AMD CEO Lisa Su reminisces about designing the PS3's infamous Cell processor

https://www.tomshardware.com/tech-industry/amd-ceo-lisa-su-reminisces-on-helping-design-the-ps3s-infamous-cell-processor-at-ibm

by rbanffy • 2 years ago

358 246 2 years ago

tomshardware.com

tithe• 2 years ago

Peter Hofstee (one of the chief architects of Cell) gave a talk at UT Austin around 2008.

At the time, UT was building TRIPS[0], a processor capable of > 1 teraflop / sec.

In the middle of his talk (with several TRIPS members present), he made an off-hand remark "I think I could build a petaflop processer...without too much trouble,"* which caused a small stir in the audience!

[0] https://www.cs.utexas.edu/users/cart/trips/

* I might be misremembering the exact quote, but this was the gist of it.

lhl• 2 years ago

Perhaps relevant to the story, he was also on the team that delivered the first petaflop supercomputer (the most powerful supercomputer at the time) around that same time frame (it was around 24K processors):

Roadrunner is a 1.38 Pflop/s-peak (double precision) hybrid-architecture supercomputer developed by LANL and IBM. It contains 12,240 IBM PowerXCell 8i processors and 12,240 AMD Opteron cores in 3,060 compute nodes. Roadrunner is the first supercomputer to run Linpack at a sustained speed in excess of 1 Pflop/s.

https://dl.acm.org/doi/10.5555/1413370.1413372

https://en.wikipedia.org/wiki/Roadrunner_(supercomputer)

bbatha2• 2 years ago

What a pain in the ass machine that was to write code for. You had 3 different processors to manage with 3 different architectures and all of the challenges the game dev community had with keeping the SPEs hot. But the SPEs were not enough to solely rely on for number crunching performance unlike today's GPU compute machines, so you also needed to do number crunching on the opterons and the main cell core for optimal performance unlike with a GPU where the CPU is mostly just keeping the GPU memory full. Then to make matters worse the cell had a different endianess than x86 making shared memory very annoying to work with.

wmf4• 2 years ago

FLOPS are easy; programming is hard. Cell was the wrong direction and TRIPS/TFlex hasn't happened either.

peutetre3• 2 years ago

The PlayStation 3 pretty much lost the generation handily to Nintendo's cheap, casual-friendly Wii and Microsoft's less powerful but easier Xbox 360.

No it didn't. The PlayStation 3 outsold the Xbox 360. Have a look at the 7th generation charts on VGChartz:

https://www.vgchartz.com/

It's true the Wii sold the most.

The 8th generation didn't go well for Nintendo. The Wii U was a flop. But they're back on top with the Switch.

EcommerceFlow• 2 years ago

You're talking about an extremely end of cycle resurgence. For the vast majority of the lifecycle, the 360 outsold the PS3 and had better compatibility with multiplatform games.

nixass• 2 years ago

PS3 ended up selling more than double what Xbox was sold "only because of extreme end of cycle resurgence"? Cmon

It's true the devs have become more comfortable develeoping games as the console matured but that's about it.

Xbox was always flop comparing to PS

peutetre3• 2 years ago

Nope. The Xbox 360 launched in November of 2005. The PlayStation 3 launched in November of 2006.

The PS3 started outselling the Xbox 360 for the year in 2007 and kept steadily gaining on the Xbox in total sales for the generation until it overtook it.

Xbox's one year head start kept it level with the PlayStation 3 for quite a while but it lost out in the end.

Look at the global yearly sales figures on VGChartz. From 2007 the PS3 outsold the Xbox 360 every year except for 2008:

https://www.vgchartz.com/yearly/

tokai• 2 years ago

Its wild that we are almost 20 years on, and consolewar posting about which is bigger/best is still going on.

Rinzler894• 2 years ago

Hey as long as the PC is still king, let them fight over the consoles :D

bee_rider3• 2 years ago

Nintendo is practically in a different market anyway. People who want a Nintendo console are mostly looking for Nintendo’s first-party characters. Lumping them in with Xbox and PlayStation is like combining Barbie and GI Joe under “dolls.” Technically correct but not very descriptive.

Snark: Nintendo’s market is kids, people with kids, and people who were previously kids. Xbox and PlayStation are fighting over the market of people who’d be better served by a PC but don’t want to deal with all that. /Snark

SahAssar2• 2 years ago

kids, people with kids, and people who were previously kids

Who is not included in that? Toddlers?

r00fus• 2 years ago

and people who were previously kids.

That's a pretty damn large market (= all adults). I think you meant to say "people who were previously kids playing Nintendo"

Clamchop2• 2 years ago

I don't disagree that Nintendo appeals to a market that the others don't, and vice versa, but I'd be surprised if there wasn't a significant number of people who will reliably buy a game console, but won't buy two consoles.

That is to say, I still think there's significant competition for sales.

And if Nintendo didn't exist, it would be easier for Sony and Microsoft to take in the market for cute, whimsical, child-friendly, instead of having to distinguish themselves by being the more serious or more powerful option.

And handhelds; Nintendo has more or less stolen that entire segment, Steam Deck still being relatively niche and PS Portal being a different kind of thing. Sony's handheld ambitions are moribund and not for lack of trying.

dehrmann• 2 years ago

but SMT wouldn't emerge until 2002

I think they're counting P4 hyperthreading, but I remember dual-CPU Pentium II boards.

buildbot3• 2 years ago

IBM had SMT in 1968 - https://people.computing.clemson.edu/~mark/multithreading.ht...

trynumber9• 2 years ago

I am not following. The page says it was investigated but not done.

wtallis• 2 years ago

SMT isn't SMP.

adrian_b• 2 years ago

SMT (including that branded as Intel HTT) is something very different from multiprocessors, like the dual-CPU boards.

The concept of multiprocessors was already well understood around 1963, i.e. the multiplication of the performance of a computer by the multiplication of the CPU hardware.

On the other hand, the purpose of SMT is to enhance the throughput of a computer by ensuring that the existing execution units do not stay idle, by executing instructions from multiple program threads in order to keep them busy.

The modern use of SMT has been started by a research paper published in 1992, as a proposed improvement to the superscalar out-of-order CPUs already popularized by IBM POWER, but the name SMT (simultaneous multi-threading) has been coined only later, in 1997 (in the IEEE Micro from September/October 1997).

IBM had already used SMT in the experimental Advanced Computer System in 1968 (and they have filed for SMT patents in 1971: US 3,728,692 & US 3,771,138), but that project has been canceled and its superior architecture has been mostly forgotten for several decades.

CPUs that could execute instructions from multiple program threads, but not during the same clock cycle, like in SMT (hence the "simultaneous" of SMT), are older than the IBM ACS, for instance such a processor with fine-grained multi-threading was the peripheral processor of CDC 6600 (in 1963). SMT can exist only in superscalar CPUs, because only there multiple instructions are initiated in the same clock cycle.

buildbot3• 2 years ago

Oh wow, I did not know Lisa Su had a hand in designing Cell - that’s really cool. In some ways the Cell architecture was very ahead of its time - if you look at it a bit funny, the PPE/SPE map to something like todays grace/hopper - big branchy core to orchestrate smaller but more numerous vector cores.

Made designing/porting games a massive headache though because nothing else was like it.

Rinzler894• 2 years ago

Correction: Cell was ahead of it's time when they started drafting the concept for it in mid-2000 and the most powerful consumer CPU at the time was the Pentium 3, but by the time it launched on the market in 2005, X86_64 was the new undisputed king of CPU performance and ARM in mobile lower power sectors, with ATI and Nvidia leading the GPU race, leaving the Cell processor as an overpriced paperweight that was a jack of all trades, master of none and difficult to program to boot, and the new market had no use for such a thing.

I think neither Sony nor anyone else expected the consumer CPU and GPU industry to evolve so quickly in that time, otherwise they would not have spent hundreds of millions developing anew architecture from scratch. Intel also made the same mistake back then when they tried to reinvent the industry with the new Itanium CPU architecture. Hindsight 20/20.

Those years were indeed wild for tech progress. You bought a new CPU, suddenly two years later you have 64 bit CPUs as the new thing. You then bought a 64 bit CPU to be like the cool kids and suddenly dual core CPUs are all the rage.

scrlk2• 2 years ago

IIRC, the original PS3 design was rumoured to have dual Cell processors. It didn't meet performance targets, so the Nvidia GPU was wedged in quite late on in the development cycle.

monocasa4• 2 years ago

Eh, the Cell would have been a fantastic processor if it weren't for hitting the end of dennard scaling like a brick wall.

In the design phase they were expecting almost 5ghz at launch, and to hit around 10ghz at the end of the product lifecycle. When the process guys came back later to let the architects know the reality we've sort of now all internalized (high end consumer electronics can expect ~3ghz with good cooling solutions) a lot of simple but fast pipeline designs stopped making sense. These are the exact same pressures that killed off the Netburst architecture too, making intel go back to a PIII design and make it smarter rather than clock markedly faster.

foobarian• 2 years ago

Those years were amazing because every new console generation had a wildly different artisanal design. NES to SNES to N64 to Switch, Jaguar, Sega, PS to PS2 to PS3, Xbox... and then at some point they all turned into disguised PCs.

h0l0cube• 2 years ago

but by the time it launched on the market in 2005, X86_64 was the new undisputed king of CPU performance

PS3 Cell could do about 154 GFLOPs across it's 7 cores (6 SPUs and 1 PPU, excluding the GPU), while Xeons were in the low double digits. What really was awkward about the Cell architecture was the DSP-style nature of it, which works great for something like shaders or audio processing with a clear pipeline, but not for general purpose computation. It was really a successor to the PS2 in this regard, where data flowed between specialized and memory constrained processors using a DMA.

Sony nor anyone else expected the consumer CPU and GPU industry to evolve so quickly in that time

PS3 used an NVidia GPU. Though there was some clunkiness around sharing memory between the Cell and GPU compared to the UMA on PS4, but not especially so compared to a graphics card on a desktop machine.

Der_Einzige• 2 years ago

There were memes circa this time on places like 4chan about "the power of the cell" and implying that it was capable of doing anything. Sony marketing about how good this chip was circa 2005-7 was strong.

scrlk2• 2 years ago

The power of the Cell bought us real time rendering of Giant Enemy Crabs, as well as real time weapon changes. All for 599 US dollars!

toast0• 2 years ago

It's a shame they didn't license blast processing.

jeroenhd• 2 years ago

The cell processor was pretty good at some calculations. Unfortunately, it wasn't great for writing video games and neither was the tooling.

I recall various news stories about governments buying up PS3s because they could be used for cheap and effective computation at scale.

astrange2• 2 years ago

Outnumbered by "the PS3 has no games" memes though.

https://knowyourmeme.com/memes/playstation-has-no-games/phot...

OCASMv2• 2 years ago

My favorite meme from the era:

https://i.postimg.cc/rwn0kT7W/Xenon-Vs-Cell.jpg

magicalhippo5• 2 years ago

Had a heterogeneous programming class, where we had to implement various things on the PS3, like a basic MPEG-ish encoder and such.

Was an interesting experience, though all I remember now is how it required careful exploitation of the vector units in the SPEs to get any decent performance out of it, and how annoying it was to synchronize between the SPEs and the PPE.

For each assignment the prof would include a benchmark, so we could compare the performance in class. Was a huge difference between the students that spent time optimizing and those which did a basic implementation.

Waterluvian4• 2 years ago

I'm super curious what program/course was set up with developer/jailbroken PS3s to use as lab material. Was this specifically about game/console dev?

selykg• 2 years ago

I believe the PS3 had an option to install Linux on it at some point. No? That would've made it a neat option for classes like this.

magicalhippo5• 2 years ago

They had Linux on them. Don't think I ever saw them, just had network access.

mywittyname• 2 years ago

Not OP, but we had PS3s at school as well. Early units ran linux. It was sandboxed to a degree in order to prevent it from being used for software piracy. OP's project was cool, ours was basically implementing PCA on one.

I got the impression that Sony encouraged this use of the console early on. They were probably aiming to establish themselves in the scientific computing niche. We also had nvidia-sponsored labs where they taught CUDA (or at least tried to - it was pretty difficult to do without support and most people took the class for access to computer labs equipped with amazing gaming machines). We all know nvidia won that war.

stevenwoo• 2 years ago

The first generation devkit PS3s cost something outrageous like 10000 dollars and were the size of large pizza boxes IIRC (shrunk from three pizza boxes stacked on top of each other - eventually they shrunk the devkit down) and has some pretty stringent requirements to acquire. On the team I worked on only the graphics developer got a PS3 devkit, the rest of us lowly programmers got PS3 test kits, that allowed us to deploy/debug but without the same level of support tools, I think it was the same situation with the XBox 360 but the Microsoft developer tools were excellent even with just the test kits. I vaguely remember the cost of the test kits being double or triple the retail price of a regular console, and we had to warranty my test xbox 360s quite regularly since they would red circle of death.

skavi• 2 years ago

that sounds like a super fun class. wonder if it’s still offered?

magicalhippo5• 2 years ago

Seems like they still do[1], though I doubt the PS3 is still in the mix.

Though from the published material for 2024 it seems not to have changed too much in principle, eg the codec63 I recall implementing parts for.

edit: oh and yes, one of the highlights of my degree.

[1] https://www.uio.no/studier/emner/matnat/ifi/IN5050/index-eng...

crote• 2 years ago

The Program Optimization course I followed did something similar with regular CPUs / GPUs. It started out as "don't do stupid things", but it quickly went into SoA vs AoS, via hand-writing AVX intrinsics, into eventually running OpenCL code on the GPU.

Part of your grade depended on the performance of your code, with a bonus for the fastest submission. It led to a very nice semi-competitive semi-collaborative atmosphere.

hmottestad• 2 years ago

I really wanted do take that course. I had taken the OS course where you build an OS from scratch, so this seemed like the next thing to do. The OS course was for 20 points, but the heterogeneous programming course was only 10 points. People said it was way more work than 10 points though and I ended up dropping it because I couldn't justify spending that much time on it when I had other courses to take and a thesis to write.

Here is the OS course: https://www.uio.no/studier/emner/matnat/ifi/IN4000/index-eng...

magicalhippo5• 2 years ago

Hah yea I quickly discovered that matnat[1] study points weren't like other study points. Felt like we had to do 2-5x the work compared to the students I knew going to other faculties. But in general the courses were very rewarding I feel.

I heard about the OS one, sounded super cool and also a lot of work from what I gathered. Heard about it too late and required a buddy. Oh well, can't have it all.

[1]: faculty of math and natural sciences

DanielHB2• 2 years ago

Just out of curiosity what kind of hard projects did you do at the bachelor level? I had to build a unix shell, implemented some CPU scheduling algorithms (but didn't actually integrate it into a proper OS), did some B-tree stuff as well as implementing JPEG-like algorithms. Probably the one that took the most time was designing a CPU that worked in a real FPGA with its own instruction set and assembler (it was such a pain in the ass to troubleshoot, my crappy laptop took 30 min to compile the project and run simulations).

When I talk to people outside my home country it feels most don't do these kind of things at the bachelor level. But, like you, people who go on a masters then might do one of these things and go much more in-depth than I did.

rbanffy5• 2 years ago

how annoying it was to synchronize between the SPEs and the PPE.

This is the kind of thing where library and OS support is essential. Developer ergonomics is not that hard to get right (because developers experience it) and yields immense benefits when it's done.

I see that today with GPUs where you need to concern yourself with moving data and setting pipelines up, and thinking about the peculiarities of that particular GPU.

Rapzid2• 2 years ago

ATI developed the graphics chips for 360 and Wii. Feels a little too revisionist to say AMD produced the chips even if they did purchase ATI a number of months before Wii released..

fourfour32• 2 years ago

It’s a bit tricky even to say that ATI developed them - they were designed by ArtX (mostly ex SGI staffers who had also worked on the N64), who were bought by ATI.

Rapzid2• 2 years ago

Perhaps but ArtX was acquired by ATI long before. The ATI logo was already on the GameCube at launch.

wmf4• 2 years ago

Maybe Wii was ArtX but wasn't the 360 using a mainstreamish Radeon GPU?

Waterluvian4• 2 years ago

I'm curious about how such a bad idea made it to market. Not that the technical concept is bad, but the business decision to build an entire console generation on such a unique, ultimately migraine-inducing architecture.

MBCook12• 2 years ago

Remember Cell was “the future”. It was going to go into the PS3, but also TVs and computers and whatever else. If you squint it’s a bit like Apple putting a ton of investment in the A series chips that then went into iPads and Macs (via the M series).

Problem is: no one wanted it. So none of that panned out. If it had gone on to be something like Intel’s Core 2 Duo that showed up everywhere in hindsight it would look smart.

initplus2• 2 years ago

Difference is that Apple invested heavily in backwards compatibility, even old x86 code performs well on their chips. Meanwhile cell requires a reworking of your entire program to take advantage of it.

duxup• 2 years ago

I remember promotional materials showing it would be in refrigerators. I can’t imagine what they thought that refrigerator would do with it….

initplus2• 2 years ago

It’s less a confusing one off decision to use a weird architecture, and more a failure to reconsider the status quo and stop using weird architectures.

Previous PlayStation hardware was just as odd. Ps2 reused old surplus Ps1 processors just for audio, and had coprocessors just like cell.

0x4574• 2 years ago

I don't think Sony expected Xbox 360 so soon. PS2 was sales were still good and there was no console on a market better than PS2 at that time.

Then out of nowhere Xbox 360 comes out, Halo that was a macOS title suddenly a main and a well selling IP on Xbox. Not just that, it also picked HD-DVD as its format posing a threat to Blu-ray.

Sony had to put something out.

cmpxchg8b• 2 years ago

Halo first came out on the original Xbox, not sure it ties into the PS3/360 calculus other than being a launch title.

Talanes• 2 years ago

The 360 didn't pick HD-DVD as it's format, the built-in drive was just regular DVD. The HD-DVD drive was an external add-on that they sold along with the 360 Media Remote.

vel0city5• 2 years ago

Halo came out in 2001, four years before the Xbox 360 was released.

The Xbox 360 released its HD-DVD drive a few days before the PS3 hit store shelves in North America. Sony had established the PS3 would use Blu-Rays a year before the Xbox 360 was even publicly announced.

I think there were a number of things Sony wasn't expecting about the Xbox 360, but HD-DVD or Halo wouldn't have been any of those things.

failuser2• 2 years ago

You forget that it was normal before. PS2 had many separate chips you need to program and balance throughout and latencies to make the most of hardware. PS2 was able to achieve 60 fps in many impressive games. Console hardware was very specific, the original Xbox being just an Intel PC was an exception.

A4ET8a8uTh02• 2 years ago

I have a weird problem with the article. It does start with a question over 'infamous' qualifier, but it goes into something deeper that has been bothering me over the past few years.

Lets start from the end, game consoles these days are basically optimized PCs. They are effectively nothing like the first, second, or even third generation consoles that had fairly unique hardware, desirable exclusives and so on.

Cell may have been difficult to program well for, but it speaks volumes about the current state of consoles that devs worry about being able to be lazy with ports. The idiot consumer will buy whatever anyway. The side effect of the first generations' relatively unique hardware profile is that games actually had to be optimized and tailored to the hardware. No lazy ports ( I still get pissy sometimes over ps3 dragon age port ). Exclusives had to be good or at least showcase what the console could do.

So is that what makes Cell infamous? The thing made news when could be run in a cluster[1].

I have a different take. Cell is the remnant of the consoles, when they were still worth one's time.

[1]https://en.wikipedia.org/wiki/PlayStation_3_cluster

MBCook12• 2 years ago

The article gives the answer.

It’s not about being lazy with ports. It’s that the asset budgets are so large that making games is way too expensive. For any normal company you basically _have_ to launch on multiple systems if you want a financial payback.

And every time graphics get better the problem gets worse. More models, more textures, more detail, more everything. $$$

Unless you have a monster hit (Zelda, Last of Us) you may be unable to make your money back on one machine. Even if you could, why not just make more money by adding in the other machines?

There’s a reason pretty much only 1st party studios do exclusives.

A4ET8a8uTh02• 2 years ago

If there is one thing that Steam has proven, it is that games absolutely do not have to focus on 8k assets, celebrity likeness and voices. In other words, that is not really an answer. Or at least not a real answer. It is, at best, a part of it that glosses over why people play games.

MisterTea• 2 years ago

To me the infamy is owed to Sony's "super computer on a chip" hype marketing (similar to the hype around the Emotion Engine int he PS2) surrounding the architecture coupled with the difficulty in programming it. The big issue was SPE's had no way to access main memory so you had to copy the data from RAM into each SPE which had its own local memory.

mrguyorama• 2 years ago

They did a similar marketing push when the PS2 was developed, saying it couldn't be exported to certain countries due to dual-use restrictions because it was "too powerful"

ProfessorLayton7• 2 years ago

While I too miss exotic hardware (N64 being my all-time favorite), ultimately what people want to play is great games and the hardware is just a means to an end.

The bigger issue with current consoles imo, specifically the PS5 and XB, is their input homogenization and stagnation. Consoles have historically been forward-looking via their controllers to enable new ways to play games:

[NOT including optional accessories, only out of the box abilities]

- N64: Analog controls, 4 controller ports, controller expansion ports

- PS1: Dual analog controls

- PS2: Pressure-sensitive buttons, built-in rumble

- Dreamcast: ...

- Xbox: ...

- Gamecube: ...

- Wii: Motion controls

- PS3: ...

- Wii U: Gamepad screen

- 360: ...

- PS4: [Correction!] Touchpad

- Switch: Detachable and sharable controllers

- XB1: Kinnect

- PS5: ...

- XB1X: ...

I'm definitely not implying that every new input type was successful (See Wii U, kinnect), but consoles up until about PS3/360 had something new to offer regarding ways to experience games. This has all but stopped, with Nintendo being the only one to consistently introduce new ways to play that cannot easily be replicated on a PC. Plugging a PS5/Xbox controller into a PC yields the same or better experience as buying a dedicated console these days, so why buy the console at all?

dfxm12• 2 years ago

why buy the console at all?

Putting console exclusives aside, I can build a PC that does everything I need it to do but play some new games + a console for less than it costs to build & maintain a "gaming rig". Also, my PS4 is still working after ~10yrs. I would have to upgrade my PC more frequently than that to keep up with new games.

You're also forgetting the two-point capacitive touch pad on the PS4 controller. It's OK, almost all the video game developers did too.

DuckConference• 2 years ago

PS5 controller replaced rumble with fancy haptics and force feedback on the triggers

MBCook12• 2 years ago

why buy the console at all?

It’s easier. I don’t have to fiddle with drivers. Or settings. Or choosing which card to upgrade to when. Or debugging mismatches. I don’t have to have the OS + Steam/Epic/whatever + games.

The games are written to what I have and I don’t need to worry about it. The OS is completely built around playing games and making it easy for me to acquire them and I don’t have to think about any of it.

I had the years of building gaming PCs and messing with all of that and doing all the research. And I had fun. And at this point in my life I just don’t wanna deal with it I just wanna play a nice game.

pezezin4• 2 years ago

While I agree with you, I think that we have reached a local optimum that is difficult to improve anymore. I mean, what else is missing from modern controllers? Unless we move to VR with direct neural control, I don't see a way forward.

numpad02• 2 years ago

DC had Visual Memory(programmable tamagocchi as memory card) and modem, 360 had near gold standard ergonomic gamepad(also first USB in console)

pezezin4• 2 years ago

They are effectively nothing like the first, second, or even third generation consoles that had fairly unique hardware, desirable exclusives and so on.

I am not sure how you are counting them, but the second generation were the Atari 2600 and its peers, and the third generation were the NES and the Master System. Consoles had unique hardware until the 7th generation (PS3 and X360).

lomase3• 2 years ago

Calling developers lazy is the pinacle of not understanding how project managment works.

driscoll42• 2 years ago

There's a wonderful blog post on the PS3 Architecture - https://www.copetti.org/writings/consoles/playstation-3/ that gives a good overview of the Cell processor with linked resources if you want more detail.

hadrien01• 2 years ago

Wow this website is absolutely amazing! Full of details about consoles, and with a really pleasing mixture of text, images, and even 3D models.

kevvok• 2 years ago

There’s a great book about the development of Cell and Xenon (which went into the Xbox 360) called “The Race for a New Game Machine” that was co-written by a couple folks who worked on them at IBM.

khalilravanna• 2 years ago

My biggest lament with the PS3 is the forward incompatibility making many great games locked to the platform (MGS4, Demon’s Souls, etc.). Meanwhile in Xbox land there are *633* 360 games you can play on any new Xbox by just slipping the disc in. Now some of this is no doubt due to differing approaches to business by the corporate masters but from what I’ve read a lot of it is the unique and befuddling architecture of the PS3.

https://en.m.wikipedia.org/wiki/List_of_backward-compatible_...

throwaway484762• 2 years ago

Xbox canceled development on their emulator though.

copx• 2 years ago

Despite its exotic architecture the PS3 can be emulated well on "normal" machines, though.

I was amazed that I am actually able to play PS3 games on my (older, not at all top of the line) laptop thanks to the brilliant RPCS3 emulator.

A PS5 is way more powerful than my laptop, so in theory SONY could just make all those PS3 games available through emulation.

failuser2• 2 years ago

It took a decade and it’s still not 100% accurate. By now PS5 is powerful enough, but Sony would rather sell you a remaster or remake.

hinkley• 2 years ago

The PlayStation 3 is the primary reason I maintained ownership of discrete dvd/Blu-ray players and never got into the PlayStation ecosystem. There were a couple PS3 games I was interested in and if they had run on the PS4 I would have owned a 4 and thus likely a 5

hehdhdjehehegwv• 2 years ago

She’s low-key one of the most successful engineers ever in terms of bringing cutting edge technology to the mass market, but AMD never gets the fawning press attention of NVIDIA, and still has the lingering image of being the poorer cousin of Intel.

Of course, as a woman engineer to make it this far she’s used to not getting the limelight in a male dominated industry - which may also explain some of AMDs recent success.

blitzar• 2 years ago

Their press should focus more on how much of a failure AMD are (vs NVIDIA in particular) rather than fawning puff pieces like this one.

noobermin• 2 years ago

The death of the powerpc is a bit sad to me as it feels like the lineage of the 6502 has really ended as of now. As much as everyone hates too on x86, I fear one day it too will disappear and everything will be on some arm isa

theandrewbailey2• 2 years ago

PowerPC's sort of descendant/predecessor POWER is still developed and sold. Raptor Computing seems to be the only company selling POWER based systems to the general public.

fb03• 2 years ago

I wonder what is the appeal of these systems, today? On what demand would one chose a POWER system like this one (which I think it's pretty cool) over a, say, more standard X86 or ARM server?

wk_end2• 2 years ago

What's the connection you had in mind between the 6502 and the PowerPC? As far as I know they're geneologically unrelated.

The 6502 was descended from the 6800 - one of a few ur-microprocessors - and somewhat clumsily evolved into the 65816 before its brilliant but highly specialized design (hard to target for HLLs and which fundamentally assumed memory could be accessed roughly as fast as the CPU was running) couldn't keep up with the times. As far as I know that's pretty much it for its branch of the CPU family tree.

Meanwhile - and I admittedly know less about this - as I understand it the PowerPC is ultimately descended from the IBM 801, which was a high-end workstation and server chip. There might've been some influence from Motorola, and Motorola made the 6800 - the 6502's father - so in that sense the 6502 and the PowerPC might be very distant cousins. Still, that's pretty tenuous.

My understanding is that the original ARM team was directly inspired by the 6502 - so if its legacy lives on anywhere, maybe it's there.

kalleboo2• 2 years ago

IIRC PowerPC was IBMs core melded to Motorolas bus design, so that Apple could easily integrate it into their existing board designs (and make upgrade cards)

kalleboo2• 2 years ago

I didn’t have high hopes for RISC-V outside of the cheap low-power chip space but watching ARM try to destroy Qualcomms new fast chips makes me believe we may see a big ARM/RISC-V showdown in the future, the instruction set wars aren’t over yet

penguin_booze• 2 years ago

Trivia: Lisa Su and Jensen Huang are relatives: https://www.tomshardware.com/news/jensen-huang-and-lisa-su-f....

TheMagicHorsey• 2 years ago

If China ever takes over Taiwan, America is so screwed.

Actually, there is such a scattered supply chain for semiconductor chips, that any war that disrupts international trade in a material way, would bring that entire industry to its knees. So many critical suppliers in the semiconductor supply chain are single companies, or single country oligopolies.

eunos• 2 years ago

Well they (and Morris Chang) are supposedly 'recent' immigrants from Zhejiang in mainland so the next gen human resources are as good as secured

rbanffy5• 2 years ago

If China ever takes over Taiwan, America is so screwed.

The 21st century will not be a remake of the 20th. Empires rise and fall, and that's in their nature to do so. So far, China has managed theirs very well, and I don't see a fall in the cards for this century.

rldjbpin• 2 years ago

side question: is there a taiwanese version to check if someone you're dating is related to you?

see for context: https://news.ycombinator.com/item?id=5561505

mrkramer3• 2 years ago

The interviewer points out that Sony's PlayStation 3 is viewed as one of its least successful consoles, which is true. The PlayStation 3 pretty much lost the generation handily to Nintendo's cheap, casual-friendly Wii and Microsoft's less powerful but easier Xbox 360.

In Europe, PS3 was widely popular especially in conjunction with Call of Duty because Xbox never really gained ground in EU, at least that was my impression growing up as an EU gamer. Wii was also popular af at the time but PS3 and Wii buyers were different audience, Wii was more casual type of gaming and it was suitable for family gaming while PS3 was for more serious gamers who played more hardcorish titles like COD, FIFA and PES and ofc there was and is PC master race for real hardcore gamers.

BolexNOLA2• 2 years ago

That may be true but overall, the PS3 was the loser at the time and it took until basically the end of that console generation, consisting of many price cuts and years of developing a strong catalogue, for them to (narrowly) outsell the 360 worldwide.

Considering the absolute dominance of PSX and PS2, it was definitely not what they were hoping for. Not quite as dramatic as Nintendo with the market loss from SNES to the N64, but I’m sure Sony had a lot of similar conversations. You also have to consider that Xbox was the distant third place console in the previous generation, so barely losing to Sony at the last minute was definitely considered a huge success. Especially after the disaster that was the RROD. If that hadn’t happened Microsoft likely would have beaten Sony’s sales.

mrkramer3• 2 years ago

My last gaming console was PS2 and yea I loved it because gaming consoles are meant for "couch gaming" with big TV screen and as one gaming friend of mine told me: "When you buy gaming console, you are set for the next 5 years." Moore's law kinda kills the fun for us PC gamers but I still prefer gaming on PC than on gaming consoles.

tekla• 2 years ago

You must not have been there for the nogaems memes.

kimixa2• 2 years ago

Yes, for much of it's lifetime it was behind the 360 in worldwide sales, and might have been a disappointment for Sony, especially after the PS2 domination, but I wouldn't say it "Lost Handily" to the xbox360. It may have taken time, but it did outsell it.

It seems an incorrectly strong wording by the interviewer.

And people seem to sleep on how powerful the xbox360 was - the CPU was a fair bit smaller than the cell, true, (~150m transistors vs 250m in the cell) not not really the multiples some articles seem to imply.

And the GPU in the xbox was arguably a fair bit more advanced - being a little larger (~300m vs ~335m transistors), and being a unified shader architecture that came to dominate GPUs from then on. And the more flexible architecture enabled more GPGPU-like tasks that the cell SPUs might have handled on the PS3. And these early GPGPU models were also a bitch to program for - like the cell SPUs were.

So it's not like one was simple and easy, the other crazy complex and advanced, but more one had the fancy experimental stuff on the CPU/Cell die, the other on the GPU die. Just the techniques used on the GPU continued to be developed and supplanted much of the Cell-style tech.

mrkramer3• 2 years ago

Yes, for much of it's lifetime it was behind the 360 in worldwide sales, and might have been a disappointment for Sony, especially after the PS2 domination, but I wouldn't call say it "Lost Handily" to the xbox360. It may have taken time, but it did outsell it.

As far as I understood, Halo was the killer app for Xbox but I suppose it was the killer app only in US, because like I said, from my EU gamer's point of view, Call of Duty was all the rage first on PC and then afterwards on PS plus FIFA and PES were huge on PS as well.

sillywalk5• 2 years ago

how powerful the xbox360 was

Interestingly, the Xbox 360 (Xenon) CPU cores were based on the Power Processing Element (not the Synergistic Processing Elements) from the Cell processor.

From [0]:

"When the companies entered into their partnership in 2001, Sony, Toshiba and IBM committed themselves to spending $400 million over five years to design the Cell, not counting the millions of dollars it would take to build two production facilities for making the chip itself. IBM provided the bulk of the manpower, with the design team headquartered at its Austin, Texas, offices. Sony and Toshiba sent teams of engineers to Austin to live and work with their partners in an effort to have the Cell ready for the Playstation 3's target launch, Christmas 2005.

But a funny thing happened along the way: A new "partner" entered the picture. In late 2002, Microsoft approached IBM about making the chip for Microsoft's rival game console, the (as yet unnamed) Xbox 360. In 2003, IBM's Adam Bennett showed Microsoft specs for the still-in-development Cell core. Microsoft was interested and contracted with IBM for their own chip, to be built around the core that IBM was still building with Sony.

All three of the original partners had agreed that IBM would eventually sell the Cell to other clients. But it does not seem to have occurred to Sony that IBM would sell key parts of the Cell before it was complete and to Sony's primary video-game console competitor. The result was that Sony's R&D money was spent creating a component for Microsoft to use against it."

( From the WSJ )

[0] https://web.archive.org/web/20150129121858/http://www.wsj.co...

bowsamic• 2 years ago

Maybe that’s a mainland thing. In the UK the Xbox 360 was definitely far more popular

lomase3• 2 years ago

In Spain too. The PS3 released with very few games and all multiplatform games ran better on the 360.

light_hue_14• 2 years ago

If I could ask a single CEO one question, it would be Lisa Su.

What is she thinking? She's sitting on what could be the 2nd most valuable company in the world. And she's doing nothing.

AMD's GPUs are as good as NVidia's. But their drivers suck. They did all the hard work. Now they need to invest a relative pittance in fixing up their drivers and open sourcing the rest. But they won't.

Absolutely nothing about this situation makes any sense. It's a company that's sitting on gold (actually, by weight, the value of GPUs is far higher than that of solid gold bars). But... they're throwing it all away for nothing? Is she getting paid off by NVidia to do this?

AMD is the single most confusing entity out there.

JackYoustra2• 2 years ago

The other day I tried to rent a spot MI-300. I couldn't find a way to do that! Whereas fairly easy to rent an H-100. I think they're just behind a bit on their tsmc allocation! Their stock seems to think so, it's priced much more aggressively than Nvidia's.

latchkey8• 2 years ago

My company Hot Aisle is working on solving this problem.

It has nothing to do with availability or pricing. You can't rent single MI300x today cause ROCm doesn't support virtualization. PCIe passthrough doesn't work yet. AMD knows about the issue and is working on it.

We will have 8xMI300x for rent soon. Once AMD fixes their stuff, we will be able to rent out one at a time.

latchkey8• 2 years ago

And she's doing nothing.

This is totally wrong. MI300x was released Dec 2023, just a few months ago. It is a fantastic product, but not into the hands of enough people, yet.

Give it time.

light_hue_14• 2 years ago

I can tell you don't work with these cards. The MI300x is bug-ridden junk at the moment. As are all AMD cards. You cannot use them for serious compute.

djmips2• 2 years ago

You might find that this is an article that's commenting on an interview and you might be equipped to interpret it better on your own.

Here's the original interview which has more details that I found interesting.

https://stratechery.com/2024/an-interview-with-amd-ceo-lisa-...

mackman2• 2 years ago

I loved working on the cell. I remember writing my first spu code when the ps3 dev kit was a huge stainless steel box. I later went in to implement predictive physics for racing line optimization for midnight club on the spus. Was a lot of fun. Weird thing was the floating point unit had a different intermediate bit size than the ppu so for deterministic results you had to schedule threads to always be dispatched to ppu or spu deterministically. That was essential for network games where the sim had to run bit for bit the same across players.

oppositelock• 2 years ago

I'm a systems nerd, and I found working with it quite challenging, but rewarding. It's been many years, but I still remember a number of the challenges. SPE's didn't have shared memory access to RAM, so data transfer was your problem to solve as a developer, and each SPE had 256k of RAM. These things were very fast for the day, so they'd crunch through the data very quickly. We double-buffered the RAM, using about 100k for data, while simultaneously using the other 100k as a read buffer for the DMA engine.

That was the trickiest part - getting the data in and out of the thing. You had 6 SPE's available to you, 2 were reserved by the OS, and keeping them all filled was a challenge because it required nearly optimal usage of the DMA engine. Memory access was slow, something over 1000 cycles from issuing the DMA until data started coming in.

Back then, C++ was all the rage and people did their various C++ patterns, but due to the instruction size being so limited, we just hand-wrote some code to run on the SPU's which didn't match the rest of the engine, so it ended up gluing together two dissimilar codebases.

I both miss the cleverness required back then, but also don't miss the complexity. Things are so much simpler now that game consoles are basically PC's with PC-style dev tools. Also, as much as I complain about the PS3, at least it wasn't the PS2.

mackman2• 2 years ago

Yep, all valid. When I started on it we had to do everything ourselves. But by the time I did serious dev on it our engine team had already build vector/matrix libraries that worked on both ppu and spu and had a dispatcher that took care of all the double buffering for me.

teekert• 2 years ago

I recently got a 2nd hand PS3, never was a gamer. I’m so impressed with it. Racing games are so real looking and GTA5 is still the most recent version of the game. The 2 Little Big Planet were an absolute joy to play with my son.

There is still so much value in this ancient machine.

jgord• 2 years ago

The actual interview text, seems to be here : https://stratechery.com/2024/an-interview-with-amd-ceo-lisa-...

AMD CEO Lisa Su reminisces about designing the PS3's infamous Cell processor

Related Stories