I presume you refer in part to the paper by Chamseddine, Connes, and Mukhanov. This is an interesting development. It find in particular the fact that they have discovered an 8-fold cyclic structure in this also interesting. This suggests something related to Bott periodicity, maybe something involved with normed division algebras, maybe E8 ... . This idea of there being a quantum of geometry is curiously dual to the smoothness of geometry one gets with the moduli space developments of Uhlenbeck, Donaldson and others with gauge fields on 4-manifolds.
This may connect up with the experimental results by the Fermi and Integral spacecrafts by NASA and ESA respectively. The observation of photons at different wavelengths found they arrived from very distant Оі-ray burstars, from very short to longer optical wavelengths, at the same time. This is dual to the prediction of loop quantum gravity, which predicted a dispersion of photons due to the graininess of spacetime. The observation is of spacetime structure with a very long baseline. In effect the transverse momentum that scatters off spacetime, or that involved with detecting these photons, is tiny. This compares to a measurement of particle that attempts to localize it in a tiny region of spacetime. This results in a massive transverse momentum transfer and the particle scatters wildly. If one attempts to localize the particle to near the Planck scale then spacetime time will behave in a quantum or discrete manner.
I am going to write on the AdS/CFT holographic principle. There is an interesting way to think about this, which is not in the literature or at least I have not seen it. It is to see that the AdS_5 is a moduli space for the CFT. In this way a Yang-Mills gauge field is a boundary of a moduli space, and the set of gauge connections as themselves the holographic projection. I will try to leave off with how this connects to the Cartan matrix for the E8 group.
The AdS/CFT holographic principle is related to the results on gauge fields in four dimensions. The moduli space for a quaternion H or SU(2)^2 = SO(4) bundle is a 5-sphere, and a hyperbolic spacetime form of this is the AdS_5
We consider pairs of quaternions with R^8 ~ H^2. This has the quaternion inner product rule
= p_1q-bar_1 + p_2q-bar_2
that is a real valued product on R^8 The vectors with real norm = 1 form the seven sphere S^7, which is a 3-fibration over S^4 ~ HP^4. This is just a quaternion version of the Riemann sphere. The principle bundle ПЂ:S^7 --- > S^4 is formed from the imaginary part of one quaternion as an internal space over the other set of quaternions with unit norm. The bundle map then sends one set of quaterionions into an H^1 inner product (q_1, q_2) --- > and the remainder is the projectivization of H^2 - = (pq_1, pq_2). The fibration is the left adjoint action of the SU(2) group. The bundle contains vertical and horizontal portions for the group action and the base manifold respectively. The vertical tangent bundle is then the set of p in Im(H) that defines a (pq_1, pq_2) as a 4-space with the group action by p on every element. The horizontal component, defined by H with elements (q_1, q_2), have the one form
П‰ = Im(q_1dq-bar_1 + q_2dq-bar_2)
and curvature form
О© = dq_1/\dq-bar_1 + dq_2/\dq-bar_2
where we consider a "slice" where dq_2 = 0. The horizontal bundle subspace is spanned then by ∂/∂q_1 and this curvature form is the pullback HP^1 of a self-dual form. The horizontal action is then two copies of the sp(2), sp(2)+sp(2) ~ sp(4) action on S^7 projected onto S^4. These group actions form the quotient of the H^2 group SL(2,H) ~ SO(5,1) with SL(2,H)/(sp(4) ~ SO(5)). This is the moduli space, which has the dimension predicted by the Atiyah-Singer theorem.
This is the Euclideanized form of the theory, for we really want a Lorentzian version of these spaces. The z_0 component of the quaternions is modified so the conjugate of z_0 is -z_0. This gives us the norm zz-bar = -|z_0|^2 + z_1^2 + z_2^2 + z_3^2, and the unit condition is replaced with the zero or null condition on a light cone. Our moduli space is then
SL(2,H)/Sp(4) ~ SO(5,1)/SO(5) = B^5 = the five dimensional ball,
and in the Lorentzian form it is
SL(2,H)/Sp(3,1) ~ SO(5,1)/SO(4,1) = AdS_5.
Of course AdS_5 is the anti-de Sitter spacetime in 5 dimensions. One perspective on this is to say that one of the Sp(2) symplectic groups with Sp(2) ~ SU(2) is replaced with sp(1,1) ~ SU(1,1) for boosts instead of rotations. Another perspective is to say that two copies of SU(2), or SO(4) is contained in sp(3,1), but where the Lorentzian change in metric is not on the group actions.
This theory is connected with the Dirac operator, and Connes' work is centered around this as well. I think these two developments are related to each other. Maybe these two are related to each other according to S-duality or П„ = Оё/2ПЂ - 4ПЂi/g^2 for a general coupling constant for the two theories. Connes' work with the Dirac matrices may connect up with Atyah's work on the Dirac operator and the elliptic bundle in some sort of dual theory.
Cheers LC