8 ms·
To be honest though building a processor these days is not exactly difficult. That is especially true if you start with someone else's ISA and they have already
by ChuckMcM 9y ago
To be honest though building a processor these days is not exactly difficult. That is especially true if you start with someone else's ISA and they have already created a GCC or LLVM back end for it. I was at Sun during the development of SPARC and as part of the Systems group we got to see a lot of the trade offs up front but these days transistors are not nearly so scarce. If you stick to 30 - 50Mhz for your first round you can simulate pretty much power on to shell prompt in a reasonable amount of time. Then there are the process specific issues that TSMC or Global Foundries or whomever will help you with translating your HDL into their most reliable node (probably 45nm or 90nm at the moment), then you'll build a test chip and it will likely do everything you want it to do and it will cost 10x what an equally powerful ARM chip will do, so you will really really want to use your chip instead of that one if you're going to design it into something. And that something has to be popular enough that you sell at least a million of them, otherwise you're going to eat a lot of Nonrecurring engineering (NRE) cost.
It is certainly possible, but it is a long game and you have to survive the early years. Go back and read the history of ARM (and Acorn), Intel (and IBM), Motorola (and Sun and Apple), and Power PC (and Cisco). Then read the history of the Z8000, the NS32032, the AMD88000, and the TI 9900. When you look at the history from where we sit today you will see that building a new CPU is the easiest thing in the world, staying alive until it is relevant is exceptionally difficult and requires quite a bit of luck in addition to good design.
I wish these guys a lot of luck, I would love to see a fully open architecture be even half as successful as ARM has been. They have to walk into this with their eyes open though, it is not going to be easy.
- bakul 9y agoAmd29000. Moto88000. And DEC alpha. But this is long due for India. Recall that the US semiconductor industry did have plenty of govt help at its inception. The Indian govt can similarly provide early support by being a customer. But still, it is going to be a tough row to hoe.
- tasty_freeze 9y ago+1 for using the archaic phrase "tough row to hoe", and for using it correctly. :-) In the few times I've heard someone say it, they say "road to hoe" which doesn't really make sense.
- draw_down 9y agoCome on
- bsder 9y ago> And DEC alpha. I really wish all these "open source" chips would just steal the DEC Alpha microarchitecture and implement it. It's been almost 20 years--even the patents on EV7 should have run out by now. It was a really clean, mostly orthogonal microarchitecture that had most of the really annoying implementation corners filed down. EV5 is a very straightforward chip and maps very well to low power implementations (see StrongARM). EV6 is pretty straightforward in modern technology nodes. And, best of all, it's documented. Compilers exist. Operating systems exist. Old hardware exists that can be used for validation and initial development. It wouldn't require you to build an entire ecosystem up front in order to bootstrap.
- bonzini 9y agoThat's what some people were doing with SuperH; I don't know how the project has progressed. https://lwn.net/Articles/647636/ https://lwn.net/Articles/647636/ [If you wanted to resurrect the Alpha, though, please do not resurrect the crazy memory consistency semantics, where you need to put a full memory barrier between r0 = load(&x); r1 = load(r0);
- wbl 9y agoHow else will I be able to bank the lower-level cache for the absolute highest in performance?
- KMag 9y agoDo you have any pointers to good documentation on the Alpha memory semantics? The memory semantics are of course so lax so that the processor doesn't need to check the computed addresses of all in-flight stores before issuing a concurrent memory operation. My understanding is that it's mostly a footgun for assembly programmers, but I'd love to see any estimates of the speed and power efficiency benefits of the lax memory semantics.
- gsmadhusudan 9y agoThis is one of the raging debates within the RISC-V group and elsewhere. Is weak memory semantics really worth it or can x86 style TSO be implemented optimally.Lot of anecdotal evidence but no conclusive experiment. As part of our work we hope to try multiple memory models in our I-class and get some definitive answers.RISC-V by the way supports ARM/PPC style weak memory model or optionally an x86 style TSO.
- pedroaraujo 9y agoThat is simply not true. Building a CPU can be "easy" if you want to build something that works. If you want to build something that is power efficient and, at the same time, quite performant, then the complexity of the problem escalates very quickly. Sticking at 30-50 MHz? What is this even supposed to mean? Digital hardware design hardly takes clock speeds into consideration, except during physical implementation in the later stages of the hardware development. TSMC or Global Foundries will help translating HDL into a reliable node? Do you even know what you are talking about? This "translation" that you refer to is called synthesis and physical implementation and it is done by the design house. The foundry only provided the logic cells their technology and they manufacture the final die that is going to ship on the product. The CPUs you mentioned were simple micro-controllers, they were not designed for heavy processing tasks, they don't even have a branch predictor which is something fundamental to have some significant performance. Nothing about your post made any shred of sense in terms of hardware design.
- inteleng 9y agoI'm not sure why pedroaraujo is being downvoted other than in response to his criticism of HN topcommenter Chuck McManis. Pedro is correct; toy CPUs , like what Chuck mentioned building, are so trivial that undergraduate students can build functioning models of them in a few days or hours. They are quite literally thousands of times less complex than the CPUs (even RISC arches) that are discussed in the article. I also concur with Pedro on the clock speed part - clock speed has absolutely nothing to do with your hardware design until you have written all the HDL, simulated its behavioral characteristics, actually synthesized the logic, completed the analog and I/O components of the design, and are actually preparing for tapeout. There's a reason only AMD and Intel make large CPUs for the consumer market, and it's not because college students are too busy. I'm going to fade into the grey background for this, but ChuckMcM's comment is yet another disappointing step toward the total calcification of this forum. It's not a commentary on the article; it simply states, lazily, and ignorantly, that whatever it is they're attempting at IIT-M, it can't be too hard. What an an incredibly condescending and blasé thing to say, especially on a public forum where what you say can't be erased. I'm actually quite glad to see another implementation of RISC-V from outside the US. The horde of nearly identical ARM cores on the market is terribly uninspiring.
- B1FF_PSUVM 9y agoAlso, MIPS, founded by the original RISC paper guys. Thirty years ago DEC was selling DECstations with MIPS R3000 CPUs, where you could enjoy the occasional program crash (and "core dump") due to memory alignment errors, under BSD 4.3 or so. https://en.wikipedia.org/wiki/MIPS_architecture_processors https://en.wikipedia.org/wiki/MIPS_architecture_processors
- tonmoy 9y agoThen why do so many companies spend so much money buying ARM cores, they have the ISA (more or less) and the compiler, why doesn’t every chip design company make their own processor? I know for a fact that some companies that had their own line of ARM competition ended up using ARM in the end
- deepnotderp 9y agoWell, first of all this is for an IoT CPU, so there will probably be less competition there and it can probably be tuned to specific niches. But I have to disagree that building a new CPU is easy. This is only true if we're talking about "toy" CPUs like some of the RISC-V implementations put out by UCB undergrads, these run at ~200MHZ clocks and with sub-1 ILP. If you want to make a high end and energy efficient CPU, it's still quite a difficult task and most novel CPU projects (e.g. the Mill, I guess the Adapteva Epiphany might count, although it's not really a CPU) don't have an existing GCC/LLVM backend.
- gumby 9y agoDepends a lot on what you are building. If you're talking about a von Neumann machine with symmetric registers -- basically a PDP-11 derivative aka "C machine" -- then you're right: you "simply" have the hard part of, you know, designing a chip, which is an art, and at any appreciable clock frequency, de facto mixes digital and analog design. But as soon as your stray off that path you're on your own (and you still have the chip design problems as well). The Mill guys have been at it 14+ years (admittedly none full time AFAICT).
- ChuckMcM 9y agoThe Mill folks are a good example of going right for the high end.
- nickpsecurity 9y agoAdapteva has nice write-ups on the challenges and a solution that worked for them in their market: http://www.adapteva.com/andreas-blog/semiconductor-economics-101/ http://www.adapteva.com/andreas-blog/semiconductor-economics... https://www.eetimes.com/author.asp?section_id=36&doc_id=1266014 https://www.eetimes.com/author.asp?section_id=36&doc_id=1266... https://www.eetimes.com/document.asp?doc_id=1330568 https://www.eetimes.com/document.asp?doc_id=1330568 There's also companies that are a cross between ASIC's and FPGA's called Structured ASIC's that can get the initial NRE down for a prototype. eASIC and Triad Semiconductor are two. eASIC has a lot of IP on it already. The last point is that operations such as DARPA show that government funding can knock out the worst of the NRE. India is using government funding. The reason we normally see things done that way still being very expensive is that those receiving the government funding are trying to maximize profit, get lots of patents, and eliminate competition that threatens that profit. An alternative for those aiming for public benefit is to work out the bugs from an initial CPU or SOC whose NRA was government backed to then release just over cost or at a decent markup. Each new wave of funding that produces new iterations soaks up the high NRE with the steady sales revenue doing the rest. Might work. OpenPITON and PULP are two things that came to mind as having potential there with one an alternative to SMP/multicore in server side and other for embedded. PULP actually used that model that I can tell. http://parallel.princeton.edu/openpiton/ http://parallel.princeton.edu/openpiton/ http://www.pulp-platform.org/ http://www.pulp-platform.org/
- zitterbewegung 9y agoI think the best way that RISC V can be successful is to market to people who want open hardware which they are doing . IoT would be a good place to market it or just electronic prototyping for Makers. The HiFive does this.