41:39,850 --> 00:41:43,600 to the people who are interested to see what kind 647 00:41:43,600 --> 00:41:47,440 of things already it's on public level that everyone can 648 00:41:47,840 --> 00:41:51,440 see on hardware and software. We have also some 649 00:41:51,600 --> 00:41:55,320 simulation level which is very important also from aim of physics 650 00:41:55,320 --> 00:41:58,880 simulation that anyone who want to run 651 00:41:58,880 --> 00:42:02,440 anything can use that from our GitHub and this 652 00:42:02,440 --> 00:42:05,840 GitHub will be more and more published in future 653 00:42:05,840 --> 00:42:09,520 whenever that we get some spec of our prototype 654 00:42:10,710 --> 00:42:14,070 and then they could have more information of each 655 00:42:14,070 --> 00:42:15,990 individual module that we have here. 656 00:42:17,910 --> 00:42:21,590 Interesting. Yeah, I think there's a lot, there's a lot to say. 657 00:42:21,590 --> 00:42:25,030 Like you know, it seems like quantum computing 658 00:42:25,510 --> 00:42:27,910 is going to need a lot of multidisciplinary and 659 00:42:29,750 --> 00:42:33,390 people. So I think would that be good advice for people like if you're really 660 00:42:33,390 --> 00:42:37,100 good at one thing, learn a little bit of something else. Yeah, yeah. And 661 00:42:37,100 --> 00:42:40,060 actually I think from the report in the 662 00:42:41,100 --> 00:42:44,900 North America I've heard some, I've seen some like a 663 00:42:44,900 --> 00:42:48,300 publication that there is shortage on the workforce 664 00:42:49,100 --> 00:42:50,540 for quantum development. 665 00:42:52,860 --> 00:42:56,700 That's interesting. And that coming, that's coming from 666 00:42:56,700 --> 00:42:59,820 different kind of principles. As I mentioned 667 00:43:01,020 --> 00:43:04,840 it could be completely not relevant to 668 00:43:04,840 --> 00:43:08,320 the. To the quantum word, but it's directly 669 00:43:08,560 --> 00:43:12,080 relevant to building the quantum computer. Right. 670 00:43:14,400 --> 00:43:17,440 So how close are we really to seeing 671 00:43:18,080 --> 00:43:21,760 quantum technologies like quantum memory 672 00:43:21,920 --> 00:43:25,760 or quantum Internet change our everyday 673 00:43:25,760 --> 00:43:29,450 lives? Yeah, 674 00:43:29,450 --> 00:43:33,130 I think we like a quantum related phenomena 675 00:43:33,130 --> 00:43:36,890 is already affect our life. Like whenever that 676 00:43:37,130 --> 00:43:40,410 you are using gps, you're using atomic clock. 677 00:43:40,890 --> 00:43:44,690 That's definitely direct use of 678 00:43:44,690 --> 00:43:48,450 the quantum word in daily application. If it 679 00:43:48,450 --> 00:43:51,210 comes to the quantum memory also 680 00:43:52,170 --> 00:43:55,840 my PhD was on the quantum memory based. There are already some 681 00:43:55,840 --> 00:43:59,080 companies that building quantum memory and quantum 682 00:43:59,080 --> 00:44:02,760 repeaters across the world testing across 683 00:44:03,000 --> 00:44:06,440 like a hundred fiber, 100 kilometer fiber and. 684 00:44:06,520 --> 00:44:09,640 And that will be crucial also for future 685 00:44:11,960 --> 00:44:15,520 quantum like a cryptography or for quantum 686 00:44:15,520 --> 00:44:19,280 computing even like some, some company already doing 687 00:44:19,280 --> 00:44:22,890 that. And for quantum computing it's a little 688 00:44:22,890 --> 00:44:25,970 bit further out. Like 689 00:44:26,290 --> 00:44:29,810 quantum memory, quantum communication, quantum sensor is closer 690 00:44:29,890 --> 00:44:33,730 and you have some product already. But quantum computing it's a little 691 00:44:33,730 --> 00:44:37,570 bit back and still we're not there to say 692 00:44:37,890 --> 00:44:41,730 this is directly used on our daily lives. But 693 00:44:41,890 --> 00:44:45,650 some problems like finding the proper drug 694 00:44:45,650 --> 00:44:49,390 for example, that's one of the things that 695 00:44:49,390 --> 00:44:52,470 the quantum computer can affect. 696 00:44:53,030 --> 00:44:56,790 If you have a quantum computer easily can break 697 00:44:56,790 --> 00:45:00,470 your cryptography that you have for 698 00:45:00,470 --> 00:45:03,910 bank account these days. That's a thread to 699 00:45:04,550 --> 00:45:08,390 a daily life if it comes to the reality. 700 00:45:09,270 --> 00:45:11,670 And that's already has a lot of people on it. 701 00:45:13,220 --> 00:45:16,940 Yeah, yeah. That also could be something that 702 00:45:16,940 --> 00:45:18,180 yeah. Will affect. 703 00:45:21,140 --> 00:45:24,020 Yeah. But yeah, it's a very tough question to say 704 00:45:24,420 --> 00:45:28,100 exactly what, what will be there? Who knows? 705 00:45:28,820 --> 00:45:32,660 That's fair. That's fair. I always like to, I ask this question in every 706 00:45:32,660 --> 00:45:36,420 single interview and I always love the answers. 707 00:45:37,300 --> 00:45:39,940 How would you explain what you do 708 00:45:41,580 --> 00:45:45,300 on a daily basis to a non technical person who's not in 709 00:45:45,300 --> 00:45:49,060 your field, like just completely non technical. How can you explain 710 00:45:49,060 --> 00:45:49,980 it and break it down? 711 00:45:54,620 --> 00:45:58,380 You mean what in terms of what. In terms of like 712 00:45:58,700 --> 00:46:02,460 in terms of what, what you're doing in quantum computing. In terms of what 713 00:46:02,460 --> 00:46:06,020 you're, what problems you're trying to solve. Is this the 714 00:46:06,020 --> 00:46:09,490 cocktail question? Kind of, yes. Things like, you know, 715 00:46:10,210 --> 00:46:13,170 like if you're at a cocktail party or whatever and what do you do? Well. 716 00:46:15,250 --> 00:46:18,450 How do you break it down? How do you unpack it a little bit. 717 00:46:21,730 --> 00:46:25,370 For like explain. Like I didn't fully get probably the 718 00:46:25,370 --> 00:46:28,970 question that explaining the, the problem that I'm working 719 00:46:28,970 --> 00:46:32,370 with in or in general, like a quantum computation. 720 00:46:33,170 --> 00:46:36,800 Which one? I would say in general. Yeah, you get the question 721 00:46:36,800 --> 00:46:40,520 what do you do? Right. And you have to assume they're a civilian, 722 00:46:40,520 --> 00:46:43,600 right? Like they're not. Yeah, yeah, 723 00:46:44,720 --> 00:46:48,560 yeah. I could go like, I could say like for example, if 724 00:46:48,560 --> 00:46:51,440 you go to the, to the beach and get one of those 725 00:46:52,560 --> 00:46:56,360 tiny sand on the beach that 726 00:46:56,360 --> 00:46:59,360 has 10 to the power of 20 atoms inside 727 00:47:01,650 --> 00:47:05,170 one of those tiny sands and here 728 00:47:05,570 --> 00:47:08,850 we are working with one single atoms. 729 00:47:10,450 --> 00:47:10,930 Wow. 730 00:47:15,730 --> 00:47:19,370 That probably can bring you 731 00:47:19,370 --> 00:47:22,690 in the scale size. How 732 00:47:22,850 --> 00:47:25,570 difficult is to work with this guy 733 00:47:28,650 --> 00:47:32,370 in terms of the scale of your system control? For 734 00:47:32,370 --> 00:47:35,810 example, if you consider Toronto to London, for 735 00:47:35,810 --> 00:47:39,530 example, you want to report the distance 736 00:47:39,530 --> 00:47:43,290 between these two cities by nanometer scale. 737 00:47:45,850 --> 00:47:49,610 That's also in terms of scaling factor 738 00:47:49,610 --> 00:47:53,450 for controlling your laser light, whatever that you do that you want to control this 739 00:47:53,900 --> 00:47:57,580 single atoms, you need to have that kind of control. And we have 740 00:47:57,820 --> 00:48:01,420 that kind of control that we can report the distance between 741 00:48:01,980 --> 00:48:05,580 New York, London with nanometer scale. 742 00:48:06,540 --> 00:48:10,260 Then you can go to your cocktail party and think about this kind of 743 00:48:10,260 --> 00:48:10,540 things. 744 00:48:14,380 --> 00:48:17,900 We have a lot of folks who are, you know, who are 745 00:48:18,220 --> 00:48:21,500 currently, they're, they're physicists, they're really in academia 746 00:48:22,700 --> 00:48:26,160 and they want to grow over the bridge into industry 747 00:48:27,520 --> 00:48:31,360 where you are now. And they don't necessarily know how to navigate 748 00:48:31,760 --> 00:48:35,480 that transition. Is there anything that you 749 00:48:35,480 --> 00:48:39,120 could talk about in relation to that, the transition you took. 750 00:48:40,000 --> 00:48:43,600 Yeah, the industry is. Yeah. As you know, it's 751 00:48:43,920 --> 00:48:47,120 completely different world compared to the academia 752 00:48:50,570 --> 00:48:54,010 or going towards that direction, like improving the 753 00:48:54,010 --> 00:48:57,770 skill set of like a 754 00:48:58,250 --> 00:49:01,770 daily activity and close some small like 755 00:49:02,010 --> 00:49:05,490 task that's very crucial for, for 756 00:49:05,490 --> 00:49:09,010 industrial level activity compared to the academia. The 757 00:49:09,010 --> 00:49:12,770 academia is more like a long term plan working on something for 758 00:49:12,770 --> 00:49:16,170 very long term. But in industry you need to deliver 759 00:49:16,330 --> 00:49:19,940 something on some specific day and that 760 00:49:19,940 --> 00:49:23,580 makes you like decide differently. 761 00:49:25,020 --> 00:49:28,740 Probably your system that you want to close, it is not perfect, but it can 762 00:49:28,740 --> 00:49:31,980 do the job that you want to do for that specific task. 763 00:49:32,380 --> 00:49:36,220 That kind of, in terms of the mind level, you 764 00:49:36,220 --> 00:49:39,100 need to change your mind from like very 765 00:49:39,820 --> 00:49:43,460 optimized for the best system that you want to make 766 00:49:43,460 --> 00:49:46,500 it to the system that it worked for that specific task. 767 00:49:47,060 --> 00:49:50,500 And whenever that task finished, then the second 768 00:49:50,500 --> 00:49:54,060 task can like improve that third task can improve 769 00:49:54,060 --> 00:49:57,620 that and then you, you will reach to the final goal but in different 770 00:49:57,620 --> 00:50:01,180 way. In industry. Yeah. You have some 771 00:50:01,180 --> 00:50:04,900 commitment to deliver some result in, in a team, 772 00:50:05,300 --> 00:50:09,060 in a big team, like a 40, 50 people, hundred people, 773 00:50:09,060 --> 00:50:12,460 they want to accomplish some, some goal 774 00:50:13,020 --> 00:50:16,740 and that's, that's very important to do your task 775 00:50:16,740 --> 00:50:19,020 on some specific tool like task 776 00:50:20,700 --> 00:50:22,060 at deadline. 777 00:50:23,900 --> 00:50:24,620 Interesting. 778 00:50:29,100 --> 00:50:32,340 Any other further questions? I know we're getting close to the top of the hour 779 00:50:32,340 --> 00:50:36,100 and. Well, I've asked my favorite cocktail question. I know, I 780 00:50:36,100 --> 00:50:39,340 know. That is my favorite one. I love doing that. 781 00:50:39,580 --> 00:50:43,260 But it does force somebody to kind of explain. Right. Like, and I 782 00:50:43,260 --> 00:50:46,620 think that's also going to be an important skill going forward for anyone in 783 00:50:46,620 --> 00:50:50,460 academia. Like you can't assume that people even know. Like 784 00:50:50,460 --> 00:50:53,980 you said 10 to the 20th power of. You can't 785 00:50:53,980 --> 00:50:57,580 assume that anyone would know. What that even means. Right. Like, 786 00:50:58,460 --> 00:51:02,140 you know, but no, I 787 00:51:02,140 --> 00:51:05,740 mean that's important. Right. Because you know, at some point if you do want to 788 00:51:05,740 --> 00:51:07,940 make that bridge, you do, you're going to have to stand in front of an 789 00:51:07,940 --> 00:51:10,100 investor in some of the, in front of a customer. You kind of have to 790 00:51:10,100 --> 00:51:12,820 explain like what this is. Right. And you know, 791 00:51:13,780 --> 00:51:17,460 that's, that's. I think that's a challenge. Every technologist, regardless 792 00:51:17,460 --> 00:51:21,060 whether it's quantum physics, whether it's AI, whether it's, you know. Yeah. 793 00:51:21,380 --> 00:51:25,060 You know, has to come across. Right. The, the person, the hippo, which I don't 794 00:51:25,060 --> 00:51:28,260 know if you've heard that acronym before. Yeah. Have I used that acronym in front 795 00:51:28,260 --> 00:51:32,020 of you, Candace? Yes, I've heard it. It's one of, it's one of Your favorite 796 00:51:32,020 --> 00:51:35,540 ones. Also NIMBY is your other one that you. Yes, 797 00:51:35,540 --> 00:51:39,000 yes. So hippo. Not my bad. Backyard. I've learned that one. 798 00:51:41,640 --> 00:51:45,000 Tell them the hippo. You know, have you heard the Hippo? The hippo acronym? 799 00:51:45,320 --> 00:51:49,080 Yeah. Highest paid person's opinion. Yeah, I, I first 800 00:51:49,080 --> 00:51:52,520 heard it at Microsoft and somebody was like, well, you know, doesn't really 801 00:51:53,240 --> 00:51:57,000 not going to name who that hippo was, but that was, that was pretty interesting. 802 00:51:57,720 --> 00:52:01,440 You know, it was like hippo, like. And my, my mind went immediately 803 00:52:01,440 --> 00:52:05,050 to hungry, hungry Hippos. And then it was like, no, no, no. This means like 804 00:52:05,050 --> 00:52:08,610 highest paid persons. But you mean the highest paid person in any 805 00:52:08,610 --> 00:52:12,370 given enterprise is probably not going to have a 806 00:52:12,370 --> 00:52:15,810 background in quantum physics, right? Yeah, I think that's a safe 807 00:52:15,890 --> 00:52:19,330 assumption. Could be wrong. But again, you know, 808 00:52:20,530 --> 00:52:24,210 but yes, you have to. The people cutting the checks or, you know, you have 809 00:52:24,210 --> 00:52:28,050 to, you have to convince them that you're a startup, the value of 810 00:52:28,050 --> 00:52:31,810 it before they write that check. Um, and I 811 00:52:31,810 --> 00:52:35,650 think that this is something that, you know, this is 812 00:52:35,650 --> 00:52:39,410 what separates kind of like the famous entrepreneurs in 813 00:52:39,410 --> 00:52:42,650 tech from the ones who are not. Right. Like, you know, Steve Jobs, 814 00:52:43,290 --> 00:52:46,890 you know, was he the most 815 00:52:47,130 --> 00:52:50,490 proficient software engineer or computer builder? Probably not. 816 00:52:50,890 --> 00:52:54,570 That was probably Woz. Steve Wozniak. But what he could 817 00:52:54,570 --> 00:52:58,010 do is sell vision, sell the end result. Yeah, I think that's really, 818 00:52:58,560 --> 00:53:02,240 that's where the magic is, right? Yeah, this is the same like in 819 00:53:02,480 --> 00:53:06,320 academia world you have some Prof. Which 820 00:53:06,320 --> 00:53:10,160 is very, very knowledgeable, but they're not good at teaching 821 00:53:10,640 --> 00:53:14,400 and vice versa. This is the same thing in industry, 822 00:53:15,040 --> 00:53:18,880 like presenting idea and compared 823 00:53:18,880 --> 00:53:22,560 to the knowing knowledge deep of that kind of area. 824 00:53:23,120 --> 00:53:26,760 The same thing we have in academic also, like. Yeah. 825 00:53:31,400 --> 00:53:34,920 Well, I think that's, I think that's a good, a good point to, to 826 00:53:35,240 --> 00:53:39,080 leave us off on. I think that we've asked some really good questions 827 00:53:39,080 --> 00:53:42,800 today. I've loved everything we've been talking about. We really haven't 828 00:53:42,800 --> 00:53:46,200 had someone who's been able to focus on trapped ion for us before. 829 00:53:47,000 --> 00:53:50,720 So that is very exciting for us. Yeah, 830 00:53:50,720 --> 00:53:53,720 that really is. Because like, I know that's one of the core. That's one of 831 00:53:53,720 --> 00:53:57,480 the oldest, like, so like it's, it's something that like it's can't be 832 00:53:57,480 --> 00:54:00,800 ignored and we've kind of ignored it up till now. So. 833 00:54:02,560 --> 00:54:05,920 Yeah, great. That, I mean, that was a good 834 00:54:05,920 --> 00:54:08,880 opportunity to talk to you. And 835 00:54:09,840 --> 00:54:13,360 also I heard about the other podcast during the last 836 00:54:13,440 --> 00:54:17,280 couple of days. That was also interesting too. Thank you. Yep. 837 00:54:17,680 --> 00:54:21,120 Thank you. Thank you. Thanks for joining us. On impact Quantum 838 00:54:21,530 --> 00:54:25,250 Today's deep dive into. Trapped ion hardware showed just how precise. And 839 00:54:25,250 --> 00:54:28,650 powerful this technology can be. If you enjoyed this episode, 840 00:54:28,730 --> 00:54:32,410 subscribe and stay curious. More conversations at the 841 00:54:32,410 --> 00:54:34,650 cutting edge of. Quantum tech are on the way.