Presumably Hawk/(Arrow?) + 5070 Ti/5080 would be much faster in games (and although also very pricey) comparable in price.
For gaming, yes, but you'd be sacrificing significant performance on the CPU side, so not really a comparable configuration.
You're paying for GDDR, extra high density PCBs and more VRM.Doesn't seem like it.
CPU shouldn't really be a bottleneck tho in games.
The advantage of course is that the OEM can use (partially) busted IGP models for a decent enough discount from Intel, and can try to get AMD to do the same.
Not really, no, there's gonna be a pile of cheaper NUCs.What you are missing is that Strix Halo will only be used on very expensive devices.
Sure does,Doesn't seem like it.
Not even lmao, remember that the things like 4070m already used the generous die chop.Blackwell mobile is like 15% faster per SKU
Frankly they can SKU an 8+40 or 8+36 config for more paletable price points.AMD can price Strix Halo comfortably under 4070m laptops if they wish, just gotta get Strix Halo into a non-premium chassis first.
Not really, no, there's gonna be a pile of cheaper NUCs.
that's n3e.how it does against M4 pro later for CPU related tasks.
Yep, slap that in a mainline gaming laptop for $1200 and we are good to go.Frankly they can SKU an 8+40 or 8+36 config for more paletable price points.
NV is dead at the lower end for gaming, what are they do use WoA?
This will take up to 70 class GPUs laptops when the prices come down in the coming years.
There will be cheaper strix point devices. Strix Halo with one CCD shouldn't cost appreciably more to produce than a 4070 mobile plus CPU.
And no, they do not need to increase the margin to "compensate for the reduced volumes". The most profitable price point is the same regardless of whether they have $100 million in upfront costs or $100.
I believe the above cut down strix halo uses both ccds (& hence more power)Cut-down Strix Halo with one CCD (I presume MAX 385 and under) is not actually a 4070m competitor though if these benchmarks are anything to go by.
View attachment 117492View attachment 117493View attachment 117493
View attachment 117494
And like, AMD (and their partners) absolutely will need higher margins through the chain to justify producing/buying these versus pumping out more mainstream dies instead.
Are you saying that they're supply constrained? That could be, I doubt it, but I don't really care to argue over that.And like, AMD (and their partners) absolutely will need higher margins through the chain to justify producing/buying these versus pumping out more mainstream dies instead.
Give it the full GPU and it would be.Cut-down Strix Halo with one CCD (I presume MAX 385 and under) is not actually a 4070m competitor though if these benchmarks are anything to go by.
Which is why they should SKU 8+36/40 for gaming specifically! yeehaw.Cut-down Strix Halo with one CCD (I presume MAX 385 and under) is not actually a 4070m competitor though if these benchmarks are anything to go by.
Hmm, really? The M4 Pro when all cores arethat's n3e.
The 8050S behaves more like a 4060. More gaming results below (timestamped), though keep in mind they're 70W for Halo versus 80+70W for Ryzen + 4070/4060. We need to see if there's anything left in the tank when power goes up to match the other systems.Cut-down Strix Halo with one CCD (I presume MAX 385 and under) is not actually a 4070m competitor though if these benchmarks are anything to go by.
It's useful for nT spam in a power-limited scenario especially, which is why Turin-D is N3e rofl.It’s seems N3E is only useful to bump up 1t, at least for mobile devices.
You will not see a 5080 for less than 3000€/S for a lot of time. And not on a 13" FFPresumably Hawk/(Arrow?) + 5070 Ti/5080 would be much faster in games (and although also very pricey) comparable in price.
It's also a chopped GB203 which is helluva lot more expensive.You will not see a 5080 for less than 3000€/S for a lot of time. And not on a 13" FF
So will Medusa Halo be 24/48?It's useful for nT spam in a power-limited scenario especially, which is why Turin-D is N3e rofl.
well, yeah, same as Gator Range.So will Medusa Halo be 24/48?
Hmm, really? The M4 Pro when all cores are
used it only goes up to 3.9GHz. 300Mhz higher than M3 pro at all core load. It’s seems N3E is only useful to bump up 1t, at least for mobile devices.
Seems me it’s more than just N3E. That ~50 watt to 70watt difference is big as well the 16 cores on the AMD vs 10 P cores show's Apple’s architectural advantage is at play as well and still the difference isn’t massive in nT tasks.
I mean if node made that big of a difference you would see that Arrow lake-H would be much better than Raptor lake-h but it’s not.
And if AMD actually intended these for sale primarily to gamers the first SKU out in the wild wouldn't be a 13-inch tablet and business laptops with MBP-rivaling prices. They'd be in gaming laptops.Give it the full GPU and it would be.
Single CCD is plenty for games.