• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Discussion ARM Cortex/Neoverse IP + SoCs (no custom cores) Discussion

Page 25 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Ghostsonplanets said:
Huawei own self-developed ISA. Currently deployed on smaller scale electronics. It's basically a second option in case US is successful at disallowing Huawei of developing Arm designs.
Click to expand...

Do you know if there are any docs available? I looked it up and only saw stuff from what seemed like very dubious blogs and "news" sites. One claimed it was a VLIW (and also insisted this was somehow revolutionary.)
 
SarahKerrigan said:
Do you know if there are any docs available? I looked it up and only saw stuff from what seemed like very dubious blogs and "news" sites. One claimed it was a VLIW (and also insisted this was somehow revolutionary.)
Click to expand...
I'm not sure what you found, so I'll post what I did, sorry if that's already known and I don't assume responsibility for site errors 🙂

en.rattibha.com

How was Huawei's in-house HarmonyOS kernel (Microkernel) created for HarmonyOS NEXT, it's quite simp... - Living In Harmony

How was Huawei's in-house HarmonyOS kernel (Microkernel) created for HarmonyOS NEXT, it's quite simple the REE of LiteOS RTOS lightweight kernel, that can go lo...
en.rattibha.com en.rattibha.com
compatible chipset ecosystem for ARM and future Lingxi (Loongson LoongArch model of binaries) with Kernel Abstract Layer
Click to expand...
Could it mean it's a derivative of Loongson?

LoongArch leads to this: https://docs.kernel.org/arch/loongarch/introduction.html
So it's supported by the Linux kernel. I also had hits with LLVM and gcc. At this point I became lazy and didn't look any further if there were pipeline models.

There are also mentions of LoongArch in An Instruction Inflation Analyzing Framework for Dynamic Binary Translators.
 
Nothingness said:
I'm not sure what you found, so I'll post what I did, sorry if that's already known and I don't assume responsibility for site errors 🙂

en.rattibha.com

How was Huawei's in-house HarmonyOS kernel (Microkernel) created for HarmonyOS NEXT, it's quite simp... - Living In Harmony

How was Huawei's in-house HarmonyOS kernel (Microkernel) created for HarmonyOS NEXT, it's quite simple the REE of LiteOS RTOS lightweight kernel, that can go lo...
en.rattibha.com en.rattibha.com

Could it mean it's a derivative of Loongson?

LoongArch leads to this: https://docs.kernel.org/arch/loongarch/introduction.html
So it's supported by the Linux kernel. I also had hits with LLVM and gcc. At this point I became lazy and didn't look any further if there were pipeline models.

There are also mentions of LoongArch in An Instruction Inflation Analyzing Framework for Dynamic Binary Translators.
Click to expand...

Yeah, Loongarch is the current ISA of Loongson - it looks like a de-legacied MIPS without a lot of the nastiness (delayed braches, hi/lo); from what I've seen it's not a bad ISA design. There are available manuals. The current-gen Loongson, which uses it, is also a very reasonable microarchitecture with decent performance numbers. Chipsandcheese had an overview recently with benchmarks here.

It seems like the amount of solid info around Lingxi is small - there's a awful lot of speculation on clickbaity sites that seems to extend from "Huawei filed a trademark."
 
Last edited:
Shivansps said:
As you can see there, things arent looking too good for Mali.
Click to expand...
Isn't the G610 from 3 IP generations ago?

Soon to be 4 in roughly a month.

I get that the driver is pants, but it's a bit OTT to say it's b0rked overall.

AFAIK they still have business with Mediatek, and likely Amlogic when they finally start announcing new SoC's again.

Given the recently announced ARM/Raspberry partnership I wouldn't be surprised to find a future RPi chip uses Mali instead of Broadcoms VideoCore GPU - between the people that work on that and the current Mali driver people I'm sure they could prevail eventually.
 
Another month, another X5 rumor. This one claims X5 has achieved higher IPC than Apple's A17 Pro and Qualcomm's Oryon. No update on power nor efficiency.

www.notebookcheck.net

MediaTek Dimensity 9400 IPC advantage over Snapdragon 8 Gen 4 and Apple A17 Pro leaks

MediaTek may have a winner with the Dimensity 9400 as the company's “internal testing progress is good”. The SoC is rumored to have a more impressive IPC than the competition. The latest leak also reiterates the Dimensity 9400 to be based on CPU designs from ARM.
www.notebookcheck.net www.notebookcheck.net

csm_MediaTek-Dimensity-9400-digitial-chat-station-leak_02443eeb0c.jpg


No claim which benchmark the IPC calculation was used on, unfortunately. For the sake of argument, I'll assume it's GB6.2 or close:

Qualcomm / NUVIA Oryon: ~690 pts / GHz (GB6.2 1T)
Cortex-X4: ~704 pts / GHz (GB6.2 1T)
A17 Pro: 775 pts / GHz (GB6.2 1T)
X5 (if +10.2% IPC): ~776 pts / Ghz
X5 (if +15.0% IPC): ~810 pts / Ghz

Thus, to beat the A17 Pro's GB6.2 1T IPC, the X5 would need a ~10.2% IPC uplift vs the X4 (and, if Arm is accurate with its early claims, it should be a 15% increase minimum). But, again, at such high IPC and with seemingly mediocre results from the X4 power / efficiency, we'll need MediaTek 9400 reviews to measure the X5 core(s).
 
ikjadoon said:
No claim which benchmark the IPC calculation was used on, unfortunately. For the sake of argument, I'll assume it's GB6.2 or close:

Qualcomm / NUVIA Oryon: ~690 pts / GHz (GB6.2 1T)
Cortex-X4: ~704 pts / GHz (GB6.2 1T)
A17 Pro: 775 pts / GHz (GB6.2 1T)
X5 (if +10.2% IPC): ~776 pts / Ghz
X5 (if +15.0% IPC): ~810 pts / Ghz

Thus, to beat the A17 Pro's GB6.2 1T IPC, the X5 would need a ~10.2% IPC uplift vs the X4 (and, if Arm is accurate with its early claims, it should be a 15% increase minimum). But, again, at such high IPC and with seemingly mediocre results from the X4 power / efficiency, we'll need MediaTek 9400 reviews to measure the X5 core(s).
Click to expand...
I guess it comes from this:
https://twitter.com/x/status/1784946320102539381

Geekbench link: https://browser.geekbench.com/v6/cpu/5827749

Note the entry is marked as 'Invalid'.
 
soresu said:
Isn't the G610 from 3 IP generations ago?

Soon to be 4 in roughly a month.

I get that the driver is pants, but it's a bit OTT to say it's b0rked overall.

AFAIK they still have business with Mediatek, and likely Amlogic when they finally start announcing new SoC's again.

Given the recently announced ARM/Raspberry partnership I wouldn't be surprised to find a future RPi chip uses Mali instead of Broadcoms VideoCore GPU - between the people that work on that and the current Mali driver people I'm sure they could prevail eventually.
Click to expand...
Mali G610 is 3rd gen Valhall intended for high-end, first launched in a SoC in 2023. The MC4/MP4 version on the RK3588 is not the best one as the best one is the MP6/MC6. But still.
4th and 5th gen Valhall is not widely used really. Hell, i expect 3rd gen to gain market in 2024 and 2025 due to the G310, G510 and G710 being 3rd gen. Naming is kind of a mess.
 
Shivansps said:
Mali G610 is 3rd gen Valhall intended for high-end, first launched in a SoC in 2023. The MC4/MP4 version on the RK3588 is not the best one as the best one is the MP6/MC6. But still.
4th and 5th gen Valhall is not widely used really. Hell, i expect 3rd gen to gain market in 2024 and 2025 due to the G310, G510 and G710 being 3rd gen. Naming is kind of a mess.
Click to expand...
Eh?

Isn't this Immortalis 720 GPU 2 generations on from the Mali x10 gen?

 
Nothingness said:
I guess it comes from this:
https://twitter.com/x/status/1784946320102539381

Geekbench link: https://browser.geekbench.com/v6/cpu/5827749

Note the entry is marked as 'Invalid'.
Click to expand...
If it’s 2.1GHz at that score that’s very good. The question though is what does the peak frequency look like for the phone, and then what does the performance per watt curve look like (for the whole mobo/phone like with Andrei/Geekerwan testing).

E.g if they can only get it to 3.6GHz, then they’ll have about a 2742 GB6 ST — in practice lower than that by some margin depending on timing stuff, L2/3 and memory, but let’s say 2650 @ 3.6GHz.

Okay, cool. That’s great. But what does power look like at that point? N3E is an easy + 10% boost iso-power from N4P, but they’ll need some real architectural changes and probably more cache to keep data movement lower in order to improve power. Right now on SpecInt2017 workloads the X4 is basically matching the A14 albeit at ~ 20% higher power - at least at 3.3GHz or whatever (maybe they have pushed their cores more than Apple by design).

Apple for all the talk about wide and slow has higher clocks than Android phones and is now pulling 4GHz in Macs at still low power, and I think nodes only explain so much. The A16 was doing 3.4-5GHz on plain N4, but the ST GB5 Geekerwan power was like 4.5W for a 20% perf uplift over A14 on plain N5, which is sure mainly frequency driven but with minimal node change that’s exactly why it’s impressive.

They did change to LPDDR5, and the SLC is 50% bigger than A14, and L2 is 100% bigger. Probably this is exactly how they lowered power in addition to the phydes changes from A15 to A16.

But that also goes to show how good the A14 really was.

Plain N5, 8MB of L2, 16MB SLC, LPDDR4x 4266, and yet the N4P 8 Gen 3 with plenty of SLC & L3 is just about matching it in Spec with 20% more power


I think Apple’s big L1 and humongous shared L2 + maybe some other phydes tricks are really killer as architectural decisions to reduce power. Even if Arm catch up on Perf/GHz for instance, and they might, it remains to be if they can achieve similar perf/W curves with similar process constraints.

Better than AMD/Intel? I think likely, even with similar phydes choices, yeah. But I am kinda concerned they won’t narrow this gap with Apple, and probably it will require some cache spending.
 
Probably that’s also a reason to be bullish on Oryon in 8 Gen 4. A priori we should expect it to have a better ST power curve than the Dimensity 9400 with the X5, because of the cache hierarchy and similarity to Apple.

But if Arm really changes something big or gets innovative, or QC blows it, then you know, maybe not.
 
New Switch 2 rumors suggest it will have 12 GB of ram (two 6 GB chips). Course they also seem to have convinced themselves that it will be comparable to the PS4 Pro/Series S in docked mode... so better take that with a huge grain of salt.
 
ARM announcement dates:

Cortex A76 : May 31st, 2018
Cortex A77 : May 27th, 2019
Cortex X1 : May 26th, 2020.
Cortex X2 : May 25th, 2021
Cortex X3 : June 29, 2022
Cortex X4 : May 28th, 2023

Cortex X3 announcement date is an outlier. All other were announced in the last days of May.

Which means... there is a high probability that ARM Cortex X5 will be unveiled in one of the coming days😱
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top