Comment by PostOnce | Hacker News Reader

Comment by PostOnce | Hacker News Reader

Parent Full thread

Perhaps it (Cell) would not have been a "dud" as you put it had IBM not been a morally bankrupt villain.

I've read that Sony was under the impression that the licensing agreement meant that IBM would market Cell tech to other customers, those customers being in other computer markets like datacenters and stuff, rather than to Microsoft, for the 360, at the same time that the PS3 was still in development.

"As the book relates, the Power core used in the Xbox 360 and the PS3 was originally developed in a joint venture between Sony, Toshiba and IBM. While development was still ongoing, IBM–which retained the rights to use the chip in products for other clients–contracted with Microsoft to use the new Power core in their console. This arrangement left Sony engineers in an IBM facility unknowingly working on features to support Sony’s biggest competitor, and left Shippy and other IBM engineers feeling conflicted in their loyalties."

from http://gamearchitect.net/2009/03/01/the-race-for-a-new-game-...

(it's a book, and worth reading)

I have written a substantial amount of code for the cell. I honestly believe that the entire approach it uses is a dead end -- no-one will voluntarily ever use such a machine again if there are any alternatives. The fundamental problem is that when you are writing code for cell, it is very hard to divide a large task into many subtasks where you do not have to constantly think of the entire problem when implementing small details. This means that writing code for a cell-like architecture doesn't scale. If you program fits on a whiteboard, a cell implementation can be easy to do and very fast to run. The moment you start building complex systems, things break down hard.

In the end, people just made each SPE do a single task, like dedicate one to audio, one to geometry, etc. There is not really enough parallelism like this in most software to support a cell-like approach now that even consoles get >4 real cpu cores. Real caches are the norm because they are extremely useful for programmers.

The villainy of IBM was just a slight additional problem over just how terrible cell was from a software standpoint.

I have not read the book, but I have a hard time imagining any world in which the Cell could possibly be successful. Its heterogeneous architecture thrust a huge amount of complexity onto software developers in exchange for meager gains. Writing good code for it was difficult and expensive compared to other platforms. Sony was just completely out of touch with reality.

In 2007, Gabe Newell famously complained that the Cell was "a waste of everybody's time. Investing in the Cell, investing in the SPE gives you no long-term benefits. There's nothing there that you're going to apply to anything else. You're not going to gain anything except a hatred of the architecture they've created."

Yep. One of the game devs I worked with on the PS3 was charged with the task of optimizing our physics and game code on the XBox 360 and PS3 platforms. The Cell architecture was bad enough, but the real icing on the cake was the XB 360 profiling/debugging tooling was a lot better too. Needless to say, he wasn't a fan of the PS3.

> I have not read the book, but I have a hard time imagining any world in which the Cell could possibly be successful. Its heterogeneous architecture thrust a huge amount of complexity onto software developers in exchange for meager gains. Writing good code for it was difficult and expensive compared to other platforms. Sony was just completely out of touch with reality.

This was a different time. At that time researchers tried to build clusters out of PS3s - because the speed advantages of the Cell made it worth and "regular" Cell clusters were much more expensive. Some years later GPGPU became feasible and one could forsee that it will become faster than the Cell, too, in near future - and at that time the same kind of researchers dropped their PS3 clusters and built GPGPU clusters. Don't tell me that particular in the beginning GPGPU was easier to program for than the Cell.

It was also the time when Apple switched to Intel CPUs. I know at that time IBM was also trying to sell the Cell to Apple, but Steve Jobs refused and decided for Intel instead.

This decision of Apple and the decisions of researchers to stop tinkering with PS3 clusters and build GPGPU clusters instead were in my opinion the two landslides after which the fate of the Cell was destinied.

I'm sorry, but I pretty much entirely disagree.

> Don't tell me that particular in the beginning GPGPU was easier to program for than the Cell.

The alternative was to use bog-standard homogeneous cores.

Yes, the air force bought a compute cluster of PS3s for some specialized calculations. I wouldn't read too much into that. It says little about the suitability of the architecture for more general purpose computing. Supercomputers were always weird.

> It was also the time when Apple switched to Intel CPUs.

I don't believe there was much chance of Apple moving to Cell. Their switch to Intel was because IBM could no longer seriously compete outside of a few niches. There's nothing positive to infer from IBM's unsuccessful pitch to Jobs.

> This decision of Apple and the decisions of researchers to stop tinkering with PS3 clusters and build GPGPU clusters instead were in my opinion the two landslides after which the fate of the Cell was destinied.

You're assigning far more importance to research group purchases than I think is warranted. They don't buy enough to create economies of scale. That's why researchers so frequently adopt consumer products already manufactured at scale, like the Novint Falcon, Microsoft Kinect, and gaming graphics cards.

The Cell was best-in-class for a few specialized use cases, but it was never going to take the world by storm. If we turn to a heterogeneous architecture in the future, it will be begrudgingly, after all simpler alternatives have been exhausted.

> You're assigning far more importance to research group purchases than I think is warranted. They don't buy enough to create economies of scale. That's why researchers so frequently adopt consumer products already manufactured at scale, like the Novint Falcon, Microsoft Kinect, and gaming graphics cards.

This is true, but in the consequences I have to disagree: Very often from this kind of "abusing" consumer products for research purposes there emerge quite interesting applications that do become quite popular and economically important. For example from such research there came the idea to use the Kinect as a 3D scanner - from this commercial applications emerged. Or from GPGPU (which at the beginning NVidia was quite the opposite of enthusiastic about) CUDA and later OpenCL emerged (which is much better to program for than abusing vertex and fragment shaders).

That is why I considered it is quite important for the future of the Cell when researchers went from tinkered PS3 clusters to GPGPU and called this a "landslide event for the future of the Cell".

It's not that it's heterogeneous. It's that each Cell processor only had 256K of RAM. K, not M. For code and data. You can pump data in and out from main memory in bulk, but random access is very slow.

So it's only useful for tasks that work like an assembly line - data flows in sequentially, gets processed, and output is pumped out. Great for audio. Lousy for everything else in games.

If you had 16MB on each CPU, the little CPUs might be useable. You might be able to run physics or pathfinding or NPCs in one.

I've heard a few times that the Cell wasn't all that bad in terms of performance, just very difficult to program. Not sure how true that is, but ostensibly the useability is just a tooling issue. Probably not a tooling issue that can be solved short-term though.

I did a university project with a Cell back in the day and it was awfully difficult to program. It was similar to GPGPU programming but instead of gigabytes, the SPUs had kilobytes of memory. Orchestrating the movement of data to/from SPUs was very hard to get right.

I never worked with cell or ps3, but that description reminds me a lot of working on the ps2.

The problem's more fundamental than that. The core of it is that parallelism requires explicit consideration in algorithm design.

The Cell demanded you to structure your program around small tasks that could be run in parallel across its seven vector cores. There's no getting around the fact that it's the programmer who has to break down problems to be small enough to fit on those cores without letting coordination overhead get out of control.

The quote says that it was the PowerPC core (PPE) was all that was used in the Xbox 360 processor and all that the PPE is is an implementation of IBM's Power ISA, one of several (see https://en.wikipedia.org/wiki/Power_Architecture#Specificati... under Power ISA v.2.03). It's no different than an ARM processor vendor re-using a core design for multiple customers.

The thing that made the Cell what it was, namely the 7 SPE units, were not used in the Xbox 360.

The Xbox did not have a cell. It had a power core without the cell.

The cell processor was intended to be the graphics solution for the PS3 [1]. The story I heard was that Sony was a hardware company and its engineers wanted technology that would work for new digital television applications and believed that the cell was going to be the perfect solution for everything, with its 8 "cpus". Except nVidia. Turns out a GPU is better at graphics than even eight very fast cpus. The PS3 wasn't going to have a GPU, and then they saw some Xbox 360 demos and had a brown pants moment. So they added a GPU at the last minute.

  [1] https://www.criticalhit.net/gaming/a-brief-history-of-the-playstation-3/

Sounds like Sony blundered when writing the agreement. If the chip is good for consoles, of course IBM will try to sell it to consoles.

Reply on news.ycombinator.com