| 'The Race For A New Game Machine' interview with author and Xbox 360, PS3 CPU architect David Shippy |
|
|
|
| Written by Alexis M. (Rivithed) | |||
| Tuesday, 17 March 2009 | |||
|
The Race For A New Game Machine is an eye-opening look into the story of the microprocessor design behind the Xbox 360 and PLAYSTATION 3 game systems. If you're a fan of either and want to know where they were born, here's where you'll find how the core architecture of both happened, and at the same place. It would all unfold beginning at the STI Design Center (Sony, Toshiba, IBM). The book's authors, David Shippy (technical lead of the PowerPC microprocessor architecture on the PLAYSTATION 3 and Xbox 360) and Mickie Phipps (project manager at STI and PowerPC core) recount the events that happened since the beginning in 2001. To meet the Christmas 2005 deadline of two next-generation gaming systems, companies and cultures would merge, risks would be taken, tempers would flare, and ultimately casualties would result. But lessons would be learned and gamers would get their systems. Not at the same time, however. How did IBM's team manage to design the microprocessors that would be the brains in both these systems? How long could they go without either Sony or Microsoft knowing? How could they take on both game consoles at the same time and deadline? The man in the middle, David Shippy, recounts it all in The Race For A New Game Machine. It's a story that explains events, histories, and the people from companies like IBM, Microsoft, Sony, Toshiba, Apple, Nintendo, and Atari. The Race For A New Game Machine is a fascinating and rare look we've never been exposed to. By the end of my read through the book, nearly every 2nd page was dog-eared and margins were written on. There's plenty of information you'll take away from it, as well as valuable learning lessons. You won't look at your Xbox 360 or PLAYSTATION 3 machine the same after. To give you more insight into The Race For A New Game Machine, I interviewed David Shippy. Read the full interview below for just a glimpse of what you'll get from the book, as well as some questions and answers on the Xbox 360's "Red Ring of Death," the thought of a one console game world, what David Shippy is doing now, and what game system he's playing on with his family. Read the interview with David Shippy below.
GamingBits: Could you please tell us what your role was in "The Race For A New Game Machine" and the time period that the book takes place? GamingBits: Who would you say is the target audience of this book? Is it engineers? Business executives? Gamers? David Shippy: The book will certainly appeal to gamers and anyone in a business who is managing large or complex-type design. But we wrote the book in such a way that it really is telling a story. It really appeals to a wide audience, including non-technical people. GamingBits: What is the underlying message you'd like to readers to get from The Race For A New Game Machine? David Shippy: What I'd like the reader to take from it is if you setup the right environment and lead a team to achieve extraordinary goals, set the bar high, that it is possible to achieve really amazing technical accomplishments. In the processing of doing that, create an environment for innovation, risk taking, fun, and keep an optimistic attitude throughout the project despite many obstacles. In the case of the Xbox 360 and PLAYSTATION 3, we did achieve supercomputer processing capability in record time, during a schedule that was unheard of. GamingBits: Yeah, I read that by the end of this, there were over 500 patents, right? David Shippy: That's right, there were over 500 patents on all areas of chip design, including the work that went into the PowerPC core. GamingBits: When did you decide that this material would (and could) make for a good book? David Shippy: As we were in the midst of this kind of crazy corporate environment and having multiple masters and working long hours, Mickie and I use to kind of laugh about "wow, this would make an interesting story." By the end of it we realized this was a story that had to be told. Really no book has told an inside story like this. GamingBits: How did you recount some of the situations dating back to early 2001? David Shippy: It was all from memory and my best recollection of what happened all the way back in 2001. Mickie joined in 2002/2003 and she had her memories of events that occurred. GamingBits: The STI Design Team (Sony, Toshiba, IBM partnership) is explained in the book. Could you tell us a little about why the partnership began and for what purpose? David Shippy: The partnership began officially in early 2001 and we formed the joint design center Sony, Toshiba, IBM, to develop the next generation PLAYSTATION 3. The charter we got from Ken Kutaragi from Sony was to develop a new, start from scratch, innovative design that was nothing like anything that had been designed before. IBM came to them with their portfolio of microprocessors and Kutaragi basically said "we don't want any of those parts, we want this groundbreaking supercomputer on a chip technology for the PLAYSTATION 3." GamingBits: There are definitely more than a handful of issues you address in The Race For A New Game Machine besides development on the PLAYSTATION 3 and Xbox 360. From getting past cultural and language barriers with Sony and Toshiba teams, having too many cooks in the pot, and corporate ethics. Besides finally delivering the technology, what were some of your highlights? What were some of the positive things you can give from it? David Shippy: From an engineering relationship standpoint, one of the things I was most proud of was the Sony and Toshiba engineers ultimately realized they were working on this common core technology that would go both into their PLAYSTATION 3, but also into their arch-rivals technology, the Xbox 360 and despite knowing that they still put in the 100% effort and high quality design. It didn't matter what company they were working for, they just wanted to create the best product in the end. From a technology standpoint, we developed this technology that has the speed record, even today, for frequency. I describe that in the book with a way to measure that with a Fanout-of-four (FO4) type measurement. We delivered the fastest technology with extreme parallel processing. On the PLAYSTATION 3 we had 8 parallel processing cores. On the Xbox 360 we had 3 cores with 2 threads each, so we had 6 parallel threads. We delivered these two completely different architectures with extremely high speed and parallel processing that was just unheard of in the computing space. Especially when you compare it to computers that have come out in a similar time frame. GamingBits: Throughout the book, it seems like you are balancing the way you address your team. Dictatorial or not. You mentioned Ken Kutaragi, Sony's "Father of the PlayStation" was like a dictator in his approach. Whatever he demanded he got. Towards the heated year right before the final delivery of the microprocessor, you said you had to adopt that approach to get things done, especially with the younger, less experienced team. Do you recommend that approach to anyone under such extreme time and project deadlines? David Shippy: It really depends on the circumstances. When Microsoft came in the picture in 2003, there were a lot of people in the design centers, particularly IBMers, who felt like they were betraying their Sony/Toshiba partners that they had worked side by side with for two years. At that point in time, I had to take a more dictator-like management that said "look guys, we got to get over this." This is no different from when Intel sells x86 chips to both HP and Dell. There were other times I listened to the team in terms of the schedule. You had to adopt different management styles at different points in the project, and with different people. Mickie and I would try to keep the team motivate and playing as a team. When morale got low, we'd try to have some fun event. GamingBits: So time was tight for a Christmas 2005 launch for the PLAYSTATION 3, but things get really complicated when Microsoft stepped in for the Xbox 360. How and when did this happen? David Shippy: We'd been working on the PLAYSTATION 3 about 2 years, from 2001 to 2003. Then Microsoft approached us in 2003 and wanted us to design a similar technology, but not exactly what we had been designing for the PLAYSTATION 3. They liked the core technology that was basically extremely high frequency circuit design and very low power. They looked at it and said "we need something like that for us" and that's when things got real interesting. That's when we had multiple masters. We were driving a chip for Sony and Microsoft to achieve a 2005 game machine. GamingBits: When would you say Sony became aware of Microsoft's presence or when did Microsoft became aware of Sony's presence in the project? David Shippy: After Microsoft approached us in 2003, there was a public release that IBM had been awarded the Xbox 360 design. Now I think it wasn't completely obvious in that time what the ramifications of that were, but I think as Sony and Toshiba started seeing this common core we were working on, they knew we were going to use this core in other products and I think they started seeing some of the requests and demands made on that core and put 2 and 2 together. They realized that this same technology was in fact going to be in both of the game machines. GamingBits: I had read that someone had approached you with a printout or was it something from the article? David Shippy: That was actually prior to that, when Sony had actually filed a patent for the Cell technology and they did that without IBM's knowledge. We sort of found out about it after there had been this patent filed on the Cell technology, which we had actually all been working on for 2 years. That was really kind of a big surprise that they had not included IBM on that patent. GamingBits: Didn't you, or someone in the team, mention that IBM had undertaken more than two-thirds of the work, even though Sony and Toshiba were involved? David Shippy: Ultimately, in the STI Design Center, from an engineering headcount standpoint, yeah we were well over two-thirds of the Design Center involved in creating the chips. GamingBits: IBM let Microsoft know they were developing a new low power, high performance processor architecture without you knowing initially? David Shippy: Yeah, that's correct. They just formed a new division called "Engineering and Technology Services" and one of the engineers that was in that division showed Microsoft our whole portfolio of microprocessor designs, including this new design we were doing for the PLAYSTATION 3. Obviously he didn't reveal that it was PLAYSTATION 3, but he revealed the characteristics of that product. Extremely high frequency at very low power, which is really what you need for a game machine. GamingBits: You mentioned there were several closed door sessions with Bill Gates and maybe that person that had presented the technology. You don't think that at any point he could have figured out or had been told this was involved with the PLAYSTATION 3? David Shippy: No, not really. It was IBM, Sony and Toshiba's right to use this technology elsewhere. That was all part of the contract. When Microsoft came along, they were approaching IBM because IBM is one of the world leaders in developing microprocessor technology and high performance microprocessor technology. I think Microsoft just looked at it and said "here's an extremely interesting technology that looks like it will fit into the needs of the Xbox 360." I don't think at all they thought about this being a PLAYSTATION 3 technology. GamingBits: What is what you refer to as the "PlayBox" (PB for short)? David Shippy: When we were working in 2003 on this common PowerPC core that was going to be the brains of both the PLAYSTATION 3 and Xbox 360, we wanted to rename this common core to something other than what we were calling it for the PLAYSTATION 3, so we adopted this new name called "PB" which is kind of a play on words for "PlayBox", PLAYSTATION 3/Xbox 360 [laughs]. GamingBits: There were several instances you noted in the book where the Xbox 360 encountered an issue and it took some of the Sony/Toshiba team to resolve it. At one point they weren't aware it was related to the Xbox 360. Later on, they do seem very aware that they are working on an issue of the Xbox 360 hardware. Do you think they cared any more and just gave in? David Shippy: I think at that point they did realize the changes they were making, because of decisions that were being made about the Xbox 360 and really just took off their Sony or Toshiba hat and really just went and delivered. Despite knowing that it was going into their competition, they just wanted to deliver the best quality product they could. GamingBits: One of the many interesting and unique moments in the book recounts the "Common Core Committee." You said it was like a trial taking place, where IBM's lead on the Xbox 360 and IBM's lead on the PLAYSTATION 3 would express their concerns to you and Chekib Akrout, VP of entertainment and embedded processors at IBM. One of the unique needs of the Xbox 360, which was irrelevant to the PLAYSTATION 3, was the need for the PowerPC chip to emulate the x86 architecture of the original Xbox system to play the original Xbox games. You noted this was an issue brought up by Microsoft late in the game and it wouldn't be addressed on the hardware. Was that what happened at the Common Core Committee? David Shippy: Yeah, I was caught in the middle because I was developing this common core for both. There were requests that came in from both sides for various needs that the other side really didn't need. In the end, I think it was very fair to both sides. We would vote on it and all would come to an agreement. GamingBits: So anything that wasn't in the hardware was addressed through software, like the backwards compatibility of the original Xbox games? David Shippy: Yeah, I mentioned this in the book. Microsoft is a software company and they would quite often tell us "we're going to do this really aggressive schedule and hit Christmas 2005 and whatever you don't put in the hardware, whatever problem you have in the hardware, we'll just fix it with software." GamingBits: Yeah, you mentioned it was the "get it to market first at all costs and we'll worry about the bugs later" approach. David Shippy: Exactly. GamingBits: You mentioned in the book Microsoft went forward with the production of the Xbox 360 against IBM's advice right before the second round of hardware revisions. David Shippy: Yeah, so we got rev 1 done, and we got hardware back and everything came up nicely (OS and games running). We were starting to work on rev 2. This was going to be the production version of that part. Rev 2 would have to ramp up to very high volumes to meet the needs of Microsoft's game machine. The only way to do that was to go ahead and start what we call a risk ramp, where we didn't really know yet whether the rev 2 was going to be a final production part. But they had to go ahead and rev up the manufacturing line to start mass producing those to achieve the Christmas 2005 launch. This was really against IBM's advice because IBM is much more conservative and use to dealing with very low volume server-type machines. This was kind of unheard of to do a risk ramp, but it was Microsoft's call. They took a risk with rev 2, and their gamble paid off. It was a high quality part that went into production. GamingBits: I'm sure you've heard of the Xbox 360's "Red Ring of Death" issue, and I don't know if Microsoft has been able to pinpoint the issue on where this is happening, but I'm on my sixth Xbox 360 system and I still have my launch day PLAYSTATION 3 and launch day Wii. Some of the Xbox 360 systems I had died within minutes of turning them on. Do you think this is a risk that could have affected them in that way of thinking or do you think this is somehow catching up to Microsoft? David Shippy: I wasn't really that involved with that. My understanding is that particular problem had nothing to do with the CPU chip that I worked on. The risk they took in getting that part out of the door really had nothing to do with this. There were other components in the design that cause the "Red Ring of Death." I'm confident that it was not CPU chip related. GamingBits: You detail in the book that IBM had successfully delivered its part to get both the Xbox 360 and PLAYSTATION 3 set for the December 2005 launch. The Xbox 360 ended up beating Sony to the December 2005 launch target. Was the PLAYSTATION 3 ultimately delayed by Sony's graphics chip (GPU) decision? David Shippy: I think there were several factors. One of them was a late decision to change the GPU chip, as well as other decisions, like adding Blu-ray. The CPU chip was there and ready to go into production. GamingBits: What do you think of the idea of a one gaming system world, such as Valve's co-founder Gabe Newell has proposed. He was pretty vocal about having to develop games on the PLAYSTATION 3 because of the learning curve. At the same time Insomniac Games has been pretty successful on the PLAYSTATION 3. Is that a good idea? In this generation, even in the Wii system, isn't your microprocessor technology the brains of it anyway? David Shippy: I think it's an interesting concept. I think there will always be competing game manufacturers that want to differentiate their product. Even in the case of the PLAYSTATION 3 and Xbox 360, they both use this common PowerPC core, but they came up with completely different programming models and completely different architectures. I think that will never change. You'll always have different ways to achieve similar results. GamingBits: The race still seems to be going as we look to the hardware sales numbers every month for a gauge on the performance based on sales. Can you say it's any one in particular at this point that's leading and why? David Shippy: I think the Xbox 360 got the lead on the market share. And they also have got some really great games available, so they've taken an early lead. The PLAYSTATION 3 is in somewhat of a catch-up role now. Then you've got the whole other side of the equation which is Nintendo's Wii system, which has interesting technology in the whole motion approach. The fundamental microprocessor in that is not nearly as sophisticated as what's in the Xbox 360 or PLAYSTATION 3. In the future, they've got some catching up to do to create some of these more sophisticated games, even though they have this interesting motion technology right now. GamingBits: With all this contribution you've done to this current generation of gaming consoles, would you consider tackling the next Sony, Microsoft or Nintendo system? Or do you think these are in the works already? David Shippy: I'm no longer with IBM. I'm not really involved with next generation game machines. I'm sure that there are next generation chips that are being worked on for those designs, but I've moved on to other exciting engineering products and I'm not really involved with game consoles anymore. GamingBits: So what are you doing these days? David Shippy: I'm currently a Vice President of Engineering at Intrinsity. We develop high performance microprocessor cores, extremely low power, for handheld devices. So really kind of the whole embedded market of microprocessor design where you need extremely high frequency at very, very low power. GamingBits: You mentioned in the book that you're enjoying playing Halo 3 with your kids. Are there any other gaming consoles you play? David Shippy: Yeah, mainly Xbox 360. We enjoy all the Rock Band stuff and Halo 3. There's a lot of really fun products available there. That's probably where we spend most of our time. GamingBits: Is there anything else you'd like to mention on The Race For A New Game Machine? David Shippy: We wrote this book because it was a really cool story that had to be told. We also wrote it to inspire young people to go into engineering. That in fact engineering is exciting and challenging work, and fun. So we hope that young people especially in high school and college will read this and aspire to go into a technology field. We also describe a lot of the management techniques and leadership principles that we applied and learned throughout the process. I think passing those along to other managers that are driving complex types of organizations will be extremely useful. ### Thanks to David Shippy, Mickie Phipps and Sarah Fogarty for making this interview possible. Find The Race For A New Game Machine at your local bookstore or see more at www.raceforxbox360ps3.com.
Powered by AkoComment 2.0! |
or 
)
| < Prev | Next > |
|---|
