Running while installing while building is tough!

Steve Nahn

Whew! I’m sure we all feel like there’s quite a lot going on these days. In CMS, we are amidst momentous times in several dimensions (but not extra dimensions, or at least not yet!).

Running the CMS detector has been a challenge, but a challenge every experiment wishes to face — too much data! In 2016 the LHC delivered 41 inverse femtobarns of integrated luminosity, a whopping 50 percent more than forecast. Coping with high instantaneous luminosity and the accompanying “pile-up” of roughly 30 to 40 overlapping interactions in a single bunch collision did push the capabilities of the experiment, but we succeeded in capturing 92 percent of delivered luminosity for offline consumption. There, the challenges faced were just as serious due to the resulting unexpectedly large data volume to reconstruct, store and analyze. But again, we must have done well, as the first comment of the January U.S. CMS Operations Review says, “The U.S. CMS Operations Program should be commended for an excellent record of successfully meeting the operational needs of the international CMS Collaboration over many years.” And we’re just about to see just the beginning of the fruits of all this hard work with winter conferences around the corner.  Sorry, no spoilers here, you’ll just have to wait and see.

We’re also on the precipice of the harvesting of early fruits of the CMS Phase 1 Upgrade, a subject near and dear to my heart. During the current LHC shutdown we are installing the pixel detector upgrade, which you’ve heard about here and there if you’ve been reading the Fermilab news with regularity. Over the last three to four years we’ve built a new pixel detector, which, relative to the old detector, has an additional layer of tracking detectors and improved electronics, and yet less mass (and therefore less loss of resolution due to scattering off the detector components), all of which will provide high tracking performance in the face of the continuing high instantaneous luminosities of the next five years.

Another facet of the Phase 1 upgrade, the forward hadron calorimeter front end, is also being installed. This involves a reconfiguration of the current photodetectors, doubling the granularity for more effective pile-up mitigation, as well as new front-end digitizing electronics, now with the addition of timing information, allowing discrimination of signal from background using another dimension (no, not an extra one).

But that’s not all for Phase 1! Finally, a rather dire forecast of degradation of the light output coming from the endcap calorimeter in 2015 instigated an acceleration of the endcap front-end upgrade, completing it in two years — half the time originally planned —to be ready for the current installation window.  However, in those intervening years we learned the light loss is not as severe as originally thought, and after revisiting the risk-benefit analysis, we’ve made the prudent decision to not change too many things at once, so this installation was deferred until the next window. It is both very satisfying and a little breathtaking to see all the hard work put in at Fermilab and the 30 CMS institutions involved in Phase 1 come to fruition in the installation season of 2017.

Members of the CMS collaboration work on the CMS forward pixel detector, which will be installed in the Large Hadron Collider later this year. Photo: Reidar Hahn

Yet, that is still not all that is going on. The big sister to the Phase 1 upgrade, the High-Luminosity LHC upgrade, has been forging ahead, becoming a full project, both on the accelerator side and the detector side. The accelerator side is the subject matter for Giorgio Apollinari, since he’s running that project. On the detector side, Fermilab’s Vivian O’Dell and Cornell University’s Anders Ryd, with a lot of assistance from the Fermilab CPO Mike Lindgren and his team, are leading all of us through the trials and tribulations of “projectizing” the upgrade, which has a scope encompassing essentially all of CMS, including the tracker, the calorimetry, muon detectors, data acquisition system and trigger. This design and prototyping phase is well under way, and design reports will be popping up with some regularity in the next two years, which, after corresponding reviews, will take us to the production phase in 2019, lasting to 2023. In parallel there is a burgeoning effort to explore and exploit new computing paradigms that we will need to cope with the 3000 inverse femtobarns (or, given the first paragraph, maybe more) that the LHC plans to deliver between 2025 and 2035.

When meeting friends in the elevator at Fermilab, my response to “how are you” lately is usually “busy” or sometimes “a bit tired.” For a while I thought maybe this was just mid-winter-in-Chicago malaise, but taking a step back and writing it all down, I guess I can see where this feeling comes from.

Steve Nahn is the head of the CMS Phase I Upgrade Project.