Interview With Cell Architect Jim Kahle - 1 Teraflop on a chip by 2010

An Interview With Cell Architect Jim Kahle

Thursday, October 26, 2006
The PlayStation 3: An Interview With Cell Architect Jim Kahle

Dean Takahashi, 12:01 AM in Dean Takahashi, Gaming

Microprocessor_party_035 Jim Kahle is the chief architect of the Cell
microprocessor and the visionary behind the multiprocessing beast that
is the heart of the PlayStation 3. He led IBM's side of the chip
alliance with Sony and Toshiba. He has been designing microprocessors
since the 1980s and was one of the founding members of the Somerset
Design Center, the chip design house formed by Apple, Motorola and IBM
at the onset of the Power PC alliance. He was also the chief architect
of IBM's Power 4 microprocessor which was used in IBM servers and
Apple's G5 Macintosh machines. I caught up with him after his recent
keynote at a Cadence Design Systems conference.

DT: It seems like you finished the Cell chip designs early. The first
prototypes came out in 2004 and this is 2006. Did you still need a lot
of development time after that first tape out?
JK: We used that first tape out to get the initial software up and
running. There were modifications we did to the chip over time. The
design center is still active and participating. Our roadmap shows we
are continuing down the cost reduction path. We have a 65 nanometer
part. We are continuing the cost reductions. We have another vector
where we are going after more performance. We have talked about
enhanced double-precision chips. Architecturally we have double
precision but we will fully exploit that capability from a performance
point of view. That will be useful in high-performance computing and
open another set of markets.

DT: That sounds like it's not a PlayStation 3 chip?

JK: Yeah, it is a different vector. For us to extrapolate. We will push
the number of special processing units. By 2010, we will shoot for a
teraflop on a chip. I think it establishes there is a roadmap. We want
to invest in it. For those that want to invest in the software, it
shows that there is life in this architecture as we continue to move
forward.

DT: Right now you're at 200 gigaflops?

JK: We're in the low 200s now.

DT
: So that is five times faster by 2010?

JK: Four or five times faster. Yes, you basically need about 32 special
processing units.

DT: AMD bought ATI Technologies and they signaled that a combined CPU
and graphics processor is not so far off. They are going to do an
initial crack at it for emerging markets in 2007. Is that something you
see coming and is Cell anticipating this world already?

JK: If you look at a gaming system, there is obviously a close
relationship between graphics and the main processing elements. Over
time we will look to see how effectively we can make the main processor
and graphics tie together. I won't go beyond that.

DT: With Cell and PlayStation 3, was there a lot of thought about
whether you needed a graphics chip?

JK: We explored that to understand the bounds of what we could do with
the architecture. If you look at some of our ray tracing, ray casting
techniques, they are very effective. People have worked on some
software caches to help out the ray tracing. I wouldn't say that is
graphics processing because ray tracing is a little different. We've
explored the bounds on this to understand where it can contribute with
pure graphics processing. Over time, we have been exploring that.

DT: With Moore's Law, is it inevitable that they will wind up on one
chip?

JK: If you look at the PlayStation 2, eventually the graphics did get
integrated into the Emotion Engine. Sony has talked about that.
Definitely from a cost reduction view. Now we have to look at it from a
performance point of view too. That is something we have to study for
the future. Even beyond PlayStation 3. I don't know if it is
inevitable. We have to understand the pros and cons of it.

http://blogs.mercurynews.com/aei/2006/10/the_playstation.html