Skip to content

Farewell Microsoft-Research Silicon Valley Lab

September 19, 2014

Today, I choose to remember the five amazing years I spent in MSR-SV Labs (which are unfortunately closing). In a place with no boarders between research areas, I was free to follow my intellectual curiosity with colleagues I wouldn’t normally have the great fortune of working with. My non-theory colleagues have left me a much more complete computer scientist than I ever been.

My theory colleagues left me in absolute awe! Being surrounded by the creativity and brilliance of a unique collection of young scientists was such a rush. I initiated Windows on Theory because I thought that this rush must be shared with everyone. I hope that the readers of this blog got a glimpse of the breadth and depth of my theory colleagues. I am confident that they will make many departments and research groups much better in the following months and years. My only regret is every free minute I didn’t spend learning from these wonderful colleagues and friends.

My email for now will be, so drop me a line.

34 Comments leave one →
  1. September 19, 2014 2:09 pm

    I was an intern at Microsoft’s SVC lab the first summer they (you) had interns, in 2002, along with Frank Dabek, Arun Venkataramani, Chris Harrelson and Neha whose-last-name-I-have-forgotten (can anyone fill me in?). Google, a few blocks away, was still a modest-sized company. The air field down the street still didn’t have a want-to-be-quantum computer. And the nearest restaurant serving edible food was a 10-minute bike ride away (at that radius, there were several choices– downtown Mountain View, the taqueria on Old Middlefield with the awesome burritos).

    Cynthia Dwork was my mentor. Inspired by a manuscript that Irit Dinur and Kobbi Nissim were circulating, we looked into how to formalize privacy in statistical databases. What was “confidentiality”, anyway? It was surprisingly hard to pin down. The existing formulations were domain- and technique-specific, and all of them broke down under a little scrutiny. The first few things we came up also had, um, significant weaknesses (c.f., Chawla et al., TCC 2005). But Cynthia’s instincts said there was gold there to dig up, so we kept digging. A few years later, differential privacy emerged from a line of work by many people (Cynthia and I among them), and my research career veered from information-theoretic crypto into some weird mix of stats, algorithms and crypto that I still have trouble defining.

    Even back then, when there were only about twelve permanent researchers, SVC was a great place to be, with two future Turing award winners hanging around, a string of amazing visitors, and close ties to other research centers in the area like Stanford, Berkeley, and IBM.

    I was a regular, if not very frequent, visitor to SVC in subsequent years, and I was constantly blown away by the amazing people they were hiring, and the work they were doing.

    So I was sad to hear that SVC will be closing. I have no doubt that everyone there will land on their feet. But our community has lost a great lab.

  2. September 19, 2014 3:07 pm

    I was an intern at SVC in 2008, together with Katrina Ligett, Grant Schoenebeck, Alex Andoni, Krzysztof Onak, and others. I was hosted by Moshe Babaioff and Liad Blumrosen, who were awesome mentors. I hadn’t spent any time at Microsoft before, so I expected that I would exclusively be working with Moshe and Liad. But the lab was so active and collaborative that I ended up also working on projects with Frank McSherry and Kunal Talwar — I’d never seen such a collaborative academic environment before, and it was hard not to be overwhelmed with all of the interesting projects going on simultaniously.

    Its actually hard to find a computer scientist of my generation who has not spent time at Microsoft Research, either as an intern, a postdoc, or a visitor. Some of us spent enough time at various labs that we joked about having our PhDs from Microsoft.

    Its really hard to understand why they would close this lab. The researchers there are so talented that they will all land on their feet, and continue to do great things. The long-term loss will be to Microsoft.

  3. Thomas Holenstein permalink
    September 19, 2014 4:15 pm

    Wow, that’s a shock!

    I’m sorry to hear this. I completely fail to understand the reasoning behind this decision. MSR SVC was a place where so many great researchers did such great work!

    I am sure that everybody who needs to find a new job will find one which is great for them. I wish them all good luck in that (though I am convinced that nobody will need any of it).

  4. September 19, 2014 4:17 pm

    I interned at SVC in 2009 and 2010 during two wonderful summers. The first time around Kunal Talwar was my mentor, the second time Cynthia Dwork. The work I did at MSR was the starting point for what became my PhD thesis later on. As Aaron said, it did sometimes feel a bit like I got my PhD from MSR SVC.

    MSR SVC was a place where I found many friends, collaborators and true academic heros that became role models in my life. They also had amazing lime-flavored sparkling water and a rock solid foosball table!

    It seems to me that there are two kinds of places in our academic world. One is the kind that tries to hire people after they did their best work. The other is the kind where people go on to do their best work. MSR SVC was in the second category and that surely is rare. It was an ideal nurturing environment in every respect.

    I spent the last two days in shock and disbelief over the decision that MSR would kill off a lab that without doubt was one of the finest research labs of all times. It took more than a decade to grow it into what it had become. Microsoft at once lost much of the trust that the academic world had slowly acrued. Some might try to shrug this off by saying this is just how industry works. But that completely misses the point. MSR SVC was really academia at its best. That’s why it’s such a painful loss.

  5. Shuchi Chawla permalink
    September 19, 2014 4:48 pm

    I have very fond memories of the short but crucial and formative parts of my career that I spent at SVC, first as an intern (twice!) and then as a postdoc. This news is saddening at multiple levels for me. At one level because of the pain it will cause to friends and colleagues at SVC. At another level, because MSR was one of the last places that really valued basic research and publications for their own sake. There was no good immediate reason for Microsoft to do this (they are not in the same boat as AT&T, Yahoo, or IBM were when they laid off researchers). I see this as mainly a symbolic move. Is there no place for basic research in industry any more?

  6. September 19, 2014 5:23 pm

    I am still coming out of this shock, given that I was there this Wednesday for the Theory seminar, which as usual was intellectually very enriching.

    I did my post-doc in the lab+Stanford until I joined Yahoo! Labs two months back. My mentor was Cynthia Dwork. I had the privilege and honor to work with Cynthia Dwork, Muthu Muthukrishnan, Guy Rothblum, Kunal Talwar, Oliver Williams and Li Zhang. Those two years as post-doc were so enriching intellectually, it is hard to put in words. As Omer said, “My only regret is every free minute I didn’t spend learning from these wonderful colleagues and friends.” I owe a lot of my PhD and in general my professional life to Microsoft Research, and all my mentors there.

    I sincerely hope wisdom prevails and Microsoft realizes what treasure they have let go.

  7. September 19, 2014 6:34 pm

    I still can’t believe this is happening. I’ve been in SVC twice: first as an intern in 2011 hosted by Moshe Babaioff and later as a post-doc in 2013. I also had the privilege to work with many members of the lab (like Moshe, Omer, Noam, Muthu, Sreenivas, …) and also to collaborate with Search Lab downstairs. I made many friends there like Shiri and Abhradeep — my fellow postdocs at the time — the summer interns, the many visitors we always had and of course the members of the lab. I remember Chuck and Andreas had just bought a 3D printer when I was there, and I spent countless hours on their lab playing with it. I learned about systems, privacy and all sorts of other things attending the seminars. Moshe, Noam and I would have long conversations on algorithms and economics while eating sushi in downtown. It was a beautiful and enriching experience and one period of my life I will always remember so fondly. It is hard to understand why Microsoft would do such a thing.

  8. September 19, 2014 7:30 pm

    What a shame. The lab had so many strong people, and was a great place to work. I really enjoyed my 3 summer visits there.

    This decision makes no sense to me. Even if they don’t care about basic research, couldn’t they appreciate that codes of Parikshit, Sergey, and others saved Microsoft 1/2 billion dollars? I’m sure there were a lot more tangible benefits I don’t know about. Moreover MSR will now face huge difficulties recruiting and retaining at their other labs. Why would they do this?

  9. Vitaly Feldman permalink
    September 19, 2014 7:37 pm

    I’m very sad to see many friends and great colleagues lose their job. It must be a very stressful time for you and I hope all of you will find other exciting opportunities.

    MSR SVC was an incredibly lively powerhouse of theory and one that I always enjoyed to visit. Its closure is a great loss to the CS theory as a whole and will be acutely felt in the Bay area. I cannot see any reasonable justification to letting such stellar people go without any regard to their merits and contribution to the company. All this at a time when MS needs more than ever to be innovative. It is a huge mistake on MSR’s part and it is hard not to suspect that it’s motivated by some internal politics rather than company’s good.

  10. September 19, 2014 8:01 pm

    Truly sad news! I was at SVC twice as an intern: once supervised by Kunal and once by Cynthia. The work I did there became the centerpiece of my thesis, so I can definitely say I am one of the MSR SVC PhDs. A large fraction of what I do now traces back to things that we started during these unforgettable six months. It was as close to a perfect research environment as I have experienced. I remember Kunal and I had been working on a problem for maybe a month, spending hours on it together every day, and we had some progress but in some places we were still stuck. Then we had a conversation over lunch with Li Zhang about geometry and statistical estimation, and realized that Li knew exactly the right tools we needed to move forward. Basically any conversation at SVC could turn into interesting research: I ended up also working with Kunal, Cynthia, Li, Alex Andoni, Abhradeep, Grigory Yaroslavtsev (who was an intern with me). And then there was this general sense of excitement about theory and computer science, and the work done in the lab. I remember Cynthia gave a lecture about differential privacy in the stats department at Berkeley and ran well over time just because she insisted on saying a little bit about an algorithm from one of my recent papers.

    I made many friends during this time, Renato and Abhradeep among them. I consider Kunal and Cynthia my mentors, and they were tremendously supportive during my time there and during my job search over last winter and spring. I am sure everyone will land on their feet, considering their talent and accomplishments, but the community has lost something.

  11. Michael Mitzenmacher permalink
    September 19, 2014 8:08 pm

    Echoing others, I’m very sad to hear about MSR SVC. MSR SVC started with a “transfer” of people from the DEC (later Compaq) Systems Research Center, where I received the mentoring and intellectual stimulation that so many other have been describing here. For several years over the summer I regularly visited MSRSV at least briefly, and enjoyed the great research atmosphere. Like so many others, I will miss the place, and am hopeful that everyone there will be able to regroup and find new positions as soon as possible.

  12. salilvadhan permalink
    September 20, 2014 2:29 pm

    I’ve been overwhelmed with sadness at the loss of this amazing place. I’ve had two long visits there, which were life-changing experiences for me. A one-month visit during 2008 was my intense and exciting entry into research on differential privacy, which is now the center of my research. Through free-form discussions in the lab and (especially) the weekly theory group hikes, I learned about this beautiful new area and started many collaborations and friendships, which led to papers with Cynthia, Guy, Ilya, Kunal, and Omer. My next visit was throughout my last sabbatical (2011-12), where I continued to have fun and inspiring interactions with all my friends in the lab and also started new collaborations with Parikshit and Raghu. Beyond having such an amazing group of researchers, what made the SVC lab so special was the incredibly warm atmosphere of collaboration and friendship that permeated the lab. Everyone talked with everyone both about what they were working on and what was happening in their personal lives, and were supportive of each other in every possible way. And the resulting research was revolutionary (differential privacy being just one example). I cannot fathom how Microsoft would fail to realize the treasure it had in this lab and in these researchers individually (who will have no trouble finding great new homes). The places they go now will be very lucky to get them, and while we may not be able to recreate it, at least we all can continue to enjoy fond memories of the extraordinary place that was MSR-SV.

  13. September 21, 2014 3:19 am

    I too am still trying to come to terms with this loss. I had an amazing visit at the lab in December 2005 (way before I joined MSR) which eventually led to a paper, but more than that, to me learning about this new area of differential privacy and getting to collaborate with wonderful people such as Cynthia, Kunal and Frank. Since then the lab only became better and stronger, and consistently hired some of the best theorists anywhere. I always wanted to visit it again, and always felt that, being at MSR, the opportunity would naturally arise at some point, and there is no rush. Now it is no more.

  14. September 21, 2014 10:33 am

    I have very fond memories of my visits to the lab and the truly magnificent team of researchers therein, many of which I came to consider personal friends.

    My warmest wishes to those from MSR SVC whose lives are now in turmoil.

  15. Guy Rothblum permalink
    September 21, 2014 7:58 pm

    The MSR lab at SVC was a unique and wonderful place. In my view, it was the epitome of what a successful industrial research lab could be. I was first introduced to the lab halfway through my Ph.D., during a career-altering internship with Cynthia Dwork. A second internship followed, as did many visits and fruitful collaborations. I was ecstatic to eventually join the group as a full-time employee. There was a palpable sense of excitement and possibility within the group, and I was continually amazed by the brilliance and creativity of my colleagues and of our phenomenal visitors and interns. The sudden loss of the lab is shocking and sad, but I’m grateful for every moment I got to spend in this wonderful place.

    Just as importantly, I think I can speak for all of us in saying that over the past few days we have been overwhelmed by the support, encouragement and solidarity of the theory community. Friends in need are friends indeed, and we are all blessed to be part of such a warm and caring community. Thank you to all of you that have reached out for your offers of encouragement and support.

    • September 21, 2014 9:08 pm

      As everyone, I am in shock over the news. I visited MSR SV several times. All of those visits were very stimulating, thanks to the very talented people that worked there, came for internships or just happened to pass by. My longest visit took place in the summer of 2012. It was incredibly productive: over two weeks, Alex Andoni and I co-wrote one paper and laid the foundations for another. Alas, this won’t happen again. Closing the lab will be a big loss to our community.

  16. Parikshit Gopalan permalink
    September 21, 2014 9:10 pm

    Like Omer, I look back at my six years at MSR SVL and my time in the theory group: learning coding theory from Sergey, pseudorandomness from Omer, probability and analysis from Raghu, game theory from Noam, a lot else from Kunal, table tennis (both practice and theory) from Udi, having an exemplary leader like Cynthia, incredible visitors, interns who walked in as mature researchers on day one … and feel incredibly privileged to have had this experience. And I will miss it.

    History tells us that research labs are mortal. Like mortals, they are finally judged by their accomplishments rather than their longevity. Even through the narrow lens of topics I understand, I think of differential privacy, the connections to convex geometry and discrepancy, locally decodable and testable codes, locally repairable codes for data storage, massive strides in pseudorandomness, nearest neighbor search, distance oracles: beautiful ideas that have had and will have future impact on both theory and practice.

    Farewell MSR Silicon Valley Lab, you will be missed dearly.

  17. Anupam Gupta permalink
    September 22, 2014 1:09 am

    The MSR SVC lab was a fantastic place — not only in the sense of being extraordinarily good, but also in the sense of being slightly incredible, unbelievable….

    Over the course of many fruitful visits, I had the fortune to meet and work and learn from an amazing group of people, leaders of their areas and of our community, colleagues and friends. They were doing top-notch research, but to me it seemed that they were making connections to practice and the company where possible (which I would learn about from lunchtime conversations, by lab meetings, and by those collections of Tetris-like granite cubes, hills of which grew on people’s windows and shelves, more numerous with every visit). Which makes recent developments more unbelievable.

    There are only a small handful of labs doing fundamental research nowadays, and now one of them — a very close one to me, personally — has been taken down in its prime. It is a real loss.

  18. AnupamG permalink
    September 22, 2014 1:11 am

    The MSR SVC lab was a fantastic place — not only in the sense of being extraordinarily good, but also in the sense of being slightly incredible, unbelievable….

    Over the course of many fruitful visits, I had the fortune to meet and work and learn from an amazing group of people, leaders of their areas and of our community, colleagues and friends. They were doing top-notch research, but to me it seemed that they were making connections to practice and the company where possible (which I would learn about from lunchtime conversations, by lab meetings, and by those collections of Tetris-like granite cubes, hills of which grew on people’s windows and shelves, more numerous with every visit). Which makes recent developments more unbelievable.

    There are only a small handful of labs doing fundamental research nowadays, and now one of them — a very close one to me, personally — has been taken down in its prime. It is a real loss.

  19. Vishesh permalink
    September 22, 2014 4:12 pm

    Like others, I still can’t believe the news! I was really fortunate to spend a few weeks of my past summer at SVC working with Cynthia. Every day was filled with excitement of doing wonderful research, I learnt a lot of things just by a few conversations and made some great friends. It was a truly amazing, one of a kind place to be treasured – a group of some of the smartest people in the world. I do not understand why Microsoft would make such a move. I am sure that everyone at SV will have no difficulty in finding a new home, and those places will be truly lucky to have them.

  20. Denis Nekipelov permalink
    September 22, 2014 7:06 pm

    I am completely stunned! I spent a year at SVC as a post-doc and it was one of the most intellectually stimulating years in my professional life. Just talking to the researchers there has defined many directions of my current research. There will be no other place like SVC!

  21. September 23, 2014 3:50 am

    You got to love our TOC community! It’s hard to imagine colleagues that will be happier for you in your successes and more compassionate in your setbacks. The offers of help by our immediate neighbors (Berkeley and Stanford, see Luca’s kindhearted post), were quickly followed with a flow of wonderful intentions from colleagues all over the world. This is how good fortune looks like 🙂

    If you want to contact our group, I’m happy to serve as the point man (or to supply new emails for any individual colleague you are looking for).

  22. Grant Schoenebeck permalink
    September 25, 2014 1:35 pm

    It is very sad to see such a vibrant institution crumble. Many from our community helped build MSR-SVC for years. Indirectly, we have all benefited from their hard work.

    I was fortunate enough to directly benefit from MSR-SVC both as a summer intern in 2008 under the guidance of Thomas Holenstein, and as an occasional visitor during graduate school. Fond memories are numerous and include the comradery of the interns (both intellectual and social), the weekly theory hikes—Cynthia led hikes in local parks where all the theorists could talk about research unencumbered by any distraction besides nature and the sometimes grueling pace—and foosball, especially with Thomas, Udi, and Mihai Budiu. I learned about many new (for me) topics including differential privacy. I wrote one paper with Thomas Holenstein, and have now written several papers with two of the other interns, Aaron Roth and Katrina Legitt, there that summer (including two on privacy).

    I am confident that the (former) members of the group will greatly enrich other institutions in the near future. However, I fear that the unique blend of talent, creativity, and infectious enthusiasm that was MSR-SVC is now lost.

  23. krishna permalink
    September 29, 2014 9:22 pm

    It’s very sad an abrupt shutdown of a great research institution has raised so many question about the research, capitalism and politics of this country. There is no other joy other than thinking freely , creatively and thinking about problems which doesn’t answerable to stake holders is the best best joy one can ever get in the world of exploration of knowledge. The joy of discovery is great but should be nurtured and protected by the government which is failing to do it’s job instead favoring the corporations by cutting taxes for them who in turn hoards the money. Why corporation start research institution to benefit them to have fat pays? Research is not about output and corps knew this. It’s like throwing the baby with bath water what they have done. Sathyan nadella is just a business man not a researcher hope he understands what is research? I still wonder why BGates did not react to it?

  24. krishna permalink
    September 29, 2014 9:34 pm

    In addition to the above comment if so many brilliant people and professors are commenting how the lab nurtured, sculpted them into a world class researchers who in turn become professors and churn out brilliant researchers some of them will practically impact the world. Who is going to benefit it’s the company. I have never seen people being so mad and praising the researchers and the Lab. Normally when people were laid off i have seen madness, angry but never heard of such praise being showered on an institution by other people. Microsoft executives should read this blog and realize what they have done is wrong. Is this an indication that company is now down the hill?


  1. Microsoft Research Silicon Valley Campus | in theory
  2. Riding the Wheel of Samsara | Windows On Theory
  3. Shtetl-Optimized » Blog Archive » Microsoft SVC
  4. microsoft silicon valley TCS research lab shuts down— easy come, easy go | Turing Machine
  5. Microsoft Closes SVC | Gödel's Lost Letter and P=NP
  6. Sum of Squares: Upper bounds, lower bounds, and open questions | Windows On Theory
  7. Microsoft Research SVC and Applied Theory | Outsourced Bits
  8. Out the Window | Windows On Theory
  9. Stanford, Here I Come | Windows On Theory

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: