 Tako toga vrteva, ki sem izgleda, se teoretik je vrteva izgleda v rovnih rovnih rovnih in vsega rečenja z Leviofe, Kitiyev in nekaj način na simulaciju Manuel Pino. Zdaj, izgleda izgleda izgleda s grupom Joe Martinis, Robin McDermott in Universtiv Medizon and Nakamura when he was at NEC and more recently took an博 to the lab, so. And. So the problem this problem started. In the late 80s when Fred was a graduate student in the group of John Clark and Berkley and basically vznikali, da počke pražami počоловjali mene iz izličenje, postoži z celikonoksaj ei tako zračo. A začali, da prilud nješte praze občasno očeljali za njič oju tudi države počovdi. Io spanno, da sem podvisljala štak sečiti poten najbolje 0,5 Kelvin. Videli, da še dashal dve vzvečenje si vse skvidi, kako je izpravljano, universtvana. Mela najskega vzvečanje vzvečavu, in ozvalo drugo što je tajh vzvečenča, ki je začeli izpravlje, in svetil. In tudi pričali možno, da je to mikroskopik origine in to je začelične, in je začelične. In... Ne znamo... ...zato je počkala in pričo je to izgleda, in tudi je vsečen, da je vsečen, pravno vsečen, in tudi je vsečen, in tudi je vsečen, in vsečen, in vsečen, vsečen, in vsečen, in vsečen, in vsečen, in vsečen, in tukaj superkondat in kubit, je skviji, in še bi jih dazili, že v nekaj neseljenih skviji, ki se v Aluminjne tukaj, nekaj na sveti, zelo na vse na vse, z češlji, nekaj na našeljenih. Nekaj, nekaj našeljenih se je zelo, nekaj na sveti, nekaj našeljenih nekaj na našeljenih, nekaj našeljenih, nekaj na šeljenih. To mi neče zelo z judgem. Zelo pošličen trsi nekaj, kako je v skupci, ki je oč원ila in neradecov i ta communitiesa, pošličena števn pošličen, je bilo všetoma Nbeča exist. V 2007 boljali bolj beguni in taka zaluminje, še in ruk jel ljubil, zash poskeljilo napšenje, zaluminj themselvesh, in še pošličena bilo pristit, zaskolj id whose what the west would have seen previously. Namely that they found the 1 over I F noise, the was a very low frequency noise. The amplitude was almost 1 over F. They found again that it was temperature independence below 500 mC. ta je se, da bi bilo sezvršo, že sezvršo je zelo na zelo vzelo, in je bilo sezvršo izelo na tudi vzelo in zelo, da je to nečo začal početnje. Zdaj na koncu še nekaj nežel. Zelo, da sem stavila, da bi sezvršo, da sem že nekaj nežel. zato se je vstajen, da tako izgovoril več nekaj dobroj temperatru, seveda, da vštje, da je izgovoril in izgovoril na vse sem, zelo tako zelo na vse mačnje energijski vse. Imediatega, je bilo vse vse, da sližujemo dušnje zelo, zelo nas ker je izgovoril, zato na našem težde, nekaj bo nekaj bolji vse, ki je izgovoril vse zelo, in začne vse, nekaj nekaj elektroni. In zelo vzlade, da je vzlade, da je zastračil nek njučenom njačin. Svečo, da smo zvalili informacijo v komandu v Nakamura KSV, da smo zvalila vseh ekko-experimentin in da so sveti, da je to njič nekaj nekih neko nekaj nekaj nekih nekih nekih nekih nekih nekaj nekih nekih nekih nekih nekih nekih nekih nekih nekih nekih nekih nekih nekih neki. In zato smo počeli, da se pričočilo, da se pričočilo, da je paramagnetik in superkondakti in solatori in terfajs. In izvah, da smo se pričočili, da se pričočilo, ker, da se pričočilo, da se pričočilo, in solatori in solatori, ki se pričočilo, da se pričočilo, da se pričočilo, vse se zemljate in ide prezistiti in vzelo, da je zelo vzelo, da se na zelo in vzelo, da je zelo, da je zozvar, da je zelo, da je zelo. Tako da se veči. Vzelo, da je veči uči, da je zelo, da je začala, da se da ne pomembna, in tako, so pričaš, da se zelo. Spind, dynamic, interaktion, was mediated through the electrons in the superconductors. And we gave some estimates, so the mechanism is r, k, k, y interaction between the spin and the typical energy scale of the interaction by typical density that we knew from experiment in other context of this typical paramagnetic spins gave us this sort of scale that was at 50 mRk. Pa posledaj, ki je vzlučila, da izvah je je generativna, legislative tron srednji zelo, da je najbolj povisan se, da bi je zelo, da jaz vse pas vičiči na različen način. Vzlučili, da je vse način že ne prihladil vse zelo, Popravili, da bi聽ili mechanizm na vade zakočnjenak. Preko jih začelil, da se različaj sem všem nemiskem dinamijan priest. In smo na toga, da je nekako svetišno, je vse početnega, atakaj, priko pekni, če ti pranočit. So pa je priča predstavila všem. In, da se početno početno dajte komodstvo, se početno se prišlo, da je to bo zelo, izveča doznih tega jezavne kanala, kako še se početno se prišlo, da so na ročnega vsega. In vsega nemačna vsega je, da when you drive the electricity through the electrical current through the squid, basically the magnetic field has a shape. zelo je nekaj veliko vziv, zelo je zelo vziv, zelo je zelo vziv. If you keep into account this surface current density for this current, basically what you end up is that by using diffusion and this geometrical effect you find that the spectrum noise could be 1 over f. But there is a crucial point that it is 1 over f depending on some frequency. Namely there is an energy scale that is this ratio between the diffusion constant and the width, the square of the width of the squid. And what you find out in this theory basically is that for frequency that are much smaller of this particular value in this frequency the spectrum should flat out should be white, saturated. While if you go to frequency that are larger of this particular value of the frequency the spectrum is 1 over f. And this is, notice that this frequency depends on the dimension on the width of the squid. So this 1 over f noise that we find for example we can find it this value of this constant is of this 10 minus 2, 10 minus 1 only for very big squid like the one that was measured. But clearly this value is quite high in the frequency range if you go to the squid loops that Martinis was using at the time. So what we were seeing by using just diffusion is that we couldn't understand the low frequency noise in very small squid. And another important feature that we find out is that there is this geometrical ratio that is depends on the radius of the squid and the width of the squid and this is basically due to the fact that you are assuming that the spins are only on the surface. And when we we did this theoretical work immediately after there were new work by a group of Robin McDermott at Madison, Wisconsin and what he studied magnetism in squid at Milly Kelvin and what was the major finding of this works was that he was able to measure what is the typical spin density that you have in the squid and this is value that was confirmed by many other groups then so it's 5 times 10 to 17 meters minus 2 and the other important thing that were coming out from this study was that the interaction between the spin was rather strong and for example here you see that we studied sociotability and dependence on temperature in this squid and in some device you see as a sort of cast that is reminiscent of the freezing in spin glass and you see that the value is around 50 milik k and this was in agreement with our first estimate in our RKKY model and in the same here basically other group in Pasadina did the geometric study more carefully and they found that correctly this scaling so I think that everyone I agree is that the spins are really on the surface and also what they clearly show is that the noise doesn't flatten out so it keeps growing it keeps being 1 over F basically so it means that one is to to see something else diffusion is not sufficient in this frequency 1 over F noise so what we started to think at the time was that in our diffusion model what the main things that we are neglecting is the fact that we are thinking that we are neglecting the fact the possibility that spins can be very close what does it mean that two spins that are very close can be locked in some singlet or triplet and in order to flip what they have to find is to have a high energy in the system so that they can flip and since these are sort of rare event if we somehow managed to show that these spin ensemble can generate this high frequency energy by itself then we can find a way in which this spin can be excited and create and being sort of flat to ethos and create noise so the question that we solved was the following the more a generic work it means that we start from an ensemble of spins and these spins have Heisenberg and Miltonia and the typical scale of the interaction of the spin is much smaller than t so we are considering that our system is at high temperature and so it doesn't exchange with the standard path like phonon or electrons so this is the case that we are considering here and what we want to calculate to see is that nevertheless because of its inherent non-linearity and because of the fact that there can be simultaneous excitation of many spins in this system one can find some frequency at high sorry, some correlators at high frequency much larger than the typical interaction scale so the theory that we did was just to consider anisotropic spin Heisenberg and Miltonia and what you need to calculate we are in the high temperature limit so this is the kind of averaging and what you need to calculate is basically the correlator that can be written as a Taylor series and we want to calculate the asymptotic behavior of these moments at any equal to infinity and find basically how which is the value of these singularities and this is a problem that we managed to solve exactly and of course the main difficulties that we had was basically the fact that you do some assumption like for example that you are considering the fields that the single spin of the interaction with the others and we assume that this spin is Gaussian but nevertheless the problem remains a bit difficult because this exchange field is a matrix and so when you try to calculate the coefficient you end up with problems that is highly non commutative nevertheless we manage analytically to solve this and what we found that was useful for our problem is that the correlator of the spin correlator the free transform has a high exponential Taylor tau frequency so this work we managed to find the bounds calculate basically this value and so this means that this ensemble of spins can generate high energy to excite these close pairs give an estimate of what could be the number of the rare spins that what could expect and immediately basically what finds out is that this mechanism gives you well one over F noise with some logarithmic correction so at this point we were quite we were quite happy about this because we found a way in which we could explain noise type frequency in combination and noise very low frequency do a combination between diffusion and these rare events but then in 2009 there was an experiment by the group of McDermott that changed completely our understanding and basically what McDermott did is the following he studied not only flux noise but also inductance noise both magnetization and susceptibility in the noise in the squid and he achieved this because he has a configuration in which he can inject some current so this current basically generate magnetic field along the plane, the x direction and the magnitude of the fields that he has is around 100 micro tesla and he studied these two spectra both of them are 1 over F but was very puzzling for us is that it founds that when you go at low temperature you see that these two noises are highly correlated and when I say highly correlated it means that the value is very close to 1 and that clearly means that first of all these noise both in the magnetization and susceptibility must be generated induced by the same phenomena and also in other things that he founds that was quite relevant for us when he follows the path of these noise he could really see some very big jumps these are jumps in the channel of the susceptibility and you see they are equivalent to changing the flux about 500 micro F0 and this is equivalent to the fact that he sees that there are fluctuations due to formation of long range magnetic ensemble and the number of these spins in this ensemble as t of the order if you do calculation t 10 to the 3 10 to the 4 this experiment it was really very difficult to try to think what could be a microscopic mechanism that was able to explain this because if you think that these spins are paramagnetic and so you think that maybe they can form a spin glass immediately you see that this correlation is zero and also it's very difficult to achieve such high degrees of correlation even if you think that some cluster with some special fractal structure can be formed and more recently in the last year we asked McDanwood to repeat again this measurement just to see if it was not just an isolated event but it turns out that even in both in aluminum and Iobium squeeze these cross correlation are present and they are as universal as the flux noises so the possibility the only possibility that we thought could be able to explain this is the fact that somehow what you must have so you have the formation of this ferromagnetic cluster and you must have some pieces ferromagnetic cluster that flips and they change the position from a ferromagnetic state to a paramagnetic state so in this case if you have some mechanisms in which a ferromagnetic in which you switch between a ferromagnetic and a paramagnetic state then you have variation of fluctuation of magnetization that are perfectly correlated but the problem was what could be these so the thing that we think is the following that basically a feature of having this RKKY interaction is that you have a broad range of interaction and what we know is that if you have RKY mechanics depending on the concentration what you can find you have a transition between a spin glass or a ferromagnetic so this was simulation that we did recently so basically we were considering this Hamiltonian with RKY interaction and we saw clearly that for some value of concentration the state of the systems that start with some magnetization either maintain the magnetization so for example is ferromagnetic or lose it and so become a glass and so basically if you allow that there are some high density concentration of these spins then you can have with this RKY formation of this ferromagnetic and the point is the following that similarly to the argument that we reason when we were we were discussing the noise in the spin with this reappair even in this feature what you can find is some spins configuration in which the spins are very close but in this case in this case live very long and basically what happen is that they can decay only by emission instead of many flip of a large number of spin ways so and another thing that we expect that is reasonable to expect here is that if the interaction between the spins are strong and they are sitting very close to the boundary so one expect to have some anisotropy so what we did here is the following we just took spins that we choose a random configuration spin in 2D and these spins are classical we simulate classical block equation and we introduce anisotropy and then we let the system evolve so what is represented here and then we start this little animation is basically the following so we the spins are initially aligned is a ferromagnetic state along the in the field along the plane so it has the x direction and this spot here is the magnetization along the z direction so when you see white means that there are spins that are in the x direction and this spot that you see basically is showing the magnetization of the spin along the z direction and when I consider big area of this spot it means that I have a space in this ensemble where I have revolution of few spins that are aligned with the z field and let me see what you see so I just want that you see this is a devolution of time and what I want that you focus at something that will happen here basically what you see here is that at a certain point you might have in this ensemble of dynamical spins the formation of some states that has a magnetization along the z and that lives quite long so and this clearly this sort of this state doesn't affect will affect create noise in the magnetization and it create noise in the sociatability so we think that a possibility to explain why there can be the existence of this very strong crystallization is due to the fact that in these systems basically you might have the possibility to have this long lift space that are very similar to the breathers and the kind of physics that we are discussing here is somehow similar to the physics that is discussed in a narration paper by Boris and other when they consider Josephson chain so I also want to show you something else so we did also simulation to see what is the spectrum that we expect and what you see is that actually in this case we consider just a block equation without anisotropy but what you see is that this was a simulation of large classical systems with these 2000 spin and you see that the noise that is generated is one over half and this one over half with this coefficient and by the way this one over half is never really one over half but it's always one over half with coefficient that varies from 0.8 to 0.9 we don't have yet done the same analysis for the sociotability namely I don't have a spectrum to show you but we expect that similar spectrum will happen also for the sociotability and it's clear that of these long live states because this low frequency noise is generated by these long live states we also expect to have basically 100% correlation between the sociotability and the magnetization so I want you to somehow conclude this talk by saying an important thing that it is very recent that we became aware of this fact a few weeks ago when we were at a meeting in the states basically there was a general feeling after these people started to work with this noise in the superconductive qubits that okay this noise is difficult to get rid of but it's a very low frequency noise so if we think of implementation in quantum computation we can echo and we can get rid of it and now there are very recent experiments that are measurement done in the group of Will Oliver at MIT Lincoln Lab that unfortunately shows that something really very surprising but that indeed this one over half noise can be a serious problem also for relaxation and what they do is that they built several qubits like here they are measured 21 qubits and these are qubits that are flux qubits and are generated with very high quality aluminium junction and they are focusing on studying the relaxation of these qubits and the first things that they were achieving is that they now have a qubit fabrication that is highly reproducible so they can generate very different qubits at different frequency and what they find out is that the major T1 is forced very nicely of all the qubits and agree with some predicted T1 where you consider when you put in this T1 basically the noise spectrum not of a homonich noise that they would expect at this high frequency these qubits are at 5 and 10 gigahertz but it fits very well all day fall into a calculation of noise with one over half spectrum with the amplitude of the low frequency noise and this is something that was hinted in some experiment that they did in 2011 where they could see that there were measurement in relaxation that when you extrapolate the level of one over half noise were matching this value but here is more consistent and so they claim really that this one over half flux noise seems to cause qubit relaxation and this is I think something remarkable because I think that this is one of the few example I don't know any other example in physics in which you can have a scaling one of bfv that cover so many orders of magnitude like 13 order of magnitude so I think that still there are things that worry sounds but somehow we have the feeling that with this understand with this fact that we are seeing that there are these long live states in these spins maybe this is the right way to explain to understand this mechanism and as you see it seems that it is quite important to be able really to understand that because echo will not help and when we start to couple and to think maybe to do some error correction then in order to couple you need the squid loops and then you see that this noise will become important even at this very high frequency so with that I think I said more or less what I wanted to say so I just want to say that I'm very I want to thank the organizers to give the possibility to be here and to congratulate Boris with this birthday and that's it so I think