Racism and HPC: a Special Podcast

June 29, 2020

Promoting greater diversity in HPC is a much-discussed goal and ostensibly a long-sought goal in HPC. Yet it seems clear HPC is far from achieving this goal. Recent U.S. events, most poignantly the killing of George Floyd, have again stirred unrest and discomfiture, forcing the U.S. to face entrenched, explicit and implicit racism. Like all segments of society, HPC is a carrier of the problem and, one hopes, an agent of constructive change. Racism, of course, is not just a U.S. problem but worldwide. In this panel, five HPC leaders from underrepresented communities discuss their perspective and experiences. The panel was conceived and moderated by Intersect360 Research founder and CEO Addison Snell.

The panelists are: Tony Baylis (Lawrence Livermore National Laboratory); Dr. Jesse Bemley (Joint Educational Facilities); William Burke (George Washington University); Jim Ganthier (Dell Technologies); and Dr. Valerie Taylor (Argonne National Laboratory)

Listen to the podcast here.

Watch the podcast video in the above frame or here.

A complete transcript follows.


Addison Snell: Welcome to a special episode of This Week in HPC with Intersect360 Research distributed by HPCwire. I’m Addison Snell with Intersect360 Research, and this week in HPC, we’re doing something new with our first-ever video panel episode of the podcast. Those of you listening to this podcast through our normal audio channels on iTunes and SoundCloud can find the complete video on the Intersect360 Research YouTube channel along with coverage on HPCwire.

We’ve adopted this video format today because of an important issue as we look at the topic of racism in America specifically as it pertains to high performance computing. For this panel I’m honored to be joined by five prominent African-Americans representing leadership roles from across the HPC community who are here to share their personal opinions and experiences, not necessarily those of their organizations.

Jim Ganthier is a senior vice president at Dell Technologies. As a seasoned executive with an engineering background and 14 patents, Jim has led HPC strategies and solutions for both Dell and HPE – the top two HPC system vendors by worldwide revenue. He was one of HPCwire’s People to Watch in 2016, and in 2019 Jim was named one of the top 10 blacks in technology by the Austin Black Business Journal. Jim, thanks for being with me on the panel today.

Jim Ganthier: My pleasure to be here, and I’m looking forward to our conversations along with the other panelists.

Snell: Thank you, Jim. William Burke is a senior HPC system engineer at George Washington University. His 15 years in HPC technical roles crosses industry, government and academia in multiple fields. His experience spans the NIH and industrial roles across pharmaceuticals, genomics and financial services. William has a master’s degree in information technology from Harvard University. William, I’m really glad you’re with us this morning.

William Burke: Happy to be here.

Snell: Dr. Valerie Taylor is the director of mathematics and computer science division at Argonne National Laboratory, where her research is in the field of performance analysis and modeling of parallel scientific applications. She’s also the executive director of CMDIT, the Center for Minorities and People with Disabilities in Information Technology. Valerie previously held CS leadership roles at Texas A&M and Northwestern University. She earned her PhD in electrical engineering and computer science from UC Berkeley. Valerie, thanks for being with me on this video panel.

Valerie Taylor: Thank you, Addison, I’m happy to be here and look forward to the discussion.

Snell: Tony Baylis is the director of the Office of Strategic Diversity and Inclusion program for Lawrence Livermore National Laboratory. Tony began his career at NCSA at the University of Illinois at Urbana-Champaign, where he earned his degree. In addition to his role at LLNL, Tony has served in student outreach roles for the SC conference series. He’s the diversity and inclusion chair for ACM SIGGRAPH and he’s a board member for the Empowher Institute. Tony, I’m thrilled that you’re with us, thanks.

Tony Baylis: Thank you, Addison, for inviting me to be on this panel. I’m very pleased to be here today. So, thank you.

Snell: Thank you. And finally, Dr. Jesse Bemley is a lecturer at Bowie State University and the director of Joint Educational Facilities introducing parallel and advanced computing techniques to inner-city high school and college students. As national education chairperson for Black Data Processing Associates, he helped to introduce the BDPA High School Computer Competition, which today has over 20 participating teams from chapters across the country. Under the BDPA the Dr. Jesse Bemley Scholarship Program facilitates the emergence of minority youth into the technology industry. Jesse, it’s great to see you again, thanks for being with me on this panel.

Jesse Bemley: It’s great to be here, Addison. You’ll find that I do a lot more listening than I do talking.

Snell: We’ll call on you when we’re ready, Jesse. All right. Thank you all for being here. I’m excited about talking about this very important issue and for our first topic today, I want to go right to the obvious, which is that blacks are dramatically underrepresented in our industry. As an HPC analyst, I do not have nor have I ever seen any official numbers on racial inclusion in HPC at a community level, but if I were asked my analyst estimate, I’d say that the HPC community is about 1% black within the U.S. and less if restricted to technical roles in science or engineering. And I’d like to just ask each of you to talk to me about your own experiences as a black person in the HPC community in the U.S. and in what ways you may have experienced racism in your career, whether it was overt or subtle or individual or systemic. Jim, I’d like to start with you. Why don’t you go ahead.

Ganthier: Well, first of all that was a pretty stunning statistic because if we look at the IT industry in general, blacks make up roughly eight percent of at least available on public websites, which obviously doesn’t match the anywhere between 18 to 22 percent of the general population.

So having said that: yes, not only very underrepresented but let’s just say that there are some hurdles that most of us have had to traverse. I’ll give you some, you know, perfect examples. I’m sure everybody in the panel has had the phrase or something along the lines of: “wow, that was really smart, or wow, you are really eloquent.” You know, that may not sound like much but if you step back and think about it it’s almost as if well, maybe you shouldn’t be or maybe you know, I’m surprised that you are. So that’s a version of a subtle scenario and by the way, you know, I know this is mostly a US topic of conversation, but I’ve I’ve walked into a room where I’ve had members of my staff – and because of the last name that I have which is you know, very European last name – people have literally brushed right by me and went to shake the hand of someone who doesn’t look like me. And they automatically assumed because of both position and last name who that individual was. So yes, I mean over my you know, 25-ish year career, I’ve never had anything really overt, but there’s just been a lot of subtle things that have occurred that make you kind of sit there and go “hmm.” Now, in most cases when that has occurred I had the type of personality where I’ve addressed it in a very polite, professional fashion, but not everybody does that.

Snell: You brought up a good point Jim that we’re talking about America here specifically. Now this doesn’t mean that racism – racism is of course a global problem. We’re talking about the U.S. today for two reasons: one is to narrow the focus because racism in Japan or China might look very different than what some looks like in the United States, and more to the point this is what’s topical in the news cycle. This is an acute issue in America right here, right now. William, I’d like to come to you next because you’ve got experience that spans a lot of different types of organizations as a senior HPC leader and when I was talking to you before you had an experience, I think that was quite similar to Jim’s.

Burke: Yeah, that is true. I can certainly relate to coming into a room and also because of my last name, you know, first and last name, I had this situation where I work with a guy for about a year, we never actually met, we had a virtual team, and this particular day we were to meet the team and he comes in – so we have a we have a nine o’clock meeting and he comes in 20 minutes late, he’s got the raincoat and it’s raining outside and he shakes off the umbrella, he sits down. He basically knows everyone else but we had never met, and so he looks around the room and he says, “hey, where’s William, is he late?” And everyone sort of turns and looks at me, and I’m senior to the project and so he realized quickly and it’s pretty subtle, you know, and he says, “oh William I didn’t see you” and so I’m just, you know I said, “oh I’ve been sitting here all along.” And so these are sort of the subtleties that we kind of have to face at some point in time and some people might have taken it in a negative way. I have a way of being able to, you know, address it in such a way that it’s looking at it from a perspective of you know, it could be an honest mistake, but these are things that you actually have to deal with. And then what was interesting is that we did kind of talk about it a little bit later and then he says to me, he’s like, “Mike,” he said, “my god, you know, I’ve been speaking with you a whole year and in my mind I pictured you as this white guy!” And so it was a really interesting topic that we had, and it was educational from the perspective of you know, I asked him, I said, “well, you know, what does a black guy sound like?” Right, so that’s what you know, one of many experiences.

Snell: I’m gonna come to Jesse Bemley next. This is something that you’ve addressed specifically with a lot of young people in your career, where do you see this both in your path and in the path of those you help?

Bemley: Let me first tell you about my incident. I was the person who was the lead person in database management systems at DEA back during the 80s. And we were subject to the Department of Justice, so we had a big meeting down at the Department of Justice. I took my team down there and we were supposed to have some information transfer and all the questions were directed to my team. And most of my team was white, so most of the questions were directed to the white boys. And they answered the questions, but every time they came to a point where they didn’t have an answer they had to defer to me and they did. And after we left, the white boys were not happy about that, but I told them, “don’t worry about that, we accomplished the mission, that’s foremost in my mind.”

Snell: This is an area where I’d like to think that science, engineering, HPC ought to be a meritocracy, right, where just the best ideas start floating to the top, but it sounds like you’re saying that’s not always true, Jesse.

Bemley: That’s not always true, not always true at all. And you know since I initially talked to you about this there are a lot of things that have been floating around in my head, and having been around for a long time – about 76 years – I’ve seen a lot of things happen. And I see that most of what the problem that we’re having has to do with all of the resources flowing in one direction and then when those people that are resource-rich need us there’s a little trickle down effect. Let’s just take computing in general. When I was in undergrad, a computer was a person who computes, and we were using those big calculators. And you know, we didn’t think about those big machines, that kind of thing. And so when the computer industry matured somewhat, the big white schools got them. In order for the blacks schools to have any kind of computational capability, what they did was they would put a leased line into a black school and a terminal on the end of that leased line. So there was no computation going on at the black schools. It was all done at the white schools.

Snell: I think many listeners you might hear white schools, black schools and think maybe we’re past that but you’re at Bowie State University, which is the oldest historically black college or university in the state of Maryland, and I think part of the conversation today has to be: what our industry is doing together with the HBCUs in this country. I’m going to move on to Tony and Valerie, both of whom are at major US supercomputing sites. Tony, where are you on this topic and where have you seen this in your career?

Baylis: I’ve seen it mostly in my conversations with students. I’ve been very I would say blessed just based on the fact of coming into an environment at the University of Illinois and the National Center for Supercomputing Applications, where I’ve had Dr. Larry Smarr, Dr. Mike Norman, Dr. Ed Sidell and James Bottom, Maxine Brown, Tom Defanti, and Dr. Valerie Taylor sitting here, be really sort of mentors to me in this space, and as a result I’ve been somewhat included. Although I didn’t necessarily feel like I would belong in some areas, especially at our – unfortunately at our Supercomputing Conference, even though the people there did work to make me feel that way, I was usually one of the few. Valerie and I were out there as one of the onlies. But what I’ve heard though from the students that are still today in 2020, they’re talking about this issue that they are not treated as if they belong in this sort of area and they feel – they see it from faculty, they see it from the system itself – the institution, the educational system – that they’re not basically, as Jesse had put it, given the opportunities. They don’t even know about HPC, the awareness isn’t really there in a lot of historically black colleges and universities and quite frankly Hispanic-serving institutions as well. So MSIs across the board unless there is a intentional effort at that particular university or they have a faculty member who they’ve recruited who has a research interest in that space and they do all the sort of lifting of you know, putting people in that space. But you know, the students themselves say that they face some of this racism that you talk about. I came into it late and like I said, I’ve been very very lucky to have wonderful stewards and mentors in the HPC arena for me.

Snell: Some of what you’re talking about is a subtle and maybe even unintentional problem where here’s me as a white man and some senior position in HPC and someone says to me, “senior HPC executive in some role,” and in my head, I haven’t met this person yet, but I’m picturing first of all a man, and second of all a white man, and and it’s just sort of situational. And that’s been really related to the stories that you’re all bringing to the table. Valerie, I didn’t mean to really put you all the way to the end, you’ve been very patient waiting.

Taylor: So I will say especially as a black woman, when the numbers are low, you know, the question that comes up has to do with the mental model and the assumptions that people have about you know, occupations, about careers and I would say especially, too, in HPC because for example with myself as a faculty member, I would go places with grad students, and one time I went someplace with a white female grad student and when the two of us walked in the room they went to her and said, you know, “how are you doing, Professor Taylor?” and she was like “no! No, you can’t say that!” It’s okay, it’s okay. You know even you know moving to Argonne, with the bank that I was working with, we never met but they knew the position and they knew the salary and so they said “I’ll help” you and they filled out the forms and filled out demographic information. So they filled out the demographic information as female, white. You know, so it has to do with the assumptions that people make and how they act on those assumptions instead of getting to know a person and giving people the benefit of the doubt. I think you know that becomes a major issue.

Snell: One last thing I definitely want to hit on before we move off of this topic is a phrase that came up in one of our preliminary conversations. William, I was talking to you and you were talking about affirmative action and said, “I’m a product of affirmative action.” Can you recap some of your experience in that regard?

Burke: Yeah, so when I said that I was saying that I’m a product of affirmative action and so is my mother and so is my father, and I think it definitely had a positive effect. However as time has gone on there’s been there’s been situations where I felt like I was part of a quota system, you know, where say they had, you know, they had to hire a hundred people and they’re looking and said, oh my god, you know, you have only three blacks here and we need four point five, so we need to go and find two more bodies. You know, I don’t want to be you know that person or have some be hired, you know, just just because of that. But I think that you know diversity means a lot of things and inclusivity and diversity embraces that unique contributions from members of the community and you know, we’ve got to, you know start to look at you know, what that means. It doesn’t mean having you know, many black faces in the room or even women faces in room or the multiplicity of you know, minority faces. The change has to come within the policies and the practices and to be quite frank the mindset. You know, we just kind of spoke about you know, some of those subtleties of what you kind of think about you know, when you think about an executive or when you think about that, you know that lead engineer, you know, you know what comes to mind is you know, it’s definitely usually, first choice is you know, like you said a male but not even just a male, a white male. So if people are thinking along those lines, you know, we have to you know, there has to be a way of being able to have people to open up and kind of look otherwise, you know, and so like for instance, there was a push to you know to embrace women in a lot of the leadership roles, there was a conscious effort to go after and push for women in leadership roles. So I would say the same thing for blacks as a minority to be able to embrace…we do have our opinions. And I think diversity is about the willingness to accept those types of diverging opinions and so that allows us to have a better realization of what our environment is and not that it’s just you know, sort of this one standard and one potential direction to go that we embrace that diversity.

Snell: And Jim, just very quickly wrapping up this topic, you’re coming from the corporate side and you were talking to me about the importance of that diversity of thought and it goes beyond the problem William’s describing where even once he’s in the organization he can feel like other people are looking at him thinking “oh you’re the token we had to hire” and the benefit of having diversity of thought in an organization and how you make better decisions.

Ganthier: Yeah, so just to give a couple of facts and data points: I mean if we look at where the world is headed, if we look at the number of folks that we’re going to need in STEM, if we don’t start hiring more diverse talent, we’re going to be short by about 1.6 to 1.8 million depending on how you want to do the math. The second thing that we found – when I say we, you know, obviously this is now top of mind for many other execs at a lot of other high-tech companies – diversity of thought, diversity of people, actually is good for business. I mean, I won’t go into specific examples, but having folks who can look at things in a different way, having folks who have had a different background, not only makes the products, offers, solutions, supercomputers that much more innovative, but in some cases that’s how the true breakthrough happens.

So whether you want to look at it as an industry problem, whether you want to look at it as an impending scenario that we categorically have to go solve or if we want to just look at it as quote unquote good business in terms of innovation and driving new thought processes, new approaches and new ways to do things – all of those are items that makes this not just a very topical conversation but a very important conversation right now.

One thing I do want to highlight a little bit of is – I was resonating with a lot of what William had mentioned. Sometimes it’s not hiring because of the person but just getting that system kick-started. I know an executive at one particular high tech firm and you know, very similar to what Tony mentioned, he said look, you know, you need to come up with four people that you want to mentor. And he had everybody write those names down on a list and then he said, okay out of the four people, how many are female and how many are underrepresented. If you don’t have at least one or two on the list, you’re obviously not doing something well as a leader. So it’s not you hire because of, but at least start to look at the consideration set, at least make it part of the initial process, because like a flywheel until you do that upfront, the wheel never starts to spin and progress doesn’t get made.

Snell: We often talk about supercomputing in general as solving the world’s toughest problems, and I want to raise an interesting question here, because when we talk about racism, to me, this is getting into “world’s toughest problems.” Now, the other major news story this year has been the global pandemic and I noticed that we’re all very fast to get into the conversation about what can supercomputing due to battle COVID-19, modeling its spread, looking for viruses, all the molecular modeling.

Now, here’s this other major news story about racism and the state of race relations in America, which strikes me as a world’s toughest problem. Can HPC in of itself, maybe with the broader aspects of data science or big data or machine learning and AI, can this scope of high performance computing potentially help address systemic racism in modeling for it, in controlling for it. I don’t know, should someone ask? I’d like to start with the major supercomputing centers here. Valerie, do you have a perspective on this?

Taylor: Yes, I think in terms of as you mentioned with HPC and solving some of the world’s grand challenge problems, I think if we look at systemic racism as a grand challenge problem to be solved, and that is along multiple dimensions. So one, you know, I think one area that’s on everyone’s minds right now has to do with policing and you know injustices and racial injustice. And so, I think if we start to have for example look at it as a data problem in terms of you know, looking at in terms of police brutality, recognize it as data and make that data available, make that data in terms of looking at trends, you can look at trends around different areas, different cities, regions as to what’s going on.

You can also have the data available such that when you have issues arise, it’s not where, for example, someone can go from one region to another region because you have data available that’s been analyzed, you have some anomalies. We’re very good with different systems of identifying extreme events,, identifying trends identifying those anomalies. We can look at that especially in terms of police reform.

I think another area that has to do with HPC – and this I think has to do more in terms of engagement – and that is HPC is a very exciting field and I think as we talk among ourselves and we talk about the excitement and what we’re doing and the problems that we’re working on and how we’re solving some of society’s major problems, whether that’s you know, COVID-19 and the pandemic, whether we’re looking at energy, we’re looking at solving let’s say cancer treatment, but often times we don’t convey that excitement to K-12, we keep it among ourselves and talk and we pray, but we need to go out and figure out ways to convey that to K-12, you know, especially looking at high school, looking at undergrads to have excitement also about the field of HPC. So I think it’s circular, and that is HPC helping to solve some of the problems having to do with racial injustices, but also looking at HPC as more engagement too in STEM and more engagement in the field. And for that I really want to commend Jesse because I know for many many years, that’s Jesse’s passion, is with the K-12 and he’s been doing that for many years and if you look at where the students go in Jesse’s program, it’s phenomenal in terms of going into STEM. So I commend you Jesse for the many years.

Snell: Well, let’s go to Jesse next then since you were just calling him up there, we’re talking about trying to solve this. Jesse, what do you think the role is of HPC in trying to solve the problem?

Bemley: Okay the role of HPC in K-12 Is just to be what I would consider a resource to get the young people familiar with and what I was doing was not so much as teaching them, you know, how to be researchers and all that type of thing – what I was doing was just introducing HPC to them, so that they would have [knowledge of it]. Let me just give you an instance of – when they were having the cluster competition, where a team would take a node of a cluster, put it in the rack, configure it, hook it up, and make it be a part of that cluster – well we had a high school student who was an 11th grader and I was trying to reach him so that I can figure out which conference that was that was an SC [Supercomputing] conference and this competition was open to high school all the way up through professionals and he was leading everybody in terms of getting all this done but he lost his lead because he was not familiar with the IO interconnects, those interfaces. I mean, we had nothing like that at Joint Educational Facilities in DC. I mean, we just barely had RJ45 interfaces. And so by the time he figured out how to get it interfaced he had lost his lead. Other than that he would have won that competition – a high school student, competing initially with everybody.

Snell: And you are doing excellent work introducing parallel and advanced computing at the high school and K-12 level. I want to come back to this idea of HPC addressing systemic racism. Tony, you’ve been at a couple of major supercomputing sites, Valerie seemed for it, if we just take the notion of we’ve got data someone ought to be able to do an application here. Do you agree?

Baylis: A hundred percent. I think the areas where HPC can actually lend itself to assisting with this systemic and institutional problem is solving that on the data front we can do visualizations that actually should talk about the modeling as well on that. I mean, if you go back to just look at the historical nature of things, the redlining that happened, you know way back when, and we do some analysis on that redlining and basically show that that was a causality of what we basically are experiencing today and why that is. It’s true – I mean, there’s research papers on this, there’s research papers studying the police brutality that that Valerie talked about, one was just in August 2019 that was produced that talked about it from ethnicity, gender, you know, and age as well – all of those things are things that could be studied more, and using high performance computing systems to do that work from a data science perspective would probably influence us to see inside that a little bit more as to, you know, okay, what can we do? At least we’ve got the data and information, what are some of the things that we can move forward with with ideas to try to help us sort of mitigate and hopefully erase what we’re seeing today.

Snell: Where does that responsibility lie? Is that a DOE supercomputing lab problem? Is it NSF? Is it an academic kind of problem? Who should run it?

Baylis: I think everybody should be coming to the table to solve this problem because this is not one institution versus another. Everybody needs to be a part of the solution and just what we’re saying, what you’ve seen in the Black Lives Matter movement in general, you know, basically everybody is saying: enough is enough. We all have to solve this, we all have to be a part of the solution moving forward, and you know, unless you have that, you know, you’re going to keep having these sort of little pitfalls and these hiccups along the road that you basically are going. It’s just not going to – it’s going to keep going. And I think if we don’t get a good majority at the very least at the table, you know. So DOE, industry, all of us, academia, should be involved in solving this. And it is a grand challenge quite frankly because it is global as well.

Snell: William, I’m very interested quickly in your perspective on this, because you’ve done those different roles – managing HPC clusters of various sizes in industry, at the NIH, at GW now. Where do you see this fit? Is this an application that you could see fitting into the workloads that you’ve encountered?

Burke: Yeah. From an application standpoint, I mean, there’s certainly use of HPC or and AI tools to combat racism, but one of the things we have to look at, we have to look at you know, it doesn’t work inside of a vacuum. And so what I mean by that – I’ll just kind of talk about something that’s really relatively low-tech. You know, there’s this YouTube video and it’s called “racist soap dispenser” and what they discovered was that in the programming of the algorithm, the soap dispensers initially didn’t seem to work for black people. So the way the IR worked was that it would send out, you know, it would send out this light, and if you had lighter skin or white skin it would reflect back, but if you had darker skin it would absorb. And so it’s amazing to me that with something that’s low-tech as that no one actually thought about you know, the range of complexions. So we have to be careful about how we apply and who is creating these algorithms and that so that we have the you know, when it comes to HPC applications, they’re much more complex. And then there was also another case where the Xbox Kinect, when it first came out it didn’t recognize black people, so it was pretty interesting that yeah, we can utilize technology to combat racism but we have to pay attention to those that are actually programming and creating the applications because they can have their biases too.

Snell: Jim, this really gets back to something you were talking about with regard to the diversity of experience in programming. We’re talking about setting up AI algorithms, and if it’s all white people or all men who are writing the algorithms, is that setting us up for trouble?

Ganthier: Well actually it kind of does. I mean, I joke sometimes that technology is like fire: it can be used for good, or it could be used for bad. So the net is that if we want to use a good example, and this kind of resonates with everything the other panelists have talked about, we’re doing some early work with some schools. I mean in order to, you know, my humble belief, solve systemic racism, you’ve got to look at it as a system. So we’re looking at which kids are going into particular STEM programs, we’re analyzing why are they falling out, we’re doing some predictive and proactive things in terms of, hey look are their grades dropping, is their attendance going down, in some cases have they registered family issues? All of that will help fix it on the front end. But similar to what William mentioned, you know, at the end of the day HPC is a computational type device. The magic is the algorithms. The algorithms are programmed by human beings. Those human beings don’t look like a lot of us on the panel. So this implicit bias is actually being placed in the algorithms.

We had the honor of having Dr. Joy Buolamwini from MIT come have a conversation with us about about six or seven months ago. And her entire conversation, very similar to what William highlighted, was the implicit bias built into facial recognition, so it either doesn’t recognize people of color – I mean brown or yellow – or in some cases it it falsely recognizes. So the net is that can HPC help and some of these grand problems? Categorically, but we just need to make sure that as we’re applying it, we’re applying it properly, we’re applying it to everybody and we’re making sure that we’re not building in some of these things that we as human beings are programming into the systems because at scale it would not be good for anyone.

Snell: A big topic that we’ve been addressing here is that the black Americans are dramatically underrepresented in the HPC community, so what I want to really get around to is our final big topic: what does it take to bring about change in this space? And I’ll start with Jesse here because Joint Educational Facilities have been trying to do exactly this work in terms of bringing more minorities up in technology roles. Jesse what’s your role there about what additional resources are needed, and what would progress look like here in the medium to long term if it’s working?

Bemley: Well what I’m seeing is us making massive resources available to people down in the K-12 level, but we have to train these people in order to use these resources. When Sputnik went up it scared the bejesus out of us, and Kennedy vowed that we were going to catch up and pass them, and so what happened? That was massive amounts of learning put into the STEM discipline. I’m the benefit of some of that money as a mathematician back in those days and so I thought that that was going to happen when we had the AI expert system surge during the 90s but it didn’t. I think that that’s what we need to do, we need to put massive amounts of resources out there – money, facilities, that type of thing. And then start identifying people to man those – well before we put the resources out there, identify people to use those resources and man those facilities and start at the elementary school level and interact with folks at all those segments of the pipeline.

Snell: Tony you’re wearing an SC conference shirt. You know, the SC conference is the biggest event in our industry every year in the U.S. Now we don’t know, with the COVID-19 pandemic, we don’t know if we’re gonna be in person in Atlanta or not but you know, my concern is: protests happen, people get tired, the news cycle moves on. By the time we get to November what should we be talking – should this be a topic at SC? What more needs to be done? How do we carry this conversation forward, and include in that: what does progress look like in this industry, what should people be doing?

Baylis: I think what we can do at SC is to – well first, we need to invite those that are in that particular community in Atlanta. Because there is a breadth of people and schools in Atlanta that we could tap into and invite them to the table as much as possible. The other thing is that I don’t think this is going to be done this time around. I think this conversation is going to continue, and I also believe that we should have some – whether it be roundtables or panel discussions at supercomputing around this topic that involves industry, academic institutions, as well as government institutions to have conversations about actionable steps that we’re going to take together to basically run this sort of pathway to understand, “Here’s what you’re doing, here’s what you’re doing, can we bring these things together,” because oftentimes, and you do need to do this, you have everyone basically doing their things but they overlap somewhat. Can we tie some of that work together perhaps? How can we make this a system to basically help us mitigate the problems of institutional racism? Can we do that? Because you know Dell, Intel, all of those folks of the world are basically putting efforts and money toward these sort of areas, but can we tie this together somehow can we have conversations about that to speak to what Jesse is talking about to help K-12, to help more college students, to partner with academia even more in this space. And there’s folks that have been doing it. Now, I would even say we should be inviting people from nonprofits even to that arena as well, because they can help us basically bring that awareness about for young people that we need to reach because there are young people out there who just don’t know, and they’re curious. They’re using a computer every day with their cell phone. So with that said they just want to know how it ticks – some of them do, some of them just want to play – and that’s okay, you know, that’s what that’s what being a child is all about – but that also sort of creates that curiosity and that enthusiasm that you know, as soon as you bring it to them that they can create apps and things of that nature, they’ll have a wonderful wonderful time of just exploring. Lastly, the only example I’ll give you is that you know, we had a program where we basically reached out to historically black colleges from our laboratory and I have a young man from rural South Carolina, he built an 11-node cluster out of Raspberry Pis himself, he learned it and then he basically created an actual we call it the Lego Pi, he created Lego Pi so we can teach young people, because he said “if I had this…” and he did a pilot with kids from his middle school, you know, where he went in rural South Carolina. And as a result of that we’re basically rolling out those Raspberry Pis, you know, and having the HPC [professionals] use those to teach young people so to get them excited about it. So if you do certain things like that I should say that will be helpful, and having role models, because he was a role model for them because he came from where they came from. Those things are important. Sorry I went on a little bit longer but I get very passionate.

Snell: Well he sounds like someone with bright job prospects in this space. Valerie, I’ll come to you – what’s your call to action here, how would you like to see this conversation evolve, and specific action items?

Taylor: So I think a couple of things just to leverage to with Tony and Jesse and that is you know, we talk about intentionality and I think that is so important. And I think recognizing with this momentum that it’s a long-term, you know, effort that we have to put in place.

And so being intentional about a long-term effort there are actionable items that we can do in the recruiting area, but I want to bring up the area about environments. And that is once you get people in the door you also want to have people stay at places or stay engaged in HPC. And so I just want to spend a few minutes talking about environments because I think that is so important. We talk about training, but I think one aspect I want to bring up is that having to go back to assumptions that people make, and mental models the people have in terms of – the assumptions you make from the first time you see a person. And one time I attended this one training that had to do with privilege. And oftentimes people talk about white privilege and that’s along the racial dimension. Well, this particular training, she presented 30 different dimensions having to do with where you went to school, where you live, you know, all of these aspects such that at the end of the day, everybody has privilege along some particular dimensions.

So if you start to look at people and individuals and blacks to say let me get to know you and get to know your background, not make assumptions, then when you start to look at the environment you start to move in the direction of an inclusive environment because people have a lot to contribute, and having people at the table contribute and being heard is so important and that has to do – I want to go back to also with William and Jim talked about, when you start to talk about HPC applications, having people at the table looking at also the data that you have as well as the algorithms. So I just want to spend some time just to emphasize the environment that people are working in becomes so important, and looking at it as a long-term issue that we need to continue this dialogue annually, you know, quarterly, looking at accountability how we’re doing and having metrics and mechanisms in place to track it.

Snell: William, you’ve been at a couple of different types of HPC organizations in a senior technical role. From that viewpoint, what should change look like here? What’s an individual viewer listener to this podcast, how should they behaving differently?

Burke: Great question. I’ll kind of talk about when I first got into HPC back in 2002, about 18 years ago. I was at Oak Ridge and it was a program that I got involved with, an internship, and one of the things that I liked about it was that it was very nurturing and and the mentors there, you know, were heavily invested. And I noticed that across the board not just within not just within our program but with many programs. So I think mentorship is very, very important – having people vested, I think that organizations, even you know, for-profit companies who have those types of programs in place, and as Valerie said, you know, being able to, you know, look at people for who they are, because there has to be a willingness to acknowledge or validate, you know people, and even if they are a little different than what you normally see it has to be a nurturing environment. And so I remember first going on Oak Ridge National Laboratories campus and just being overwhelmed. Now I had some engineering background, you know, I start off in the computer engineering industry so it wasn’t all new to me, but it was just like fantastic because it was like, hey, this is where at that time, the world’s fastest supercomputer was. And then it’s starting to meet people who you know who were really really smart who were willing to share, so I think more of that needs to be seen, and not just at the governmental labs but just across the board, you know, whether it be for-profit companies or non-profit companies, I think that we need to see industry and academia work hand in hand, bringing more and more of the STEM programs not just you know, not just looking specifically for underrepresented groups but to really have an investment because with diversity, you know, I think with corporations if you put it in terms of economics in terms of financials and you show the benefits, I think it will move the needle towards that social justice. That’s what I’d like to see moving forward.

Snell: Jesse I saw you coming off mute, is there something you wanted to inject there?

Bemley: No.

Snell: I’m going to go to Jim and wrap up then. William was just talking about industry and you’re you’re on the industry side, you’re not necessarily speaking for your organizations but to you what does this look like going forward? How should this change? What are the action items for our industry?

Ganthier: My fellow panelists have said it really well. If we want to call it, you know, pick your favorite – institutional or in some cases systemic racism – we’re gonna have to look at it as a system. We’re gonna have to look at it as a life cycle. Because even things that you know we grew up with may or may not be applicable. Love the idea of you know Jesse talking about how to get this into the K-12, but here’s the fun stuff: a lot of folks in tech and a lot of companies in tech may no longer require a four-year degree. You know, some of our partners slash co-opetition are starting to look at folks who not only have two-year degrees, some of them they’re hiring directly out of high school and some of them collectively were setting up an apprentice. So we’ve got to start them young. We got to get them into a STEM program. We got to get them excited about here are some of the world’s biggest problems that you can potentially not only be a part of but help solve. That once they do get into a STEM-like program, we help them. We help them with things such as – and the digital divide is very real, we give them access to the right types of tools and the right types of training. We may want to look at something, and again, you know thinking about the larger industry – a lot of folks in industry are now working with historically black colleges, a lot of folks are also working with what we call MSI’s, which is minority serving institutions. I happen to have had a conversation with one of the largest graduator of underrepresented STEM on the east coast yesterday, but It’s not just getting them in, excited, trained and on board – it’s once they actually get into industry and you know, Dr. Taylor was highlighting this very well, how do you give them a nurturing environment where they not only can do their best work, they can bring their authentic selves, they can be the innovative types of folks we’re looking for, they can collaborate across the companies – and how do we build ecosystems where they stay?

You know, I’m not gonna go through any particular company’s data, but you can look at it very well. If you look at underrepresentation, somewhere between years three to eight they typically fall out of companies that they don’t feel comfortable with. So how do you build a sponsorship mentorship nurturing component, then ultimately how do you help move them into senior leadership? Because again this is a longer-term play, we don’t want to be looking back five years from now and saying well the number hasn’t budged or better yet, they’re really aren’t a lot of folks not just at the bottom of the pyramid but all the way through the pyramid. I mean, folks want to be able to look up a corporate ladder and see folks that look like them. They also want to be able to look across governments and do the same. So we’re gonna have to look at this not just long term, we’re gonna have to look at this systemically, we’re gonna have to go after every single one of those points where we either have a fallout point or where we end up losing individuals only because we didn’t make them feel comfortable calm and to some extent being able to bring their authentic and best selves to work every day because that’s where the real innovation happens.

Snell: I’d like to thank my panelists Jim Ganthier from Dell, William Burke from George Washington, Valerie Taylor from Argonne, Tony Baylis from Lawrence Livermore and Jesse Bemley from Bowie State. Thank you for taking the time to allow me to hear your perspectives on this crucial topic. I hope it won’t be too long before we can all see each other in person again at an HPC community event. I’m Addison Snell with Intersect360 Research and thanks to you, our listeners, and viewers for tuning in.

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industry updates delivered to you every week!

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion XL — were added to the benchmark suite as MLPerf continues Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing power it brings to artificial intelligence.  Nvidia's DGX Read more…

Call for Participation in Workshop on Potential NSF CISE Quantum Initiative

March 26, 2024

Editor’s Note: Next month there will be a workshop to discuss what a quantum initiative led by NSF’s Computer, Information Science and Engineering (CISE) directorate could entail. The details are posted below in a Ca Read more…

Waseda U. Researchers Reports New Quantum Algorithm for Speeding Optimization

March 25, 2024

Optimization problems cover a wide range of applications and are often cited as good candidates for quantum computing. However, the execution time for constrained combinatorial optimization applications on quantum device Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at the network layer threatens to make bigger and brawnier pro Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HBM3E memory as well as the the ability to train 1 trillion pa Read more…

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing po Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HB Read more…

Nvidia Looks to Accelerate GenAI Adoption with NIM

March 19, 2024

Today at the GPU Technology Conference, Nvidia launched a new offering aimed at helping customers quickly deploy their generative AI applications in a secure, s Read more…

The Generative AI Future Is Now, Nvidia’s Huang Says

March 19, 2024

We are in the early days of a transformative shift in how business gets done thanks to the advent of generative AI, according to Nvidia CEO and cofounder Jensen Read more…

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, codenamed Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from Read more…

Nvidia Showcases Quantum Cloud, Expanding Quantum Portfolio at GTC24

March 18, 2024

Nvidia’s barrage of quantum news at GTC24 this week includes new products, signature collaborations, and a new Nvidia Quantum Cloud for quantum developers. Wh Read more…

Alibaba Shuts Down its Quantum Computing Effort

November 30, 2023

In case you missed it, China’s e-commerce giant Alibaba has shut down its quantum computing research effort. It’s not entirely clear what drove the change. Read more…

Nvidia H100: Are 550,000 GPUs Enough for This Year?

August 17, 2023

The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…

Shutterstock 1285747942

AMD’s Horsepower-packed MI300X GPU Beats Nvidia’s Upcoming H200

December 7, 2023

AMD and Nvidia are locked in an AI performance battle – much like the gaming GPU performance clash the companies have waged for decades. AMD has claimed it Read more…

DoD Takes a Long View of Quantum Computing

December 19, 2023

Given the large sums tied to expensive weapon systems – think $100-million-plus per F-35 fighter – it’s easy to forget the U.S. Department of Defense is a Read more…

Synopsys Eats Ansys: Does HPC Get Indigestion?

February 8, 2024

Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…

Choosing the Right GPU for LLM Inference and Training

December 11, 2023

Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…

Intel’s Server and PC Chip Development Will Blur After 2025

January 15, 2024

Intel's dealing with much more than chip rivals breathing down its neck; it is simultaneously integrating a bevy of new technologies such as chiplets, artificia Read more…

Baidu Exits Quantum, Closely Following Alibaba’s Earlier Move

January 5, 2024

Reuters reported this week that Baidu, China’s giant e-commerce and services provider, is exiting the quantum computing development arena. Reuters reported � Read more…

Leading Solution Providers

Contributors

Comparing NVIDIA A100 and NVIDIA L40S: Which GPU is Ideal for AI and Graphics-Intensive Workloads?

October 30, 2023

With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…

Shutterstock 1179408610

Google Addresses the Mysteries of Its Hypercomputer 

December 28, 2023

When Google launched its Hypercomputer earlier this month (December 2023), the first reaction was, "Say what?" It turns out that the Hypercomputer is Google's t Read more…

AMD MI3000A

How AMD May Get Across the CUDA Moat

October 5, 2023

When discussing GenAI, the term "GPU" almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word "GPU" is assumed to mean "Nvidia" products. (As an aside, the popular Nvidia hardware used in GenAI are not technically... Read more…

Shutterstock 1606064203

Meta’s Zuckerberg Puts Its AI Future in the Hands of 600,000 GPUs

January 25, 2024

In under two minutes, Meta's CEO, Mark Zuckerberg, laid out the company's AI plans, which included a plan to build an artificial intelligence system with the eq Read more…

Google Introduces ‘Hypercomputer’ to Its AI Infrastructure

December 11, 2023

Google ran out of monikers to describe its new AI system released on December 7. Supercomputer perhaps wasn't an apt description, so it settled on Hypercomputer Read more…

China Is All In on a RISC-V Future

January 8, 2024

The state of RISC-V in China was discussed in a recent report released by the Jamestown Foundation, a Washington, D.C.-based think tank. The report, entitled "E Read more…

Intel Won’t Have a Xeon Max Chip with New Emerald Rapids CPU

December 14, 2023

As expected, Intel officially announced its 5th generation Xeon server chips codenamed Emerald Rapids at an event in New York City, where the focus was really o Read more…

IBM Quantum Summit: Two New QPUs, Upgraded Qiskit, 10-year Roadmap and More

December 4, 2023

IBM kicks off its annual Quantum Summit today and will announce a broad range of advances including its much-anticipated 1121-qubit Condor QPU, a smaller 133-qu Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire