Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Arm's Cortex X925: Reaching Desktop Performance

chipsandcheese.com

Post reply on HN
Print view View original

121–130 of 169 posts

Re: Arm's Cortex X925: Reaching Desktop Performance

#121

Post by drzaiusx11 »

ThrowawayR2 wrote: post #112
drzaiusx11 wrote: post #75

Earlier quoted context omitted.

Afaict the "desktop" target is meaningless these days. Desktops aren't really a thing anymore in the general sense are they? Only folks I know still hanging on to desktop hardware are gamers and even those I see going by the wayside with external video cards becoming more reliable. "Daily driver" is probably a better term, but everyone's daily usage patterns will vary. I could do my day job with a VT100 emulator on a…

There's a zillion office workers that have low cost mini PCs from the big OEMs on their desk. After all, all those off-lease mini PCs on eBay that are so beloved by home lab enthusiasts have to come from somewhere.

For whatever I don't really register those little hockey pucks (mac minis, NUCs, etc) the same way as the desktop tower PCs of old. A me problem for sure, but those mini device things vary _wildly_ in capabilities manufacturer to manufacturer, from full blown intel i9s to little more than headless phones running ChromeOS on an underpowered ARM. Desktops _used_ to be fairly standardized to one CPU arch, same order of magnitude RAM, ran the Windows du jour, etc. Today the landscape isn't so monotonous (and thats a good thing!)

Re: Arm's Cortex X925: Reaching Desktop Performance

#122

Post by jadbox »

exabrial wrote: post #117

Hoping someday we can get ARM System76 laptops that meet Apple M* chip performance.

For most, it doesn't need to 'meet' Apple's performance. It just needs to be competitive to general hardware of around the -the same price point- category. This is the same problematic statement I hear that a ~$1500 PC laptop just isn't as good as a ~$3000 macbook.

Re: Arm's Cortex X925: Reaching Desktop Performance

#123

Post by guerrilla »

jakogut wrote: post #120
guerrilla wrote: post #114

Earlier quoted context omitted.

Are M* chips even beating AMD anyway?

On average according to Geekbench, the M5 compared to the 9950X is ~17% faster in single thread performance and ~30% slower in multithread performance. Individual benchmarks tell the bigger picture. These two are optimized for different use cases, with Apple heavily leaning towards low latency single thread throughput with low sustained power usage. https://browser.geekbench.com/v6/cpu/compare/16833358?baseli... EDIT…

Alright, thanks. Seems like a tradeoff issue.

Re: Arm's Cortex X925: Reaching Desktop Performance

#124

Post by megous »

BTW, does anyone have some pointers to where one can find an oldish in-order Cortex-A core (like A53) in verilog RTL form? I know ARM must give this out to companies that implement ARM based SoCs for eg. purpose of validation on FPGA.

So far I've only found various M cores online. It would be fun to have something to experiment with on a cheapish FPGA like Kintex XC7-K480T, that may have enough resources for some in-order A core, and can be had for $50 or so.

Re: Arm's Cortex X925: Reaching Desktop Performance

#125

Post by adgjlsfhk1 »

megous wrote: post #124

BTW, does anyone have some pointers to where one can find an oldish in-order Cortex-A core (like A53) in verilog RTL form? I know ARM must give this out to companies that implement ARM based SoCs for eg. purpose of validation on FPGA. So far I've only found various M cores online. It would be fun to have something to experiment with on a cheapish FPGA like Kintex XC7-K480T, that may have enough resources for some in-…

You're going to have a much better time finding RiscV cores.

Re: Arm's Cortex X925: Reaching Desktop Performance

#126

Post by llm_nerd »

senko wrote: post #70
charcircuit wrote: post #63

Earlier quoted context omitted.

M4 and M5 are literally general purpose computing parts. Apple literally owns the most profitable general purpose computing platform with the iPhone.

Perhaps this was worded poorly, but the parent is referring to inability to source these processors from Apple and use them in other (non-Apple devices). As in, they don't sell you the parts, they only sell you the entire product. If you don't want the entire package, the processors alone are irrelevant.

None of us can buy or make an X925 in isolation. We can't get one and stick it in our motherboard. It has literally zero relevance to the desktop space. You can buy a DGX Spark and use it, just like you can buy a Mac Mini.

The tested machine is an nvidia GB10 which nvidia makes and sells as a whole unit and various vendors stick it in different devices to try to differentiate (although in the end they're all basically identical).

And yes, it is extremely weird for it to never mention the Apple chip, which has a little something to do with who they thank for lending them the device. The arbitrary claims for why they ignored the enormous, class-leading ARM processor in the space is not convincing.

I mean, the other claim that this is an "industry blog" and not a "consumer blog" was equally silly. It's basically for curious hobbyists. Zero industry insiders follow this to see about the core in the GB10. It's basically Anandtech.

Re: Arm's Cortex X925: Reaching Desktop Performance

#127

Post by megous »

adgjlsfhk1 wrote: post #125
megous wrote: post #124

BTW, does anyone have some pointers to where one can find an oldish in-order Cortex-A core (like A53) in verilog RTL form? I know ARM must give this out to companies that implement ARM based SoCs for eg. purpose of validation on FPGA. So far I've only found various M cores online. It would be fun to have something to experiment with on a cheapish FPGA like Kintex XC7-K480T, that may have enough resources for some in-…

You're going to have a much better time finding RiscV cores.

Yeah, I don't need help with that one. :)

Re: Arm's Cortex X925: Reaching Desktop Performance

#128

Post by ezst »

spiderfarmer wrote: post #78
ezst wrote: post #38

Earlier quoted context omitted.

Pretty simply because I don't want to use MacOS, its terrible window management, quirks and idiosyncrasies. In your comparison, my gripe wouldn't be about the hassle of finding 3rd-party compatible batteries, but about the daily handling of the Makita while knowing the DeWalt to be more ergonomic and better suited to my needs.

That’s not what a commitment is though. If I use a Makita because the battery life and resale value is twice that of a DeWalt, I wouldn’t say Makita is asking for a commitment to their ergonomics.

We clearly have different values and priorities, and just to be clear, that's perfectly fine. I haven't considered "battery life" to be a bottleneck for about a decade, which is when all my devices started to be able to last me a whole day of work. Similarly, I only change device when I must, which most often equates to "when they die", so "resale value" doesn't matter to me (and in that regard, Apple takes themselves out of the selection pool due to poor repairability and no upgradeability). My devices are tools, I care that they help me do the task at hand while stepping as little as possible in the way.

Re: Arm's Cortex X925: Reaching Desktop Performance

#129

Post by hajile »

amelius wrote: post #13
pdpi wrote: post #2

Kind of weird to see an article about high-performance ARM cores without a single reference to Apple or how this hardware compares to M4 or M5 cores.

Apple does not produce general purpose computing parts. This is an industry blog, not a consumer oriented blog.

Chips and Cheese covers Apple products in a LOT of their posts.

The real reason is probably because they are supported by patrons and can only get new equipment to review when people donate (either money or sometimes the hardware itself).

If you like what they do (as pretty much the last in-depth hardware reviewers), consider supporting them.

Re: Arm's Cortex X925: Reaching Desktop Performance

#130

Post by hajile »

rayiner wrote: post #80

ARM designs are effectively paper launches. You get these press releases saying the new ARM matches Apple and AMD, but its years before you can buy a product with it. Google Pixels that came out in the fall are still on the X4, which was introduced in 2023. At this rate, Pixel 11 will launch with X925, which is an Apple A17/M3 tier core, when Apple is on the A20: https://wccftech.com/apple-a20-and-a20-pro-all-technol…

This core was released in the MediaTek 9400 in October 2024 some 16 months ago.

The successor of x925 is C1 Ultra and even that was released 6 months ago in September 2025 with the MediaTek 9500 and GeekerWan even has a phone review they did with that chip last year.

Post reply on HN