Question AMD Phoenix/Zen 4 APU Speculation and Discussion

Page 52 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Sort by date Sort by votes
tamz_msc

tamz_msc

Diamond Member
Jan 5, 2017
3,828
3,661
136
TESKATLIPOKA said:
That RAM speed is only 4800 MHz, that's not enough even for Rembrandt!
The highest gain is in Assassin's Creed Valhalla +23%.
The lowest gain is in Shadow of the Tomb Rider +9%.
Give them 6400MHz LPDDR5 with 33% higher BW, then we can talk how good or bad 780M is.

BTW, I wonder If this is really a proper comparison.
We know memory is slow and likely with bad timings, but 680M has only 2200MHz and 780M has 2800MHz clock speed.
I don't think they have the same TDP limit.
Click to expand...
Doesn't matter how fast the RAM is. The pre-announcement Chinese review had LPDDR5-6400 on both systems, and the 780M was 12% faster.

TDP is irrelevant. Both the -H and -HS in laptops like the Asus TUF will happily boost to 60-70 W in their highest power mode.

This is real-world comparison using similar type of systems that you can buy right now.
 
  • Like
Reactions: insertcarehere
Upvote 1 Downvote
T

TESKATLIPOKA

Platinum Member
May 1, 2020
2,373
2,868
136
tamz_msc said:
Doesn't matter how fast the RAM is. The pre-announcement Chinese review had LPDDR5-6400 on both systems, and the 780M was 12% faster.

TDP is irrelevant. Both the -H and -HS in laptops like the Asus TUF will happily boost to 60-70 W in their highest power mode.

This is real-world comparison using similar type of systems that you can buy right now.
Click to expand...
RAM speed matters, stop saying It doesn't.
TDP is also relevant, Rembrandt can also consume 60-70W.

680M which boosts to 2400MHz was limited to 2200MHz, which should mean It had a limited TDP while 780M didn't because boosted to 2800MHz.
And this is supposedly a proper comparison in your opinion?
Give them 6400MHz RAM and either unlimited TDP or limit It to same TDP for both of them. That's a proper comparison.
 
Last edited:
  • Like
Reactions: Tlh97
Upvote 0 Downvote
tamz_msc

tamz_msc

Diamond Member
Jan 5, 2017
3,828
3,661
136
TESKATLIPOKA said:
RAM speed matters, stop saying It doesn't.
TDP is also relevant, Rembrandt can also consume 60-70W.

680M which boosts to 2400MHz was limited to 2200MHz, which should mean It had a limited TDP while 780M didn't because boosted to 2800MHz.
And this is supposedly a proper comparison?
Give them 6400MHz RAM and either unlimited TDP or limit It to same TDP for both of them. That's a proper comparison.
Click to expand...
What I meant was faster RAM benefits both the 680M and 780M. So giving them both faster RAM would end up giving the same net result.

680M boost frequency depends on the SKU. 2400 MHz is for the -HX CPUs IIRC. It is 2400 MHz on the 6900 SKUs.

Both systems are boosting to 60-70W. So it is as close to an apples to apples comparison as it can get.
 
  • Like
Reactions: Mopetar
Upvote 0 Downvote
T

TESKATLIPOKA

Platinum Member
May 1, 2020
2,373
2,868
136
tamz_msc said:
What I meant was faster RAM benefits both the 680M and 780M. So giving them both faster RAM would end up giving the same net result.

680M boost frequency depends on the SKU. 2400 MHz is for the -HX CPUs IIRC.

Both systems are boosting to 60-70W. So it is as close to an apples to apples comparison as it can get.
Click to expand...
Ok, I didn't know R7 6800H is limited to 2200 MHz, but then HX would be a better comparison.

Faster RAM benefits both of them, true, but giving them the same faster RAM shouldn't net the same improvement.
Let's say 2200MHz 680M paired with 5600MHz is not bottlenecked, but then 780M clocked 27% higher would be bottlenecked even with 6400MHz.
Guess the LPDDR5x-7500 support is not just for show.

Don't take me wrong, I am also disappointed with this level of performance, It's not even as fast as my 1650M, so buying Asus ROG Ally is pointless for me, but I am still interested in how much performance is lost due to this BW bottleneck.

edit:
I don't understand how Strix Point can supposedly have 16CU(+33%) but no Infinity cache based on leaks. This won't perform like 35W 3050M.
 
Last edited:
  • Like
Reactions: Tlh97 and Kryohi
Upvote 0 Downvote
L

leoneazzurro

Senior member
Jul 26, 2016
952
1,516
136
mikk said:
Extremely limited availability after months and expensive. And it's Asus only as usual, I call it paper work. It's the usual AMD thing, get 1-2 showcase models out 3-4 months after launch (usually Asus exclusive) and people like you are happy. The next devices come out 6+ months after launch , and then rinse and repeat. At CES next January AMD will announce or launch whatever the next paper work and only then we see a reasonable low volume from the last generation, see Rembrand!
Click to expand...
Uhh.. limited availability in the number of models yes, in quantities I would not say so because the stock is always full everytime I look for it. But for sure this is retail so it has limited impact and we will see something more substantial once the other OEMs roll out their models, which were showcased anyway.
 
Upvote 0 Downvote
SteinFG

SteinFG

Senior member
Dec 29, 2021
458
521
106
Triskain said:
According to recently posted AGESA changelogs there is something called "PHX2 AM5". I thought Phoenix2 is the one with the 2xZen 4 + 4xZen 4c CCX Setup? Any idea if/what APUs will come to the Desktop?

(Excuse the off topic post...)
Click to expand...

If Phoenix 2 comes to desktop (6 cores, 16MB L3, 4 CUs), it only makes sense if it's cheap.
With its IPC and Clocks and Cache, it'll probably match R5 5600. So it should cost 139 at most.
I do think it costs AMD about the same to do a 4nm 137mm² die compared to 7nm 80.7mm² + 12nm 125mm²
After AMD releases PHX2 on desktop, they'll probably stop making 5600 and any higher tier processors for AM4.
Not sure when it will happen though.
For example, Renoir (Zen 2 APU) got released in January 2020, but the desktop chips got to the DIY market 2 years later.
 
Upvote 0 Downvote
K

Kaluan

Senior member
Jan 4, 2022
500
1,071
96
TESKATLIPOKA said:
That RAM speed is only 4800 MHz, that's not enough even for Rembrandt!
The highest gain is in Assassin's Creed Valhalla +23%.
The lowest gain is in Shadow of the Tomb Rider +9%.
Give them 6400MHz LPDDR5 with 33% higher BW, then we can talk how good or bad 780M is.

BTW, I wonder If this is really a proper comparison.
We know memory is slow and likely with bad timings, but 680M has only 2200MHz and 780M has 2800MHz clock speed.
I don't think they have the same TDP limit.
Click to expand...
AFAIK Phoenix also supports LPDDR5X-7500, some announced handhelds seemingly come with it, no idea on ultraportable laptops yet tho.

Judging by the frametime graph in those tests, which look way more spiky than 680M/RDNA2's.... 780M/iGPU RDNA3 is clearly still on early drivers in all these tests.

Also NBC's review is laughable. Only interesting tidbit is the Port Royal test (hybrid/ray tracing), where it's over 50% faster than last gen.

Seem the wait for fully fleshed out drivers and reviews of Phoenix IGP is still ongoing.


Here's 680M w/ DDR5-4800 vs LPDDR5-6400:

v2-11589ecdc9f013fbf79ef2d5476d8e39_r.jpg
v2-45c8f9ab314d2f40ea89b648b4cae198_r.jpg


And we're supposed to believe it matters LESS for 780M, when it's not even using natively supported 5600 in all of these so called reviews, let alone LPDDR5 or 5X. lol
 
Last edited:
  • Like
Reactions: Tlh97, lightmanek, igor_kavinski and 1 other person
Upvote -1 Downvote
I

insertcarehere

Senior member
Jan 17, 2013
639
607
136
All this goalpost moving about RAM speeds when somebody has done a review already comparing to 680m with the same RAM, tested bandwidth scaling, and power scaling to boot..
insertcarehere said:
Radeon 780m review here:

With both systems running LPDDR5-6400, 780m/7840HS is around 12-13% faster than 680m/7735HS:
View attachment 80142
View attachment 80143
LPDDR5-6400 looks to be adequate enough for 780m, not a ton of difference moving to LPDDR5-7500
View attachment 80144

Quoting the reviewer: "Instead of calling the iGPU RDNA3, might as well call it RDNA2.5".
Click to expand...
 
  • Like
Reactions: Mopetar
Upvote 0 Downvote
DrMrLordX

DrMrLordX

Lifer
Apr 27, 2000
21,709
10,984
136
AMD should be able to get as many wafers as they like of N5-family nodes. If AMD wants more Phoenix/Phoenix2, they should be able to get it. The question is, how much do they really want to bring to market right now?
 
Upvote 0 Downvote
tamz_msc

tamz_msc

Diamond Member
Jan 5, 2017
3,828
3,661
136
Also, check this:

1683305231430.png

And this:

1683305327725.png

No bandwidth advantage is going to overcome a 1 GHz GPU clock speed deficit and a 1.3 GHz CPU clock speed deficit to give you the same FPS.
 
Upvote -1 Downvote
ryanjagtap

ryanjagtap

Member
Sep 25, 2021
109
130
96
tamz_msc said:
Something is off. RDR2 was showing >60 FPS in the frame rate counter but gameplay was obviously sub 20 FPS.

For now I'm not going to rely on testing by this guy.
Click to expand...
Well we should never really rely on a single tester, but the memory is also quad channel so that can also be doing something.
 
Upvote 0 Downvote
ryanjagtap

ryanjagtap

Member
Sep 25, 2021
109
130
96
tamz_msc said:
Something is off. RDR2 was showing >60 FPS in the frame rate counter but gameplay was obviously sub 20 FPS.
Click to expand...
Here's another reviewer testing the 7840U on the GPD Win Max 2. It's just RPCS3 PS3 emulation testing. He gets different FPS on RDR2, so something is really not right on ETA Prime's review.

 
Upvote -1 Downvote
SteinFG

SteinFG

Senior member
Dec 29, 2021
458
521
106
tamz_msc said:
Something is off. RDR2 was showing >60 FPS in the frame rate counter but gameplay was obviously sub 20 FPS.

For now I'm not going to rely on testing by this guy.
Click to expand...
I see 60+ FPS.

You can see it for yourself by using . and , buttons while youtube video is paused. Those allow you to go frame-by-frame in any youtube video. And in this 1080p60fps youtube video, when the character is moving, each youtube frame shows a diferent frame on the device, meaning it's at least matches youtube FPS (60).

It simply can't be 20 or 30 FPS, otherwise we'd see repeated frames in the video.
 
Last edited:
  • Like
Reactions: Kaluan
Upvote 0 Downvote
T

TESKATLIPOKA

Platinum Member
May 1, 2020
2,373
2,868
136
VideoCardz: AOKZEO A1 Pro
AOKZOE-RADEON-780M.jpg

Time Spy and Fire Strike colours are mixed up.
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom