Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)

[DisEnchantment]

Speculate at will

 

[Ghostsonplanets]

It's just a cheaper strix.

He probably means in term of price target and battery life. Granted, I expect Lunar Lake to be priced very high due to N3B and tiled approach. Intel will probably ask outrageous prices from OEMs to mandate that LNL is to be used on premium T&L devices, which to me is the same error they did with Core M/Core Y.

So AMD will have the advantage here with KRK being slotted into <$1000(?) machines while still offering competitive battery life and better performance.

But LNL should be the undisputed x86 battery champion next year, even if Intel isn't as competitive in performance. Otherwise, LNL as a whole would a huge waste.

 

[adroc_thurston]

He probably means in term of price target and battery life

Well no.
Not a premium tablet part, but a mainstream Strix.

Granted, I expect Lunar Lake to be priced very high due to N3B and tiled approach

And MoP. That's a direct cost adder.

Intel will probably ask outrageous prices from OEMs to mandate that LNL is to be used on premium T&L devices

not anymore.

still offering competitive battery life

LNL should still be more efficient in that case.
Hopefully. lmao.

Otherwise, LNL as a whole would a huge waste.

it's a product of many new's for Intel so worth it in itself.

 

[Thunder 57]

For such a cache/bandwidth starved implementation of RDNA, that's really good. AMD basically giving an integrated RTX 3050 35 - 50W with every STX SoC, but without the VRAM bottleneck.

In theory, OEMs could create T&L entry level Gaming Laptops with it. Or users could simply buy an office/premium T&L laptop and still be able to game at 1080p Low/Medium with reasonable settings.

I must be getting old. You keep saying T&L and I keep reading it as "Transform & Lighting". Hardware T&L was a big deal back in the day.

Then I realized you meant Thin & Light.

 

[biostud]

I must be getting old. You keep saying T&L and I keep reading it as "Transform & Lighting". Hardware T&L was a big deal back in the day.

Then I realized you meant Thin & Light.

yup

Full Review NVIDIA's new GeForce256 'GPU'

NVIDIA's TreeMark is an OpenGL-benchmark that uses a very high polygon count. You've certainly heard of it, it's simply a scene with a tree that is very detailed. The camera moves around the tree clos

www.tomshardware.com

 

[Ghostsonplanets]

Well no.
Not a premium tablet part, but a mainstream Strix.

And MoP. That's a direct cost adder.

So I'm assuming what you mean is that LNL is targeted at >$1000 premium devices while KRK is targeting <$1000 mainstream devices. So very different markets.

LNL should still be more efficient in that case.
Hopefully. lmao.

Aye. But if KRK can offer similar or close to battery life while offering nearly 2x CPU performance and a better and more mature GFX solution while at much lower price points, it makes LNL much less appealing in my view.

Granted, LNL should enable and be offered in novel products designs and will also target a much lower power curve. So might as well be apples to oranges.

it's a product of many new's for Intel so worth it in itself

Right. Special products like Lakefield or LNL still offer much learnings that can be re-used in the future, even if they themselves fail. They're worthwhile experiences.

I must be getting old. You keep saying T&L and I keep reading it as "Transform & Lighting". Hardware T&L was a big deal back in the day.

Then I realized you meant Thin & Light.

Transform and Lighting Laptops would be a very rad name 😁

 

[tamz_msc]

MTL has 8 Xe-LPG cores, not 4!
What we should expect from Lunar Lake IGP I don't know.
But I am not even sure, If AMD will have a direct competitor against It, maybe Kraken Point? 4+4 and 8CU looks pretty similar.

My bad. I forgot that LNL is a replacement for MTL-U which has 4 Xe cores.

Have you seen the SiSoft Leak, comparing full MTL 8 Xe core vs LNL 8 Xe2 cores?

MTL-U 8 Xe cores:



LNL 8 Xe2 cores:



Almost 3x compute vs full-fat MTL-H.

 

[tamz_msc]

???

No, you're underestimating how arse MTL is. LNL being as good as it is at 15W relative to MTL only puts it on par with PHX at 15W.

MTL is utterly unusable at 15W, it consistently falls far short of Phoenix on average frame rates and has utterly unplayable frame times to boot at 25W. It just gets even worse at 15W.

That "review" is more about MSI screwing up the firmware than the incompetence of Arc.

On Notebookcheck the MSI prestige Evo 16 needs 110 W boost power to achieve the same TS score as a Lenovo Yoga 7 with 50 W boost power.

 

[adroc_thurston]

That "review" is more about MSI screwing up the firmware than the incompetence of Arc.

no, MTL-H just falls apart below 30w.
you're welcome.

 

[PJVol]

AMD has most definitely decided. They are just holding their cards close to their vest. Almost all of these decisions/changes occur internally a year a two b/4 they are leaked. Even board partners are getting specs far later than they did 10 years ago. Such is the state of a very competitive market.

Obviously not yet, every Zen DT dev.cycle involves two separate designs (as well as two masks, two MB prototypes, dev teams, etc.) for the current and future platform, and when the time comes (usually couple of days before launch) just decide which one to throw to hell

 

[tamz_msc]

no, MTL-H just falls apart below 30w.
you're welcome.

Is that why Asus is able to make a 14 inch laptop with the 155H and give it 17% faster memory than the MSI 13 inch model and score 1.7x the MSI at 64W/20W PL2/PL1 for both?

 

[Hitman928]

Is that why Asus is able to make a 14 inch laptop with the 155H and give it 17% faster memory than the MSI 13 inch model and score 1.7x the MSI at 64W/20W PL2/PL1 for both?

Do you have a link to said results?

 

[adroc_thurston]

Is that why Asus is able to make a 14 inch laptop with the 155H and give it 17% faster memory than the MSI 13 inch model and score 1.7x the MSI at 64W/20W PL2/PL1 for both?

no it's just that 15/20W MTL-H gfx results are bad.
can you not?

 

[tamz_msc]

Do you have a link to said results?

Sure:

https://www.notebookcheck.net/Intel-Arc-8-Cores-iGPU-Benchmarks-and-Specs.782930.0.html

 

[Hitman928]

Sure:

https://www.notebookcheck.net/Intel-Arc-8-Cores-iGPU-Benchmarks-and-Specs.782930.0.html

Can you be a little more specific than an article with a score of benchmarks across multiple models?

 

[tamz_msc]

no it's just that 15/20W MTL-H gfx results are bad.
can you not?

Yes, Time Spy takes long enough to run that eventually package power will stabilize at 20 W if you've set PL1 at 20 W.

Looks to me like you've never used a recent Intel iGPU in a laptop before.

 

[adroc_thurston]

Time Spy takes long enough to run that eventually package power will stabilize at 20 W if you've set PL1 at 20 W.

I'm not even talking timespy.

Looks to me like you've never used a recent Intel iGPU in a laptop before.

Alder-U was pretty awful so hopefully never again.

 

[uzzi38]

That "review" is more about MSI screwing up the firmware than the incompetence of Arc.

On Notebookcheck the MSI prestige Evo 16 needs 110 W boost power to achieve the same TS score as a Lenovo Yoga 7 with 50 W boost power.

You're not seriously trying to suggest that in any situation a standard laptop chip would see significant gains to iGPU performance above 50W, are you?

That's well, well, well beyond the region of diminishing returns. So common sense would state that something is clearly more wrong than you're letting on.

After a bit of googling later to find the article you conveniently forgot to post, yeah obviously it didn't make sense. In The Witcher 3 this laptop doesn't peak above 70W. As an actual game, TW3 actually puts a load on the CPU as well, TS doesn't really do that well. I've no doubt during the TS test this device was actually pulling less than 70W.

Anyway, performance at low power is something Intel acknowledged was an issue very close to launch, and they said they'd release firmware updates to fix it. They still haven't done so, and this isn't on MSi to fix. There's nothing for them to do.

 

[tamz_msc]

Can you be a little more specific than an article with a score of benchmarks across multiple models?

MSI prestige 13 AI Evo scores 2178, while the Asus Zenbook 14 scores 3427.

 

[gdansk]

Move MTL junk to MTL thread

 

[tamz_msc]

You're not seriously trying to suggest that in any situation a standard laptop chip would see significant gains to iGPU performance above 50W, are you?

That's well, well, well beyond the region of diminishing returns. So common sense would state that something is clearly more wrong than you're letting on.

After a bit of googling later to find the article you conveniently forgot to post, yeah obviously it didn't make sense. In The Witcher 3 this laptop doesn't peak above 70W. As an actual game, TW3 actually puts a load on the CPU as well, TS doesn't really do that well. I've no doubt during the TS test this device was actually pulling less than 70W.

Anyway, performance at low power is something Intel acknowledged was an issue very close to launch, and they said they'd release firmware updates to fix it. They still haven't done so, and this isn't on MSi to fix. There's nothing for them to do.

The MSI handheld having performance issues and the MSI laptop needing twice the power to achieve the same benchmark results as other OEMs doesn't make you pause to think that MSI might be the culprit?

 

[Abwx]

MSI prestige 13 AI Evo scores 2178, while the Asus Zenbook 14 scores 3427.

Scores at barely 25W for the MSI and at 45W for the Asus, so the latter get roughly 2x the power for the GPU, this can be seen in the power measurements with Furmark, so you re left one more time using gross lies to prop up your intelmania.

https://www.notebookcheck.net/Asus-Zenbook-14-OLED-review-The-1-2-kg-subnotebook-with-120-Hz-OLED-and-Core-Ultra-7.805236.0.html

https://www.notebookcheck.net/MSI-Prestige-13-AI-Evo-laptop-review-Core-Ultra-7-and-OLED-weighing-under-1-kg.810670.0.html

Now if we can get back to the topic subject and stop all this thread trashing with your
usual fairy tales.

 

[tamz_msc]

Scores at barely 25W for the MSI and at 45W for the Asus, so the latter get roughly 2x the power for the GPU, this can be seen in the power measurements with Furmark, so you re left one more time using gross lies to prop up your intelmania.

https://www.notebookcheck.net/Asus-Zenbook-14-OLED-review-The-1-2-kg-subnotebook-with-120-Hz-OLED-and-Core-Ultra-7.805236.0.html

https://www.notebookcheck.net/MSI-Prestige-13-AI-Evo-laptop-review-Core-Ultra-7-and-OLED-weighing-under-1-kg.810670.0.html

Now if we can get back to the topic subject and stop all this thread trashing with your
usual fairy tales.

You'll need to check with an optometrist - your eyes are failing - when you cannot see that it is clearly stated in the reviews what power levels the models run at.

 

[Hitman928]

MSI prestige 13 AI Evo scores 2178, while the Asus Zenbook 14 scores 3427.

Thanks, but is it really that hard to see why the Asus scores much higher? Just because they have the same PL2/PL1 setting, doesn't mean they have the same power consumption.



There's a huge difference with how long each is able to maintain higher power levels but once they both settle to 20 W, they get the same performance.





So yeah, nothing unusual here, MSI's just configuring for more power savings with a very short turbo window and a steep roll off versus a soft roll off on the Asus. This won't change anything in a completely power constrained situation though, like with the MSI Claw.

 

[tamz_msc]

Thanks, but is it really that hard to see why the Asus scores much higher? Just because they have the same PL2/PL1 setting, doesn't mean they have the same power consumption.

That's total power consumption you are trying to show in the table.

So yeah, nothing unusual here, MSI's just configuring for more power savings with a very short turbo window. This won't change anything in a completely power constrained situation though, like with the MSI Claw.

If your firmware power management is whack at higher power, chances are that it might be the same at lower power.

Which is my interpretation of the MSI claw results.

 

[adroc_thurston]

If your firmware power management is whack at higher power, chances are that it might be the same at lower power.

no, it just sucks at low watts.

Which is my interpretation of the MSI claw results.

please stop the cope.
you only have 2 months left, then you're waiting all the way till 2026 for a chance at relevance.
cheers!