Discussion AMD SoC Halo series GPU discussion

Page 41 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Darkmont

Junior Member
Jul 7, 2023
19
50
61
If Adroc has a million fans, then I am one of them. If Adroc has ten fans, then I am one of them. If Adroc has only one fan then that is me. If Adroc has no fans, then that means I am no longer on earth. If the world is against Adroc, then I am against the world.
 

LightningZ71

Platinum Member
Mar 10, 2017
2,104
2,544
136
Personally, I don't really have a problem with Adroc. Yes, they are a bit abrasive, but I do like that they have strong convictions. I know that they don't care, and that's fine too.

As for the topic at hand, I'm sure that, if you look hard enough, you can cherry pick a benchmark and use it to support a broad conjecture like KRK being much better than HPT, or use one particularly narrowly defined class of operations that are really only relevant in DC or DB servers and use it to hype Zen5 to the moon. KRK won't be in most general cases. It's scarcely smaller than HPT while delivering the most minimum of ST improvements and possibly MT parity while regressing in iGPU, all so they can have an NPU that qualifies for MS CoPilot+ branding.

Why would I put the previous generation tech on a new node? Easy, when the next Gen tech wasn't enough of an improvement to be worth it. Zen5 MOBILE isn't a huge improvement over Zen4. Those benchmarks are readily available. What cost more for AMD; designing another whole SOC from scratch, or taking an existing design and doing the bits that need doing to move it to a new node. Judging by the benchmarks, they only needed about 4-500 mhz of boost clock to match or beat Zen5 in mobile roughly across the board. Now, they've got cores so big that half of them had to be compact design rules to even fit 8 total, and those don't match Zen4. Yes, RDNA3.5 is an improvement over 3.0, there's no denying it, but as has been said many times, the difference isn't that big.

HPT on N3P would be a better product than KRK. Similar performance across the board, more dies per wafer, and more PCIe lanes with likely better power draw characteristics. The competition does the same thing on a regular basis. AMD renamed stuff and sells it as new for half a decade, even after they have completely abandoned all but emergency security fixes for most of it's drivers! About the only thing AMD might have to change is the NPU if they want copilot+ for it.
 

adroc_thurston

Diamond Member
Jul 2, 2023
5,438
7,609
96
you can cherry pick a benchmark and use it to support a broad conjecture like KRK being much better than HPT, or use one particularly narrowly defined class of operations that are really only relevant in DC or DB servers and use it to hype Zen5 to the moon. KRK won't be in most general cases
I don't have to, it's just better.
It's scarcely smaller than HPT while delivering the most minimum of ST improvements and possibly MT parity while regressing in iGPU
It's whatever 1t, sizeable chunk more nT and the same synthetic higher real gaming iGFX perf.
Easy, when the next Gen tech wasn't enough of an improvement to be worth it
It is worth it.
Zen5 MOBILE isn't a huge improvement over Zen4.
Strix? Yeah, it's a bit suboptimal turns out.
But it really, really wasn't inteded to be like that.
but as has been said many times, the difference isn't that big.
The difference is gigantic.
HPT on N3P would be a better product than KRK
lmao.
Similar performance across the board, more dies per wafer, and more PCIe lanes with likely better power draw characteristics.
putting older, worse IP on a newer more expensive node is something between oxymoron and schizophrenia.
Please stop.
The competition does the same thing on a regular basis.
who. what.
Tick-tock is a decade like dead. And even then ticks had some CPU increments and usually a sizeable (Gen7/Gen8, Gen10 with CNL if it ever actually worked) GFX IP bumps.
 
Reactions: madtronik

LightningZ71

Platinum Member
Mar 10, 2017
2,104
2,544
136
Most recently meteor lake refresh on Intel3. Same RTL as o.g. MTL, but now shrunken on Intel3. Sees a notable increase in fMax and better power characteristics. AMD half way did it with Rembrandt with Zen4 moving from N7 to N6, just with an updated iGPU block, then rebranded it a couple times.

The VAST majority of customers have ZERO idea what's inside of the SoC. A few know the relative performance without understanding why. The rest of us on the forums make up such a tiny portion that they don't really care about us. N3P expands the PPA triangle in every way over N4P, and that's before you incorporate fin-flex.
 

adroc_thurston

Diamond Member
Jul 2, 2023
5,438
7,609
96
Most recently meteor lake refresh on Intel3. Same RTL as o.g. MTL, but now shrunken on Intel3
No it sucks balls.
Intel isn't even fabless, same rules do not apply to them.
AMD half way did it with Rembrandt with Zen4 moving from N7 to N6, just with an updated iGPU block, then rebranded it a couple times.
Wrong, Rembrandt had massive SoC-level changes.
N3P expands the PPA triangle in every way over N4P, and that's before you incorporate fin-flex.
It's a more expensive node kiddo, you don't put old IP there.
 

Joe NYC

Platinum Member
Jun 26, 2021
2,962
4,337
106
It's a more expensive node kiddo, you don't put old IP there.

From MLID Zen 6 leaks (regarding IODs) it seems that AMD may be moving some of the high end IODs (SoC) to N3, and maybe some of the monolithic designs.

Which would imply newer IP, including Low Power cores, RDNA > 3.5, maybe new iteration of NPU.

A lot of people are looking at the gap between now and when new Zen 6s are going to start to release. That's where the speculation is coming from that it is a big gap to not release any meaningful updates.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |