https://www.amd.com/en/products/adaptive-socs-and-fpgas/soc....
The demand isn't there for the RISC-V product. AMD is exploring this space[1][2] but they aren't bringing them to market because sufficient demand isn't there.
1. https://www.amd.com/en/products/software/adaptive-socs-and-f...
2. https://www.amd.com/en/products/software/adaptive-socs-and-f...
They would have to persuade MS to create Windows for RISC-V in this case.
Could be a revival but for different purposes
Would make much more sense to compare with Qualcomm trajectory here as they dominate the high end ARM SoC market.
Basically AMD missed the opportunity to be first mover on a market which is now huge with a project Apple proved to be viable three years after the planned AMD release. Any way you look at it, it seems like a major miss.
The fact that other good decisions in other segments were made at the same time doesn’t change that.
I don't think this is a fair position. It could as well be that focusing in K12 would have delayed Zen, maybe delaying it enough that it could have become irrelevant by the time it got to market.
Remember that while Zen was a good CPU, the only reason it made as much impact as it did was because it also was released in a good time (when Intel was stumbling with 10nm and releasing Skylake refresh after Skylake refresh).
Agree. AMD stock was under $2 prior to Zen. Buying was a bet that Zen would be competitive with Intel in which case the stock would come back, otherwise they were doomed. The first Zen chips were in fact competitive but beat Intel in some benchmarks and lost in others. That would have brought back competition, but who knew Intel would flounder for many more years while Zen got a nice uplift with each generation! Delaying Zen would have been bad for AMD, but in hindsight that wouldn't have mattered so long as they could stay afloat til it launched.
The thing about being broke is you may know about good opportunities but not have the resources to actually make use of them.
No man, apple basically had the power to frog march it's app devs to a new cpu arch. That absolutely would not have happened in the windows ecosystem given the amount of legacy apps and (arguably more importantly) games. For proof of this you need look no further than Itanium and windows arm
If most Intel hardware makers had gone full ARM, they would simply have lost market share. Apple customers are going to buy Apple hardware—whatever it has inside.
But of course Apple controls not just the hardware but the OS. So ya, if only Apple hardware will run your application, you are going to port to that hardware.
Apple has a massive advantage in these transitions for sure.
Microsoft's ARM transition execution has been poor.
Apple's Rosetta worked on day one.
Microsoft's Prism still has some issues, but at release its compatibility with legacy x86 software was abysmal.
Apple's first party apps and developer IDE had ARM versions ready to go on day one.
Not so for Microsoft.
Apple released early Dev Kit hardware before the retail hardware was ready to go (at very low cost).
Microsoft did not.
Apple had already switched cpus in Macs twice, it's not surprising that they could do it again, but would they have switched from Intel x86 to AMD ARM when they never used any AMD x86? Seems unlikely.
Focusing on a product that would sell on day one rather than one that would need years to build sales makes sense for a company that was struggling for relevance and continued operations.
I think Apple would have switched anyway though. They designed Apple Silicon for their mobile devices first (iPhone, iPad) which I doubt they would have made x86. The laptops and desktops are the same ISA as the iPhone (strategically).
Sure they Apple and Arm worked together but it wasn’t developed by Apple and given to Arm.
AMD cannot go and tell its customers "hey we are changing ISA, go adjust.". Their customers would run to Intel.
Apple could do that and forced its laptops to use it. Developers couldnt afford losing those users, so they adjusted.
Nobody supports the new ISA because there is no SoC and nobody makes the new SoC because there is no support. But in this case, that’s not really true because Linux support was ready.
More than forcing volumes, Apple proved it was worth it because the efficiency gains were huge. If AMD had release a SoC with numbers close to the M1 before Apple targeting the server market, they had a very good shot at it being a success and leveraging that to success in the laptop markets where Microsoft would have loved to have a partner ready to fight Apple and had to wait for Qualcomm for ages.
Anyway, I stand that looking at how the stock moved tells us nothing about if the cancellation was a good or a bad decision.
Apple proved that creating your own high end consumer SoC was doable and viable idea due to TSMC and could result in better chips due to designing them around your needs.
And which ISA they could use? X86? Hard to say, probably no. So they had RISCV and ARM
Also about Windows...
If PantherLake on 18A actually performs as good as expected, then why would anyone move to ARM on Windows when viable energy eff. cpus like lnl and ptl are available
Well yes, exactly, that’s the issue with arriving 10 years later instead of being first mover. The rest of the world doesn’t remain unmoving.
Thing is, those efficiency gains are both in hardware and software.
Will a Linux laptop running the new AMD SoC use 5 W while browsing HN like this M3 pro does?
A huge amount of Apple's competitive edge is in the "other 90%", but they don't seem to get the headlines.
Does Windows have working sleep now? I hear it's dangerous to throw a wintelmd laptop in a backpack without shutting it down.
Data centers and hosting companies are probably the biggest customers buying AMD CPUs, no?
If those companies could lower their energy and cooling costs that could be a strong incentive to offer ARM servers.
1% 3% 6% 10% 30%?
But all of this is a decade before what we are discussing here. I didn’t even remember XScale existed at Intel while writing my first comment.
When the Microelectronics Group was transferred to Intel,
that included the StrongARM Group. A month later, everybody
in the StrongARM Group had pretty much quit.
From 2:03:30 he points out that the only purpose of the DEC lawsuit was to facilitate the sale to Compaq without the microelectronics group.
I don't think AMD should be following Intel in markets outside x86. I want to see them go RISC-V with a wide vector unit. I'd like to see Intel try that too, but they're kind of busy fixing fabs right now.
SoC market is mcdonalds. its huge in the same way the soybean industry is huge. zero margin commodity.
But, don't get me wrong, I wouldn't spit on McDonalds 6 billions either and the soybean market is one of the fastest growing in the agrifood business, with huge volume traded, probably one of the most profitable commodity at the moment.
How much of Qualcomm's profit comes from providing yet another ARM chip vs. all the value-added parts they provide in the ARM SoC's, like all the radio modem stuff necessary for mobile phones?
Now that's kind of a rhetorical question, not sure a clear answer exists, at least not outside Qualcomm internal finance figures. Food for thought, though.
(That's sort of the logic behind RISC-V as well. The basic ISA and the chip that implements it is a commodity, the value comes from all the application specific extra stuff tacked on to the SoC.)
Maybe the folks at Intel just didn't want to StrongARM their competitors?
I mean Keller is talking about a decision to not pursue an ARM chip that he’d apparently been working on after(?) Zen 2 (or maybe in parallel). So AMD was already back on a good path at that point.
[0] https://web.archive.org/web/20210622032535/https://www.anand...
I believe Jim Keller is now working on RISC-V which could take the server market by storm in the next 5 years or so.
There are already RISC-V server offerings:
Look at intel's various arm or embedded offerings it keep canceling. It can't find buyers. Qualcomm and Samsung other vendors just keep eating up sales in ARM.
Now I imagine AMD sees ARM servers as the future and wants to make sure not to be left behind, on top of ARM desktop/laptop and further embedded.
I think this mostly a sign the world is now moving away from the old x86/64 system that ruled technology for so long. AMD is needs to stay competitive here.
Stuff like this, https://www.amazon.de/-/en/Microsoft-Corporation/dp/15723171...
Performance per watt is increasing due to the lithography.
Also, Devon’s paradox.
Traditionally x86 has been built powerful and power hungry and then designers scaled the chips down whereas it's the opposite for ARM.
For whatever reason, this also makes it possible to get much bigger YoY performance gains in ARM. The Apple M4 is a mature design[0] and yet a year later the M5 is CPU +15% GPU +30% memory bandwidth +28%.
The Snapdragon Elite X series is showing a similar trajectory.
So Jim Keller ended up being wrong that ISA doesn't matter. Its just that it's the people in the ISA that matter, not the silicon.
[0] its design traces all the way back to the A12 from 2018, and in some fundamental ways even to the A10 from 2016.
ISA is just ISA
Had been arm so weighted by backwards compatibility i doubt it would be so good as it is.
I really think intel/amd should draw a line somewhere around late 2000 and drop compatibility with stuff that slow down their processors.
That’s a blast from the past; native Java bytecode! Did anyone actually use that? Some J2ME phones maybe? Is there a more relevant example?
I would need some strong evidence to make me think it isn't the ISA that makes the difference.
Basically, x86 uses op caches and micro ops which reduces instruction decoder use, the decoder itself doesn't use significant power, and ARM also uses op caches and micro ops to improve performance. So there is little effective difference. Micro ops and branch prediction is where the big wins are and both ISAs use them extensively.
If the hardware is equal and the designers are equally skilled, yet one ISA consistently pulls ahead, that leads to the likely conclusion that the way the chips get designed must be different for teams using the winning ISA.
For what it's worth, the same is happening in GPU land. Infamously, the M1 Ultra GPU at 120W equals the performance of the RTX 3090 at 320W (!).
That same M1 also smoked an Intel i9.
I'm not saying the skill of the design team makes zero difference, but it's ludicrous to say that the ISA makes no difference at all.
The claims about the M1 Ultra appear to be marketing nonsense:
https://www.reddit.com/r/MachineLearning/comments/tbj4lf/d_a...
That's not true.
We will see how big improvement is it's successor panther lake in January on 18A node
>I would need some strong evidence to make me think it isn't the ISA that makes the difference.
It is like saying that Java syntax is faster than C# syntax.
Everything is about the implementation: compiler, jit, runtime, stdlib, etc
If you spent decades of effort on peformance and ghz then don't be shocked that someone who spent decades on energy eff is better in that category
Not by a long shot.
Over a decade ago, one of my college professors was an ex-intel engineer who worked on Intel's mobile chips. He was even involved in an Intel ARM chip that ultimately never launched (At least I think it never launched. It's been over a decade :D).
The old conroe processors were based on Intel's mobile chips (Yonah). Netburst didn't focus on power efficiency explicitly so and that drove Intel into a corner.
Power efficiency is core to CPU design and always has been. It's easy create a chip that consumes 300W idle. The question is really how far that efficiency is driven. And that may be your point. Lunar Lake certainly looks like Intel deciding to really put a lot of resource on improving power efficiency. But it's not the first time they did that. The Intel Atom is another decades long series which was specifically created with power in mind (the N150 is the current iteration of it).
It might be the same with x86 and power-efficiency (semantics being the issue), but there doesn’t seem to be a consensus on that.
I love the saying "i dont trust benchmarks that i didn't fake myself"
Java and C# are very similar so that analogy might make sense if you were comparing e.g. RISC-V and MIPS. But ARM and x86 are very different, so it's more like saying that Go is faster than Javascript. Which... surprise surprise it is (usually)! That's despite the investment into Javascript implementation dwarfing the investment into Go.
The efficiency came solely from the frontend which is a lot heavier on x86, and stay up longer because decoding is way more complex. The execution units were the same (at least mostly, I think, might be misremembering) so once you are past the frontend there's barely any difference in power efficiency.
... the rest is history.
Acorn won the bid to make the original BBC home computer, with a 6502-based design.
Acorn later designed their own 32-bit chip, the ARM, to try to leapfrog their competitors who were moving to the 68000 or 386, and later spun off ARM as a separate company.
Nvidia can design super clean solution fron scratch - i can bet 50$ that its gonna be more efficient in MIPS/watt
It does not matter whether you are a believer in horses for courses when it comes to ISA, or a believer in "frontend ISA does not matter because it's all translated away anyways": when buyers don't want what you have, you are out. And buyers are more like a stampeding herd than like rational actors when it comes to ISA choice. I'd see offering CPU for multiple ISA as an important hedge against the herd changing direction.
But.. ..why? Of all things, I would have expected the webcam to not be cpu-related..
Cameras used on x86-64 usually just work using that usb webcam standard driver (what is that called again? uvcvideo?). But these smartphone-land cameras don't adhere to that standard, they probably don't connect using USB. They are designed to be used with the SoC vendor's downstream fork of Android or whatever, using proprietary blobs.
For things like the automotive industry or industrial applications, it could have some sense. Most are high-margin industries ready to commit to specific architectures.
https://chipsandcheese.com/p/evaluating-the-infinity-cache-i...
ARM isn't nearly as interesting given the strides both Intel and AMD have made with low power cores.
Any scenario where SoundWave makes sense, using Zen-LP cores align better for AMD.
AMD does not have any product that can compete with Intel's N-series or industrial Atom CPUs, which are designed for power consumptions of 6 W or of 10 W and AMD never had any Zen CPU for this power range.
If the rumors about this "Sound Wave" are true, then AMD will finally begin to compete again in this range of TDP, a market that they have abandoned many years ago (since the AMD Jaguar and Puma CPUs), because all their resources were focused on designing Zen CPUs for higher TDPs.
For cheap and low-power CPUs, the expensive x86-64 instruction decoder may matter, unlike for bigger CPUs, so choosing the Aarch64 ISA may be the right decision.
Zen compact cores provide the best energy efficiency for laptops and servers, especially for computation-intensive tasks, but they are not appropriate for cheap low-power devices whose computational throughput is less important than other features. Zen compact cores are big in comparison with ARM Cortex-X4, Intel Darkmont or Qualcomm cores and their higher performance is not important for cheap low-power devices.
A cursory search shows that the AMD APU used in the Valve Steam Deck draws 3-15W. Limiting the TDP to 6W on a Steam Deck is fine for Linux in desktop mode.
“IT Home News on October 13, @Olrak29_ found that the AMD processor code-named "Sound Wave" has appeared in the customs data list, confirming the company's processor development plan beyond the x86 architecture”
I think that means they are planning to export parts.
I think there still is some speculation involved as to what those parts are, and they might export them only for their own use, but is that likely?
Apple isn’t going to switch back to AMD64 any time soon. Cloud providers will switch faster if X64 chips become really competitive again.
The limit is power capacity and quite often thermal. Newer DCs might be designed with larger thermal envelopes, however rack space is nearly meaningless once you exhaust thermal capacity of the rack/isle.
Performance within thermal envelope is a very important consideration in datacenters. If a new server offers double performance at double power it is a viable upgrade path only for DCs that have that power reserve in the first place.
EDIT: Haha, I was going off our workloads but hilariously there are some HPC-like workloads where benchmarks show the Graviton 4 smoking a 9654 https://www.phoronix.com/review/graviton4-96-core/4
I suppose ours must have been more like the rest of the benchmarks (which show the 9654 faster than the Epyc).
I've always heard it's cooling capacity. I'm also pretty confident that's true
Clearly, they want them, because there's demonstrated power savings.
Any pointers regarding that? How does the computing power to watts ratio look these days across major CPU architectures?
- Low power when only idling through events from the radio networks
- Low power and reasonable performance when classifying objects in a few video feeds.
- Higher power and performance when occasionally doing STT/TTS and inference on a small local LLM
The chip here is an interesting mix. Fast ddr5-9600! But less GPU CU's than most APUs: 4 down from 6. But if it comes with the other fixings like a good video engine & AMD's very good drivers it could be a real win.
Also a little hopeful that AMD rebadging it's Zen 1 and Zen 2 chips again might possible open up some decent low end space, but Sound Wave with more modern solutions would be a very nice to have power efficient low end.
[1]: https://learn.microsoft.com/en-us/windows/arm/arm64ec-abi
So I went out looking for an ARM-based server of equivalent strength to a Mac Mini that I could find and there's really not that much out there. There's the Qualcomm Snapdragon X Elite which is in only really one actual buyable thing (The Lenovo Ideacentre) and some vaporware Geekom or something product. But this thing doesn't have very good Linux support (it's built for ARM Windows apparently) and it's much costlier than some Apple Silicon running Asahi Linux.
So I'm eventually going to end up with some M1 Ultra Studio or an M4 Mini running Asahi Linux, which seems like such a complete inversion of the days when people would make Hackintoshes.
The fact that you have to argue with ARM about what you are allowed to do is the main reason not to use ARM. RISC-V is not about cost; it is about control. ARM suing Qualcomm to stop Elite X should be everything the industry needs to know to choose RISC-V wherever possible.
If you are going to launch a chip for yourself (like Apple did with Apple Silicon) or Amazon did with Graviton, I would choose RISC-V over ARM if starting today. That is what Tenstorrent did for their platform. I can see NVIDIA releasing their own RISC-V chip.
In the case of AMD, what are their customers asking for? Probably not RISC-V at this point (sadly). So ARM makes a lot of sense for them.
To get back to the original suggestion, replacing the ARM decoder in SoundWave with a RISC-V one, I do not know how feasible that is in practice. The entire chip is designed around the ISA, especially registers and memory model. It is not like compiling Kotlin instead of Java. Or rather, it could be like that if both ARM and RISC-V instructions were designed to compile down to the same micro-architecture (but they are not).
Why have both to run native arm64 code? Nearly anything you'd want is cross compiled/compilable (save some macOS stuff but that's more than just CPU architecture).
My understanding is that ARM chips can be more efficient? Hence them being used in phones etc.
I guess it would let you run android stuff "natively"?
Or perhaps you imagine running Blender in x64 mode and discord in the low wattage ARM chip?
Chip manufacturers need to focus on making power-efficient, high-performance workhorses. Apple figured this out first and got frustrated enough with Intel, who was more preoccupied with vendor lock-in than with doing the one thing they were supposed to do: developing best-in-class chips. The jump from x86 to M1 completely destroyed Intel’s reputation on that front. Turns out all those incremental changes over the years were them just moving deck chairs around. AMD was just tagging along and did not offer much more than them. They too got sidelined by Apple’s move. They never were much better in terms of efficiency and speed. So them now maybe getting back into ARM chips is a sign that times are changing and x86 is becoming a legacy architecture.
This shouldn’t matter. Both Apple and Microsoft have emulation capability. Apple is of course retiring theirs, but that’s more of a prioritization/locking strategy than it is for technical reasons. This is the third time they’ve pulled off emulation as a strategy to go to a new architecture: Motorola 68000 to PowerPC to x86 to ARM. Emulation has worked great for decades. It has broken the grip X86 has had on the market for four decades.
There is more to a CPU instruction set than just instruction encodings. For instance, x86 has flags which are updated (sometimes partially) by a lot of instructions, and a stronger memory model (TSO), while RISC-V has its own peculiar ideas on the result of an integer division by zero.
> What Apple has shown is that emulating x86 can actually rival or be faster than a natively running x86 chip.
AFAIK, Apple has special support in its processors for emulating x86. It has a hardware mode which emulates the x86 memory model, and IIRC also has something in hardware to help emulate the x86 flags register.
Yeah, it does. Its architects knew that it is cheaper to, when necessary, check and branch if the divisor is zero than it is to deal with exceptions.
Thus that hardware budget can instead be used for making the chip faster and more power efficient.
;-P
imho. (!)
i think this would be great!!
personally i totally understood why AMD gave up on its last attempt - the A1100 opterons - about 10 years ago in favor of the back then new ryzen architecture:
* https://en.wikipedia.org/wiki/List_of_AMD_Opteron_processors...
but what i would really like to see: an ARM soc/apu on an "open"*) (!) hardware-platform similar to the existing amd64 pc hardware.
*) "open" as in: i'm able to boot whatever (vanilla) arm64 linux-distribution or other OS i want ...
i have to add: i'm personally offended by the amount of tinkering of the firmware/boot-process which is necessary to get for example the raspberry pi 5 (or 4) to boot vanilla debian/arm64 ... ;)
br, a..z
ps. even if its a bit o.T. in this context, as a reminder a link to a slightly older article about an interview with jim keller about how ISA no longer matters that much ...
"ARM or x86? ISA Doesn’t Matter"
> * https://chipsandcheese.com/p/arm-or-x86-isa-doesnt-matter
Some people, for some strange reason, want to endlessly relitigate the old 1980'ies RISC vs CISC flamewars. Jim Kellers interview above is a good antidote for that. Yes, RISC vs CISC matters for something like a simple in-order core you might see in embedded systems. For a big OoO core, much less so.
That doesn't mean you'd end up with x86 if you'd design a clean sheet 'best practices' ISA today. Probably it would indeed look something like aarch64 or RISC-V. So certainly in that sense RISC won. But the win isn't so overwhelming that it overcomes the value of the x86 software ecosystem in the markets where x86 plays.
But, wouldn't it make more sense for amd to go into risc-v at this point of time?
128-bit LPDDR5X-9600 is about 150 GB/s, that's 50% better than an Orin NX. If they can sell these things for less than like $500 then it would be a pretty decent deal for edge inference. 16 GB is ridiculously tiny for the use case though when it's actually more like 15 in practice and the OS and other stuff then takes another two or three, leaving you with like 12 maybe. Hopefully there's a 32 GB model eventually...
For me personally I’d love it if this made it to a framework mainboard. I wouldn’t even mind the soldered memory, I understand the technical tradeoff there.
It looks like it is intended to run Windows Arm.
They will be very happy.
So far, Tenstorrent promises Ascalon devboards for 2026Q2.
Performance should be similar if not above AMD Zen2 or Apple M1.
By 2027, I expect there will be no gap left to close.
Longer term, I think the future looks bright for RISC-V. If nothing else, at least the Chinese are investing heavily into it, for obvious reasons wrt avoiding sanctions and such.