The Woman Behind Intel’s Leading Architectures

Rani Borkar has led engineering teams from Pentium 4 to Intel Core and now SoCs.

Rani Borkar has been with Intel since 1988 and currently leads key engineering teams that design and develop Intel’s Core, Atom, Xeon and Itanium processors. As vice president and general manager of the Intel Architecture Group, she’s also been tasked with a big challenge: leading Intel’s efforts to design a new crop of SoCs. These integrated system-on-chip designs incorporate core functions into the silicon for PCs, tablets and servers as well handheld devices where Intel is considered lagging. Recently, Borkar discussed the shift in focus from designing traditional microprocessors to SoCs and how Intel stacks up to the competition.

Rani Borkar Intel Architecture Group

"Winning in SoCs is not just about changing how we build chips. Intel is in the midst of a huge and critical transformation. No one organization can do it alone, and many pieces have to come together to make us a successful SoC company," said Rani Borkar, vice president and general manager, Intel Architecture Group.

One of her first major projects she worked on in her Intel career was the P6 CPU that first came to market with the Pentium Pro in 1995. Since that time she has been involved with several major architectural developments at Intel, including Intel’s Pentium 4 and leading the engineering team behind Intel’s 4th generation Intel Core processor, which is expected to be launched soon. In this conversation, Borkar speaks about Intel’s SoC efforts as well as an issue she believes strongly in: bringing more women into science and engineering careers.

What’s required for Intel to compete in SoCs?

There are three guiding principles for SoC development. One, you need leadership features, which are essential to deliver breakthrough experiences. The strides we’ve made in our development methodology have made us more nimble, but we’re in an extremely dynamic environment. We have to take some risks. Two, cadence is important. You have to go very, very fast, and you have to establish a cadence so that the market knows what’s coming — they are expecting it. The third thing is cost. And, which is where I truly believe we can leverage our IDM [integrated device manufacturer] advantage against our competition.

At the beginning of 2012 the company had zero percent market share in phones. Today, there are Intel-powered phones shipping in more than 20 countries. How does Intel get from where it is now to being a major player in mobile?

Until January of last year we didn’t have a single phone in the market. Now we have a dozen designs shipping with Atom SoCs. I carry one of them [Borkar holds up her Orange smartphone]. We didn’t have a tablet SoC, now we have “Clover Trail” [the codename for the dual-core Intel 32nm Atom SoC]. Will we win in 2013? We are making progress and building momentum.

I was here back when we were nowhere on servers and look where we are today on servers. The difference with SoCs is that the pace is changing in this market and some of the parameters have changed. It’s a multi-OS, fragmented environment, and there are a lot of players, but Intel has the software expertise and assets to deliver Windows and Android products.

Rani Borkar VP and GM of Intel Architecture Group

"The strides we’ve made in our [SoC] development methodology have made us more nimble, but we’re in an extremely dynamic environment. We have to take some risks," said Rani Borkar, vice president and general manager, Intel Architecture Group.

If you look at [Intel] from outside, who has the assets we do? Forbes called Intel “the biggest software company you’ve never heard of.” We have the latest process technology in-house and the manufacturing capacity to scale. There are a lot of fabless companies, but we have the luxury of working directly with our manufacturing partners. When I say that I believe we will get there, I’m not just saying it as a philosophical belief; I’m saying it because I look at what we have.

How big is the transition from a CPU focus to a SoC focus?

Winning in SoCs is not just about changing how we build chips. Intel is in the midst of a huge and critical transformation. No one organization can do it alone, and many pieces have to come together to make us a successful SoC company. The strength of the collaborative spirit within Intel is amazing from software to platform to silicon and we have to work well across the organization to deliver the end product.

In the past, we were the CPU supplier. We tuned our process technology to our products. Our OEMs and ODMs had the system expertise, and they put the software stack and hardware together.

That’s changed. Now, to deliver the best end-user experience, we need to significantly invest in software and software-hardware co-design. On top of that, we have to deal with the challenges of a rapidly changing and fragmented OS ecosystem.

When you start to think about, “OK, what does the chip look like?” It’s not sufficient to lead in traditional areas like CPU and graphics, we need to be the best in emerging capabilities like display, media, imaging, and so on. As computing goes mobile, these are the capabilities end users really care about. We’ve invested a great deal in those capabilities and we’ve led in those areas for years. So it’s a shift to an outside-in view versus inside-out.

How much more complexity is there with SoCs?

There have always been interdependencies, but with SoCs it’s growing exponentially because there are parts that you’re not building yourself. There are things that we get externally. There are things we get from other groups within Intel.

Software is a key thing — it doesn’t matter if my silicon is ready, I need the software to be ready. We are co-designing and co-developing so that we find issues early because every time you spin new silicon it interrupts your process and that has the potential to slow the cadence and increase time to market.

How far is Intel along the path to being a major player in SoCs?

To build a house first you have to lay a foundation. I can’t start thinking about chandeliers and faucets and appliances if I haven’t poured the cement. So we have poured the cement — we have identified areas of improvement and made focused investments that have started to pay off. We’ve invested in FFRDs, we have a strong IP portfolio and we’ve increased our investments in software to support hardware-software co-design.

We are not done completely but we continue to put that tension in the system. So the foundation is laid. The machine is cranking. This year we’ve introduced a 32nm low-power Atom SoC [codenamed "Lexington"] for value smartphones and a 32nm dual-core Atom SoC [codenamed "Clover Trail+"] platform for smartphones and Android tablets. This year we’ll also be introducing a 22nm quad-core Atom SoC [codenamed "Bay Trail"] and the 4th Generation Core 22nm SoC [codenamed "Haswell"].

Rani Borkar in Intel Archiecture Group engineering team meeting

"There have always been interdependencies, but with SoCs it's growing exponentially because there are parts that you’re not building yourself," said Rani Borkar, vice president and general manager, Intel Architecture Group.

As someone who has worked in the tech industry for nearly 25 years, what do you think is needed now to bring more women into technical fields?

It has to start as early as the middle school level. By the time women get to high school, they’ve made up their mind on a lot of things. Maybe they go to college and change, but not that dramatically from non-engineering to engineering or even non-science to science. It has to start in middle school at least if not earlier.

If I reflect upon myself, I can still remember the encouragement from my grade 8 math and science teacher. It was a long, long time ago when I was in 8th grade, but yet I can remember their faces, I can remember the support.

Each one of us who is in the field has to make sure that we are sharing our experiences, how we handle challenges and we must be honest. Do I want to give up my child’s [musical] performance today and go to some big meeting? No. I have to prioritize that because I’m not just an exec at Intel, I’m also a wife and a mother. It applies to men, too. They have to make choices too. I don’t see that as a gender one. We talk a lot about it as women because we think this is something we can’t overcome, but all you have to do is basically say what your priorities are and how you’re going to address them. There’s a whole, big community out there who’s willing to help you if you let them.

The Woman Behind Intel’s Leading Architectures

Rani Borkar has led engineering teams from Pentium 4 to Intel Core and now SoCs.

Rani Borkar has been with Intel since 1988 and currently leads key engineering teams that design and develop Intel’s Core, Atom, Xeon and Itanium processors. As vice president and general manager of the Intel Architecture Group, she’s also been tasked with a big challenge: leading Intel’s efforts to design a new crop of SoCs. These integrated system-on-chip designs incorporate core functions into the silicon for PCs, tablets and servers as well handheld devices where Intel is considered lagging. Recently, Borkar discussed the shift in focus from designing traditional microprocessors to SoCs and how Intel stacks up to the competition.

Rani Borkar Intel Architecture Group

"Winning in SoCs is not just about changing how we build chips. Intel is in the midst of a huge and critical transformation. No one organization can do it alone, and many pieces have to come together to make us a successful SoC company," said Rani Borkar, vice president and general manager, Intel Architecture Group.

One of her first major projects she worked on in her Intel career was the P6 CPU that first came to market with the Pentium Pro in 1995. Since that time she has been involved with several major architectural developments at Intel, including Intel’s Pentium 4 and leading the engineering team behind Intel’s 4th generation Intel Core processor, which is expected to be launched soon. In this conversation, Borkar speaks about Intel’s SoC efforts as well as an issue she believes strongly in: bringing more women into science and engineering careers.

What’s required for Intel to compete in SoCs?

There are three guiding principles for SoC development. One, you need leadership features, which are essential to deliver breakthrough experiences. The strides we’ve made in our development methodology have made us more nimble, but we’re in an extremely dynamic environment. We have to take some risks. Two, cadence is important. You have to go very, very fast, and you have to establish a cadence so that the market knows what’s coming — they are expecting it. The third thing is cost. And, which is where I truly believe we can leverage our IDM [integrated device manufacturer] advantage against our competition.

At the beginning of 2012 the company had zero percent market share in phones. Today, there are Intel-powered phones shipping in more than 20 countries. How does Intel get from where it is now to being a major player in mobile?

Until January of last year we didn’t have a single phone in the market. Now we have a dozen designs shipping with Atom SoCs. I carry one of them [Borkar holds up her Orange smartphone]. We didn’t have a tablet SoC, now we have “Clover Trail” [the codename for the dual-core Intel 32nm Atom SoC]. Will we win in 2013? We are making progress and building momentum.

I was here back when we were nowhere on servers and look where we are today on servers. The difference with SoCs is that the pace is changing in this market and some of the parameters have changed. It’s a multi-OS, fragmented environment, and there are a lot of players, but Intel has the software expertise and assets to deliver Windows and Android products.

Rani Borkar VP and GM of Intel Architecture Group

"The strides we’ve made in our [SoC] development methodology have made us more nimble, but we’re in an extremely dynamic environment. We have to take some risks," said Rani Borkar, vice president and general manager, Intel Architecture Group.

If you look at [Intel] from outside, who has the assets we do? Forbes called Intel “the biggest software company you’ve never heard of.” We have the latest process technology in-house and the manufacturing capacity to scale. There are a lot of fabless companies, but we have the luxury of working directly with our manufacturing partners. When I say that I believe we will get there, I’m not just saying it as a philosophical belief; I’m saying it because I look at what we have.

How big is the transition from a CPU focus to a SoC focus?

Winning in SoCs is not just about changing how we build chips. Intel is in the midst of a huge and critical transformation. No one organization can do it alone, and many pieces have to come together to make us a successful SoC company. The strength of the collaborative spirit within Intel is amazing from software to platform to silicon and we have to work well across the organization to deliver the end product.

In the past, we were the CPU supplier. We tuned our process technology to our products. Our OEMs and ODMs had the system expertise, and they put the software stack and hardware together.

That’s changed. Now, to deliver the best end-user experience, we need to significantly invest in software and software-hardware co-design. On top of that, we have to deal with the challenges of a rapidly changing and fragmented OS ecosystem.

When you start to think about, “OK, what does the chip look like?” It’s not sufficient to lead in traditional areas like CPU and graphics, we need to be the best in emerging capabilities like display, media, imaging, and so on. As computing goes mobile, these are the capabilities end users really care about. We’ve invested a great deal in those capabilities and we’ve led in those areas for years. So it’s a shift to an outside-in view versus inside-out.

How much more complexity is there with SoCs?

There have always been interdependencies, but with SoCs it’s growing exponentially because there are parts that you’re not building yourself. There are things that we get externally. There are things we get from other groups within Intel.

Software is a key thing — it doesn’t matter if my silicon is ready, I need the software to be ready. We are co-designing and co-developing so that we find issues early because every time you spin new silicon it interrupts your process and that has the potential to slow the cadence and increase time to market.

How far is Intel along the path to being a major player in SoCs?

To build a house first you have to lay a foundation. I can’t start thinking about chandeliers and faucets and appliances if I haven’t poured the cement. So we have poured the cement — we have identified areas of improvement and made focused investments that have started to pay off. We’ve invested in FFRDs, we have a strong IP portfolio and we’ve increased our investments in software to support hardware-software co-design.

We are not done completely but we continue to put that tension in the system. So the foundation is laid. The machine is cranking. This year we’ve introduced a 32nm low-power Atom SoC [codenamed "Lexington"] for value smartphones and a 32nm dual-core Atom SoC [codenamed "Clover Trail+"] platform for smartphones and Android tablets. This year we’ll also be introducing a 22nm quad-core Atom SoC [codenamed "Bay Trail"] and the 4th Generation Core 22nm SoC [codenamed "Haswell"].

Rani Borkar in Intel Archiecture Group engineering team meeting

"There have always been interdependencies, but with SoCs it's growing exponentially because there are parts that you’re not building yourself," said Rani Borkar, vice president and general manager, Intel Architecture Group.

As someone who has worked in the tech industry for nearly 25 years, what do you think is needed now to bring more women into technical fields?

It has to start as early as the middle school level. By the time women get to high school, they’ve made up their mind on a lot of things. Maybe they go to college and change, but not that dramatically from non-engineering to engineering or even non-science to science. It has to start in middle school at least if not earlier.

If I reflect upon myself, I can still remember the encouragement from my grade 8 math and science teacher. It was a long, long time ago when I was in 8th grade, but yet I can remember their faces, I can remember the support.

Each one of us who is in the field has to make sure that we are sharing our experiences, how we handle challenges and we must be honest. Do I want to give up my child’s [musical] performance today and go to some big meeting? No. I have to prioritize that because I’m not just an exec at Intel, I’m also a wife and a mother. It applies to men, too. They have to make choices too. I don’t see that as a gender one. We talk a lot about it as women because we think this is something we can’t overcome, but all you have to do is basically say what your priorities are and how you’re going to address them. There’s a whole, big community out there who’s willing to help you if you let them.