[HN Gopher] Google Axion Processors - Arm-based CPUs designed fo... ___________________________________________________________________ Google Axion Processors - Arm-based CPUs designed for the data center Author : ksec Score : 187 points Date : 2024-04-09 12:12 UTC (10 hours ago) (HTM) web link (cloud.google.com) (TXT) w3m dump (cloud.google.com) | DrBazza wrote: | Buzzword soup and lots of X% better than Y or Z times something. | Any ELI5 version of this with concrete numbers and comparisons? | sapiogram wrote: | Nope, and I don't think Google will publish concrete numbers | anytime soon, if ever. | wmf wrote: | Phoronix will benchmark it at some point. | HankB99 wrote: | I wonder if TOS precludes publishing benchmark results like | some SQL database products. | | I also wonder if home users will ever be able to buy one of | these? Will the predecessor show up in the used market? | djoldman wrote: | also at techcrunch: | | https://techcrunch.com/2024/04/09/google-announces-axion-its... | | > Google did not provide any documentation to back these claims | up and, like us, you'd probably like to know more about these | chips. We asked a lot of questions, but Google politely declined | to provide any additional information. No availability dates, no | pricing, no additional technical data. Those "benchmark" results? | The company wouldn't even say which X86 instance it was comparing | Axion to. | xyst wrote: | They must be really trying to pump the Next conference | attendance numbers. | thebytefairy wrote: | Probably not, given it starts today and is already sold out. | Takennickname wrote: | Other way around probably. Team had the next conference as a | deadline and this is all they were able to come up with. | devsda wrote: | Recently there was an article on the frontpage titled | "headline driven development". | | I guess it happens more often than not. | mtmail wrote: | https://news.ycombinator.com/item?id=39891948 - That was | posted Apr/1st so authors were half joking | miohtama wrote: | Goole taking a page from Apple's playbook. | delfinom wrote: | I remember when Apple marketed swift as being xx times | faster! | | In fine print, compared to python, lul. | refulgentis wrote: | This didn't happen | tyrd12 wrote: | https://www.apple.com/in/swift/ | | Upto 8.4x faster Than python | refulgentis wrote: | Claim: "In fine print, compared to python," | | Reality: In bold call out: | | Up to 2.6X faster than Objective-C | hu3 wrote: | Not sure what your point is but Apple also claims up to | 8.4x faster than Python: | | https://i.imgur.com/CZ6vTZV.png | refulgentis wrote: | My point is we shifted from "compared it to Python in | fine print" to "compared it to Objective-C and Python in | bold focused callout". People are jaded, we can argue | that's actually the same thing, that's fine, but it is a | very different situation. | miohtama wrote: | As a Python developer, I am happy that Apple mentions | Python and acknowledge its existence. | danielbln wrote: | At least Apple really delivered with the Ax and Mx chips. | Let's see if that pans out here as well. | FdbkHb wrote: | It's google, so probably not. Their claims should always be | taken with a grain of salt. They went with their own stuff | route for the SOC on their Google Pixel phone lines and | those SOC are always much worse than the competition and | are the main reason why those phones have such terrible | battery life compared to Qualcomm powered phones. | madeofpalk wrote: | At least with Apple's Bezos charts usually there's fineprint | that'll tell you which 3 year old pc laptop they're comparing | it to. | raverbashing wrote: | Amazing | | Sounds like some middle manager will retire this even before | launch so they can replace it with Google Axion CPU or some | meaningless name change | dmitrygr wrote: | Summary: no timelines, no specifics, NOT a custom core (just | neoverse), benchmarks with no references. | jsheard wrote: | Nonetheless, Intel and AMD must be sweating as another | hyperscaler breaks away from using x86 exclusively. | mtmail wrote: | They probably already knew it was coming. Announcing it 2 | weeks before Alphabet/Google quarterly results might be | timed. | jsnell wrote: | It's obviously timed to the Google Cloud Next conference, | which starts today. | refulgentis wrote: | The Cloud Next Conference is obviously timed to the | quarterly results which is obviously timed to the ARM | processors, wait, no, the ARM processors are obviously | time to the Cloud Next Conference which is obviously | timed to the quarterly results | osnium123 wrote: | This is probably one reason why Intel is moving towards | providing foundry services. The barrier to entry for doing | chip manufacturing is higher than for designing chips now. | It's still an open question if Intel can compete with TSMC | and Samsung though. | ajross wrote: | Intel 4 is superior to anything Samsung offers, and not | nearly as far behind TSMC's 3nm density as people are lead | to believe. The "open question" is mostly about fab scaling | and the business side of foundary management. Their silicon | works just fine in the market as it stands. | chx wrote: | I read a lot about how important it is for a foundry to | be able to work with customers and this used to miss from | the company DNA of Intel. Cell library, process that sort | of thing. We shall see. | cmpxchg8b wrote: | I doubt it, Google has been using Ampere for some time. | bigcat12345678 wrote: | Right, Google invested in non x86 since 2016 afaik (I was in | the team supporting arm powerpc inside Google). At the size | of Google, it's pretty much can break from any vendors | without damaging it's core businesses | mtmail wrote: | Eight testimonials and it's clear the companies haven't been able | to test anything yet. "[company] has been a trusted partner for | customers adopting Arm-based virtual machines and an early | adopter of Arm for our own operations. We're excited about Google | Cloud's announcement of the Axion processor and plan to evaluate | it [...]' | shrubble wrote: | Downside is that GCP still has the same faults regardless of | which CPU is being used. Things like poor customer interaction, | things not working as designed etc. Switching to ARM won't solve | that. | CuriouslyC wrote: | As much as I don't trust google and their customer service is | trash, their infrastructure is mostly good and some of their | services are very aggressively priced. I wouldn't 100% buy into | GPC but a multi cloud solution that leverages some of the good | bits is definitely a top tier approach. | thebytefairy wrote: | In what ways? In my experience I've found GCP to easiest to use | redman25 wrote: | Feels like they've changed their pricing structure for | bigquery multiple times in the past couple years. They've | never turned the status page yellow or red but there have | been a few incidences where from our perspective the service | was clearly degraded. | gigatexal wrote: | yeah on the whole (data eng by day -- data eng contractor by | night) using both AWS and GCP I much prefer GCP to AWS. I | find it far more simple to use, has sane defaults, a UI that | isn't harmful to users doing clickops, etc. | | AWS gives you low level primitives to build the moon or shoot | yourself in the foot. | shrubble wrote: | It's not about easiest to use but the way in which problems | are handled. I am aware of case where something failed after | being assured it would work by GCP ... when they got someone | a GCP tech lead on the video call he started by saying it was | all 'best effort' and that it might take several days to | resolve. Ultimately it was fixed in 8 hours but that sort of | laissez-faire attitude has led to the company in question | making plans to go elsewhere. | formerly_proven wrote: | How do you do, fellow "gluing neoverse cores to synopsys IP" | hyperscalers? | cmpxchg8b wrote: | We walk among you | bgnn wrote: | Synopsys should make its on to stop all this nonsense zt this | point. | brcmthrowaway wrote: | LMAO | | Azure / AWS / Ampere silicon are all Neoverse wrappers, | amirite? | martinpw wrote: | Looks like GCP has been using ARM processors from Ampere since | mid 2022: | | https://cloud.google.com/blog/products/compute/tau-t2a-is-fi... | | So I guess this may be the end for Ampere on GCP? | xyst wrote: | I'm all for investment in less power hungry chips. Even if it's | from Google (for a short period of time. Who knows how long these | chips will be supported) | qhwudbebd wrote: | Looks like it's cloud-scam only, rather than a product one can | actually buy and own? | sofixa wrote: | That'd be like complaining AWS don't sell Graviton chips to the | public. Why would they, they're a cloud provider building their | own chip to get a competitive edge. Selling hardware is a whole | other business. | synack wrote: | All I wanna know is how it compares to AWS Graviton2/3/4 | instances. Axion needs to be cheaper, faster, or lower emissions | to be worth even considering. Everything else is just talk and | vendor lock in. | ksec wrote: | Graviton2 is based on Neoverse N1. Graviton3 is based on | Neoverse V1. | | The Graviton4, announced last year at AWS re:Invent. is based | on Neoverse V2. | | So you should expect similar performance to EC2 R8g. I say | similar because obviously there may be some difference with | clock speed and cache size. | | It terms of x86. We are expecting AMD Zen 5c with 160 Core / | 320 vCPU later this year. | sapiogram wrote: | > lower emissions | | Do you know of any cloud providers that publish actual data for | this? Preferably verifiable, but I'll take anything at this | point. | jeffbee wrote: | Google. See https://www.gstatic.com/gumdrop/sustainability/go | ogle-2023-e... and below. | synack wrote: | Google published PUE | (https://en.wikipedia.org/wiki/Power_usage_effectiveness) | numbers a while ago, I haven't seen anything that specific | from Amazon. | | It's difficult to quantify emissions because the power | generation is mixed and changes based on demand. Water | consumption should also be a factor, but there's even less | data available for that. | my123 wrote: | Why vendor lock-in? It even uses the same CPU core as | Graviton4, so it's clearly quite a fungible offering to me. | synack wrote: | By lock-in, I'm referring to my EC2 committed use savings | plan that would prevent me from considering migrating to GCP | until it expires next year, even if Google's instances are | quantifiably better. | mcmcmc wrote: | Why are you grousing about vendor lock in when you chose to | sign an extended contract to get a better deal? You locked | it in for them. | jbverschoor wrote: | I'm curious to see when Apple will migrate off AWS and run their | own datacenters on their own ARM socs | ericlewis wrote: | Xcode cloud isn't running on Apple silicon, arguably a place | where it would make tons of sense. | jbverschoor wrote: | That does not make any sense | ein0p wrote: | It's not at all a given that this would be profitable. Apple is | not paying the same prices for AWS/GCP that mere mortals do. | jbverschoor wrote: | Well they're paying minimum 30 mil per month until this year | ein0p wrote: | I think it's a heck of a lot more than that. 30 million | seems puny compared to the revenue their services | businesses generates. Though to be fair much of it probably | doesn't run on public clouds. | jbverschoor wrote: | From what I understood is that everything runs on public | clouds. They tried Microsoft, Google, and Amazon. Sooo | they should have enough experience by now. | | The contract was 1.5B over 5 years | refulgentis wrote: | Sourcing here is 2019 article about just one AWS | contract. Apple also uses Google Cloud and Azure | extensively, not just tryouts, they were one of Google | Cloud's biggest customers. They are also building their | own data centers. (TL;DR it's much more complicated than | these comments would indicate at their face) | tudorw wrote: | Working hard to avoid lock-ins, how the other half lives. | freitzkriesler2 wrote: | I wonder if this is part of their new SoC chips they've been | building to replace tensor on the pixel line. | MobiusHorizons wrote: | I would be very surprised if there is much overlap between this | server sku and any mobile sku. Mainly because Google doesn't | design the compute core, they integrate some number of cores | together with many different peripherals into an soc. Mobile | hardware will need vastly different integrations than server | hardware and the power / area optimizations will be very | different. (Mobile is heavily limited by a maximum heat | dissipation limit through the case) the reusable bits like arm | cores and pcie / network interfaces might be reusable between | designs, but many of those come from other venders like arm or | synopsis | sega_sai wrote: | What's up with physics names ? Axion, Graviton, Ampere, Electron. | I guess it sounds cool and is not trademarked... | | What's next ? -- boson, hadron, parsec ? | winwang wrote: | If I had to guess, it's because its a hardware change... like | you're doing better at implementing software on physics :P | the_panopticon wrote: | There is a lot of room in the periodic table, too, as MS showed | starting with https://www.servethehome.com/microsoft-azure- | cobalt-100-128-... | Koffiepoeder wrote: | Generic names are often used to avoid copyright and trademark | strikes. Other easy naming schemes are rivers, minerals, | cities,... | amelius wrote: | Is "Tesla" a generic name too? | tekla wrote: | Yes | ThinkBeat wrote: | With more and more big players starting production of customized | private proprietary hardware compatibility becomes increasingly | difficult. Which works out well for the big players who can | leverage it as lock ins. | | Regular people wont be able to buy the chips from Microsoft, | Google, and you only get M* chips if you buy Apple hardware. | | Good luck with the frankenMacs now. | | At the same time devices that you buy get locked into company as | well, and if that company goes out of business you are screwed. | | Like I was when Ambi Climate closed up shop and left with | e-waste. All the hardware I need is in there but I can do | anything with it. | | Or when Google decided to close down access for Lenovo Displays, | because they didnt want to support 3rd party Displays anymore. | Two more e-waste devices for me. (There might be a way to save | the Displays I just haven't got in working yet) | | Open, compatible, standardized omni purpose hardware seems to be | dying. Much more profit in lock ins | soulbadguy wrote: | PC hardware (and hardware in general) has never been | particularly open. We simply seem to move from one dominant | player to the next. I don't think AWS/GCP using custom chip for | their cloud offering changes much of the situation (well at | least before they start having weird custom instructions). | walterbell wrote: | _> you only get M* chips if you buy Apple hardware._ | | Former M* team members are now at Qualcomm and HP/Dell/Lenovo | will ship their Arm laptops later this year, including Windows | 11 and upstream Linux support. | a_wild_dandan wrote: | If they can match the battery life, then I might switch back | to a Linux laptop. I'm so sick of Apple's anti-competitive, | anti-consumer practices. | ThinkBeat wrote: | Dell and Microsoft has been shipping ARM laptops for years. | | Creating an ARM cpu is ok but will they copy the entire M* | architecture? | LettuceSand12 wrote: | Aren't all of these ARM chips? Why is compatibility such a big | issue? | astrange wrote: | It isn't a big issue. But ARM doesn't have a universal boot | loader/device discovery/etc standard like EFI/ACPI, so there | is some more work to support them. | jonmasters wrote: | Arm servers do precisely have exactly that set of standards | ((U)EFI/ACPI). See Arm SystemReady. You'll notice in the | blog linked above that it mentions Arm SystemReady-VE, | which uses those standards. | Aurornis wrote: | These are ARM processors using standard ARM instruction sets. | | I don't see any lock in here. | soulbadguy wrote: | Another interesting challenge for intel (and AMD to a lesser | extent). Between the share of compute moving to AI accelerator | (and NVIDIA), and most cloud provider having in house custom | chip, I wonder how intel will positioned themselves in the next 5 | to 10 years. | | Even if they could fab all of those chips, the margin between the | fab business and CPU design is pretty drastic. | onlyrealcuzzo wrote: | Doesn't TSMC just fab? | | They seem to be doing just fine. | ethbr1 wrote: | TSMC fabs leading node, and has consistently for several | cycles now. So its margins probably benefit from a premium. | | If Intel can make their foundry business work and keep parity | with TSMC, the net effect is that margins for leading node | compress from the increased competition. | | And there's a lot of various US fab capacity coming online in | 2024/5: https://www.tomshardware.com/news/new-us-fabs- | everything-we-... | parasense wrote: | > the net effect is that margins for leading node compress | from the increased competition. | | That is true in perfectly competetive markets, but I'm | skeptical about that idea holding true for high-end chip | nodes. | | I'm not sure there is enough competition with Intel joining | the market alongside TSMC, Samsung, and all the other | (currently) minor players in the high-end space. You might | see a cartel form instead of a competative market place, | which is a setup where the higher margin is protected. | | My best guess is the price will remain high, but the | compression will happen around process yields. You could | successfully argue that is the same as compressing margins, | but then what happens after peak yield? Before prices | compress, all the other compressable things must first | squeeze. | onlyrealcuzzo wrote: | > You might see a cartel form instead of a competative | market place, which is a setup where the higher margin is | protected. | | Wouldn't it more likely be that players just carve out | niches for themselves in the high-end space where they | DON'T compete? | | If you're Intel - it seems like a fools errand to spend | $50B to _maybe_ take some of TSMC 's customers. | | You'd probably rather spend $10B to create a new market - | which although smaller - you can dominate, and might | become a lot larger if you execute well. | ethbr1 wrote: | I figured you'd see margin compression from the major, | volume-limited buyers: e.g. Apple, Nvidia, etc. | | Paying a premium to get a quota with TSMC looks | different, if there's competitive capacity, at least for | large customers who can afford to retask their design | teams to target a different process. | | Even if only as a credible stalking horse in pricing | negotiations with TSMC. | jeffbee wrote: | I am interested in the market impact of offloads and | accelerators. In my work, I can only realistically exploit | capabilities that are common between AWS and GCP, since my | services must run in both clouds. So I am not going to do any | work to adapt my systems to GCP-specific performance features. Am | I alone in that? | wmf wrote: | They're accelerating networking and block storage. Do you use | those? | jeffbee wrote: | Of course, but even if Google makes somehow networking 50% | more efficient, I can't architect my projects around that | because I have to run the same systems in AWS (and Azure, for | that matter). | wmf wrote: | It appears to me that Google is just matching what AWS | Nitro did years ago. | jeffbee wrote: | That's the flavor of it, but they didn't give us enough | data to really think about. But the question stands | generally for accelerators and offloads. For example | would I go to a lot of trouble to exploit Intel DSA? | Absolutely not, because my software mostly runs on AMD | and ARM. ___________________________________________________________________ (page generated 2024-04-09 23:00 UTC)