Waiting for Pussy Lake.

Gen11 is always compared with uhd/hd 620 when REALLY it's best compared with iris 640 and higher. Efficiency is probably a LOT better, but performance/EU (or whatever) really isn't that great from what I've been able to deduce. Mostly seems to have more EU's, not much else.

Just a little rant on why Intel WON'T crush AMD APU's. AMD can still add more EU's themselves (especially with 7nm)! ;)

RavenlordWaiting for Pussy Lake.

Head to your nearest springbreak destination.

Hugh MungusGen11 is always compared with uhd/hd 620 when REALLY it's best compared with iris 640 and higher. Efficiency is probably a LOT better, but performance/EU (or whatever) really isn't that great from what I've been able to deduce. Mostly seems to have more EU's, not much else.

Just a little rant on why Intel WON'T crush AMD APU's. AMD can still add more EU's themselves (especially with 7nm)! ;)

Creating a competitive space in your market to lure investors is something great, and when your stock goes up by 50% since you have a competitor puts things in perspective. But eh, who cares.

Hugh MungusGen11 is always compared with uhd/hd 620 when REALLY it's best compared with iris 640 and higher. Efficiency is probably a LOT better, but performance/EU (or whatever) really isn't that great from what I've been able to deduce. Mostly seems to have more EU's, not much else.

Just a little rant on why Intel WON'T crush AMD APU's. AMD can still add more EU's themselves (especially with 7nm)! ;)

You would need to look at resource counts for the iGPUs. Here are the main contenders with resource counts formatted in this way - SP:TMU:ROP.
Vega 8 and 11 have 512:32:8 (8 CU) and 704:44:8 (11 CU) respectively. Vega 3 has 192:12:4 (3 CU).
The common UHD 630 (GT2) has 192:24:3 (24 EU), current iteration of Iris Plus (GT3e) is 655 with 384:48:6 (48 EU).
Iris Pro 580 (GT4e) exists with 576:72:9 (72 EU) on high-end mobile-ish Skylake chips but these are rare, as far as I remember Skull Canyon was the only one with it that got some traction.

Out of widely available parts, Vega 3 would be the best (but not perfect) comparison to UHD 630. Comparing these directly is complicated by the different CPUs these are attached to but when compared they are roughly on par - not exactly due to different focus in resource balance. Search for Athlon 200GE reviews, some of these have iGPU benchmark results.

Intel's iGPU performance is fairly appropriate to the amount of execution units (and clocks) they have. AMDs iGPUs are simply far larger at this point. There is an argument to be made about Intel's iGPU using more die space but probably no real way to have a definite answer to that.

dj-electricCreating a competitive space in your market to lure investors is something great, and when your stock goes up by 50% since you have a competitor puts things in perspective. But eh, who cares.

Look at Intel stock price once they release 10 nm in quantity...

Huawei MateBook D 14 shows how great a Ryzen notebook could be.
Still, they've spared on screen, but at least they didn't cripple battery life and stay at 9+ hours, while most competitors manage to bring it down to 4.

As Intel APU will never cost much less than comparable AMD's, the only company that could suffer from Intel's iGPU buff is the greedy green.

Get a good look, boys. It's the only 10nm you'll ever see lmao

TheGuruStudGet a good look, boys. It's the only 10nm you'll ever see lmao

It's also an Atom, it's not even a proper laptop chip. Previous gen of them had TDP of max 10W. It has no AVX at all.

ncrsIt's also an Atom, it's not even a proper laptop chip. Previous gen of them had TDP of max 10W. It has no AVX at all.

Watch out, intel will be sending suits your way for exposing them lol. Remember, yields are improving and they're making good progress :roll:

I mean, is it me, or has anyone else noticed they're using stacking, b/c they literally can't make real chips on 10nm? Sure, it sounds nice and dandy for package size, but what it screams to me is that they want as small of dies as possible. And will probably use it anywhere possible (heat be damned no doubt). If 10nm didn't suck so bad, then there wouldn't be any reason to invest in stacking right now. The dies are already small. Or did they just not want to "glue" chiplets together right away and admit AMD did it right? I don't believe it's an economical solution.

ncrsIt's also an Atom, it's not even a proper laptop chip. Previous gen of them had TDP of max 10W. It has no AVX at all.

The chip supports a max of 15W. Intel or manufacturers can choose to lower that TDP to less than 15.