Live data from Hacker News

The Ampere Altra Review: 2x 80 Cores Arm Server Performance Monster

anandtech.com

141–150 of 266 posts

Re: The Ampere Altra Review: 2x 80 Cores Arm Server Performance Monster

#141
post #91

Earlier quoted context omitted.

I think Chromebooks did ARM years before Apple, no?

Has anyone seriously used chromebooks for development work? That's the difference.

I did for about a year, though with Linux installed instead. It's not really a bad experience when most if not all of your pipeline is punted remotely anyhow.

Re: The Ampere Altra Review: 2x 80 Cores Arm Server Performance Monster

#142
post #122

Earlier quoted context omitted.

I'd be curious to see AES encryption performance compared between the same CPUs from Ampere, Intel, and AMD. In my experience, Intel have a significant performance advantage with their AES-NI implementation over AMD. I wonder what Ampere's cores can do. For some workloads, special things like AES-NI or other optimizations which only Intel (or another vendor) have can make a huge difference in real world performance w…

> I'd be curious to see AES encryption performance compared between the same CPUs from Ampere, Intel, and AMD. In my experience, Intel have a significant performance advantage with their AES-NI implementation over AMD. I wonder what Ampere's cores can do. AMD in Zen1 had 2x AES-pipelines, allowing 2x 128-bit AES operations per clocktick. Intel caught up later, and eventually also added 2x AES-pipelines of their own.…

I haven't been able to test any Zen 3 CPUs. Now I'm curious!

From benchmarks listed here: https://calomel.org/aesni_ssl_performance.html it seems that Intel had similar performance in their 2015 release of the i7-6700 as AMD's 2017 release of Ryzen 7 1800X.

Raw general purpose benchmarks I found via Google put the Ryzen 7 1800X at about twice the compute capability as an i7-6700, but their AES performance is quite even.

Re: The Ampere Altra Review: 2x 80 Cores Arm Server Performance Monster

#143
post #91

Earlier quoted context omitted.

I think Chromebooks did ARM years before Apple, no?

Yes but they were terrible. Just being ARM isn't some kind of magic dust. It has to actually be fast.

I never had an ARM one, but imagine they would be a bit slow. However, they are still a fairly wide use case on a laptop, so to speak. And that gets blurred further by tablets I guess.

The M1 is a nice product. But let's not pretend it's the first ARM device in the world. It has plenty of merits to rest on that it don't need that qualifier as justification, anyways.

Re: The Ampere Altra Review: 2x 80 Cores Arm Server Performance Monster

#144
post #2

These things look awesome, but are pretty expensive, still. I'd love to put an ARM server in my homelab, but the price range between 100$ (RPi & Co) and 10000$ (enterprise server) seems pretty empty right now :/

If you are just building a home lab server (ie. don't need a video card, well... usually), you can build a surprisingly powerful, inexpensive system. Probably in the $500-$1000 range depending on CPU choice.

Re: The Ampere Altra Review: 2x 80 Cores Arm Server Performance Monster

#145
post #75

This line is the kicker for me: "Intel’s current Xeon offering simply isn’t competitive in any way or form at this moment in time." The rest of the paragraph twists the knife more, and declares there to be an open market on server hardware for AMD and Ampere to fight it out. With application stacks being what they are and optimizations for Xeon so deeply integrated, it's unlikely that folks running private datacenter…

AMD should see the writing on the wall and start designing ARM CPUs. I think that would be very exciting because they have the engineering expertise and talent necessary to compete against Apple's M1. I don't know of any other company that can do that (besides Intel, but you know...). Additionally, there is a huge hole open right now for ARM on everything else that is not an Apple product. Apple normalized ARM on con…

It's not totally clear to me if the M1 gains are from the 5nm process, which even AMD hasn't transitioned too yet while still smoking everything Intel on 7nm, from ARM itself, or from some mix of both.

Of course, it's probably a mix, but I don't know if ARM alone has much in that mix. It does seem to me that, even if ARM held no performance benefit over x86, Apple would still decide to use it based on their use of it in their hugely successful mobile products.

Re: The Ampere Altra Review: 2x 80 Cores Arm Server Performance Monster

#147

How close are Intel to releasing anything that's a "next gen" core design and not just incremental design change or node shrink? When will Intel's "Zen2" arrive? Are there any rumors?

SemiAccurate thinks no where on the horizon. They have a good track record on these things. See this for example. https://www.semiaccurate.com/2020/11/12/intel-delays-a-mains...

No need to go by rumors. Intel themselves have announced that Ice Lake SP won't happen until mid-2021. They showed off some stuff at SC'20 last month, but since it's possible that Zen3 EPYC "Milan" will launch at the same time or even before, it's impossible to say what their competitive position will look like when they finally ship it.

Re: The Ampere Altra Review: 2x 80 Cores Arm Server Performance Monster

#148
post #17

32cores for 800$, feasible for a nice home server rig

but really, in what way would 32 cores become useful in a home server, except running a bunch of VMs that do nothing or donate money in terms of electricity to folding@home

It would chew through large compiles (e.g., kernel, BSD userspace, anything Gentoo) in no time flat. Especially if you load the thing up with RAM to use as a tmpfs...

Re: The Ampere Altra Review: 2x 80 Cores Arm Server Performance Monster

#149
post #145
post #75

Earlier quoted context omitted.

AMD should see the writing on the wall and start designing ARM CPUs. I think that would be very exciting because they have the engineering expertise and talent necessary to compete against Apple's M1. I don't know of any other company that can do that (besides Intel, but you know...). Additionally, there is a huge hole open right now for ARM on everything else that is not an Apple product. Apple normalized ARM on con…

It's not totally clear to me if the M1 gains are from the 5nm process, which even AMD hasn't transitioned too yet while still smoking everything Intel on 7nm, from ARM itself, or from some mix of both. Of course, it's probably a mix, but I don't know if ARM alone has much in that mix. It does seem to me that, even if ARM held no performance benefit over x86, Apple would still decide to use it based on their use of it…

I think M1's gains are mainly from the specific configuration they chose to build the ARMs in (after doing the necessary quantitative analyses), not from actual development of CPU tech.

Re: The Ampere Altra Review: 2x 80 Cores Arm Server Performance Monster

#150
post #136
post #75

Earlier quoted context omitted.

AMD should see the writing on the wall and start designing ARM CPUs. I think that would be very exciting because they have the engineering expertise and talent necessary to compete against Apple's M1. I don't know of any other company that can do that (besides Intel, but you know...). Additionally, there is a huge hole open right now for ARM on everything else that is not an Apple product. Apple normalized ARM on con…

AMD might already be working on an ARM design[1]. [1] https://www.techspot.com/news/87851-amd-rumored-working-arm-... .

From what I recall, the K12 core was essentially the same micro-architecture as Zen, but with ARM instruction decoders. There were some vague statements that the ISA allowed some optimizations compared to the x86 version, but we never got to see what that really means, since the K12 project was shelved to focus on x86. I wouldn't be surprised if AMD has been sitting on K12 and keeping it just updated enough so that they could revive it whenever the market becomes more interested in ARM servers.
Post reply on HN