Support NeoGAF

[LordOfChaos]

Intel Core Ultra 7 155H Arc iGPU surpasses Radeon 780M, new leaked OpenCL result shows - VideoCardz.com

ASUS laptop with Core Ultra 7 155H and ARC iGPU surpasses Radeon 780M in Geekbench OpenCL test Intel is set to unveil the upcoming Core 100 series in three weeks. The integrated graphics department will receive one of the biggest upgrades. The Core series will soon include Intel Arc GPUs. The...

videocardz.com

Looking decent, and Intel is still in the era of rapid GPU driver improvements. With an IGP this good and Meteor Lake promising a 50% efficiency improvement over Intel last gen, this seems ripe for someone to make a handheld out of

 

[nkarafo]

Lets hope the next gen of Intel CPUs are good. Right now Intel has bad power consumption at load and AMD at idle. I'm waiting for something that's efficient at both before i upgrade.

 

[LordOfChaos]

Lets hope the next gen of Intel CPUs are good. Right now Intel has bad power consumption at load and AMD at idle. I'm waiting for something that's efficient at both before i upgrade.

I'm hopeful that this is it, with the LP-E cores on the SoC tile so it can literally turn off the whole compute tile at idle, and 50% better task efficiency

 

[twilo99]

Wouldn’t some of that be up to Windows to take full advantage tho?

 

[Celcius]

I just hope with the next generation Intel dials back the heat and power requirements so that it's easier and more comfortable using an air cooler with the top chip. (or the whole stack for that matter)

 

[LordOfChaos]

I just hope with the next generation Intel dials back the heat and power requirements so that it's easier and more comfortable using an air cooler with the top chip. (or the whole stack for that matter)

That's the whole reason for Meteor Lake, Raptor Lake performance in mobile at 50% the power, plus a much better IGP

 

[SABRE220]

I hope this is a good sign for Intel's next gen gpus, Ive honestly given up on amds gpu side they have thrown in the towel. Honestly, the only thing that can save from nvidias complete domination is a intel breakthrough or someone like apple for some reason decides to leverage Imagination Studios into the gpu market.

 

[Silver Wattle]

People are thinking this bench means more gaming perf than the 780M, they are wrong.
Also the wattage is unknown, and considering all the other benchmarks show this as between 20-28k I'd suggest people lower their expectations.

 

[LordOfChaos]

People are thinking this bench means more gaming perf than the 780M, they are wrong.
Also the wattage is unknown, and considering all the other benchmarks show this as between 20-28k I'd suggest people lower their expectations.

That's fair but it largely checks out on what we know on paper, accounting for AMD's doubling of flops but not doubling of performance, and Intel's stated 2x Perf per watt, and where they already were

It's possible the 28K ones are 15W and this is 28, which would still be a decent result

They should be pretty comparable on execution hardware, it's down to the drivers after that

I hope this is a good sign for Intel's next gen gpus, Ive honestly given up on amds gpu side they have thrown in the towel. Honestly, the only thing that can save from nvidias complete domination is a intel breakthrough or someone like apple for some reason decides to leverage Imagination Studios into the gpu market.

Yeah, the first attempt didn't light the world on fire, but I think that's fine, it already had better ray tracing and upscaling performance when using the extensions then AMD, which seems to be continuing to do everything together pumped through somewhat beefed up CUs, and I'm not sure that strategy can keep them in the game long term.

2-3 generations and Intel could be a real threat to AMD

 

[Buggy Loop]

Will be interesting if for next gen consoles, someone seeks an advantage over a certain competitor and the best way to do it is probably to go with different APU supplier.. as otherwise you’re almost on the same baseline.

 

[winjer]

LOL, judging performance of a CPU or GPU with Geekbench.
When will people learn that synthetic benchmarks does not reflect real world performance.

Took us years, for people and reviewers to move away from 3Dmark and focus on real games.
Now we are going into the same mistake, but with Geekbench.
We already saw this non-sense with Qualcomm vs Apple, just a few weeks ago.....

 

[Rickyiez]

Performance at low wattage is extremely important. Like for example what is the point for it to be faster than 780m at 50 watt

 

[Hudo]

No one I know or I have ever talked to uses OpenCL.

 

[LordOfChaos]

No one I know or I have ever talked to uses OpenCL.

LOL, judging performance of a CPU or GPU with Geekbench.
When will people learn that synthetic benchmarks does not reflect real world performance.

Took us years, for people and reviewers to move away from 3Dmark and focus on real games.
Now we are going into the same mistake, but with Geekbench.
We already saw this non-sense with Qualcomm vs Apple, just a few weeks ago.....

There's a degree of implication that anyone said this is the be all end all of benchmarks and performance tests here, I'd assure you I and probably no one here thought that.

Generally there's a proportional scaling up with the added efficiency of Vulkan, some chips scale better, some less, but if this test is showing it in 780M/M2 ballpark, it would generally mean it would stay in the ballpark with Vulkan as well

It's just the benchmark that leaked out, so it's the news right now. Everything else has yet to be seen. But we've also seen them being damn serious with improving their dedicated card drivers which should carry over to this.

 

[Hudo]

There's a degree of implication that anyone said this is the be all end all of benchmarks and performance tests here, I'd assure you I and probably no one here thought that.

Generally there's a proportional scaling up with the added efficiency of Vulkan, some chips scale better, some less, but if this test is showing it in 780M/M2 ballpark, it would generally mean it would stay in the ballpark with Vulkan as well

It's just the benchmark that leaked out, so it's the news right now. Everything else has yet to be seen. But we've also seen them being damn serious with improving their dedicated card drivers which should carry over to this.

Not questioning Intel's intent itself, I sure hope so that they finally fucking get their head out of their ass in the dedicated consumer CPU market, since what they're trying to sell to consumers for years now is overpriced, repolished shit.

It's just that I'll still hold out until more leaks or official benchmarks appear since OpenCL is not the best measure for performance, especially for CPUs. Maybe OpenMP would be the better library to use. Oh well. It was not aimed at you or the news. It's just weird, that's all.

On that note, I sure do wish that AMD and Intel would team up and invest heavily into OpenCL so that we might have a serious competitor to CUDA. Because that shit keeps me (and others) locked into NVidia's bullshit. Not a fan of having to pay 2000€ for cards that should cost 800€.

 

[LordOfChaos]

It's just that I'll still hold out until more leaks or official benchmarks appear since OpenCL is not the best measure for performance, especially for CPUs. Maybe OpenMP would be the better library to use. Oh well. It was not aimed at you or the news. It's just weird, that's all.

This is an IGP benchmark, for the integrated graphics chip

14th gen Meteor Lake on laptop is very different from the 14th gen rewarming (literally) on desktop, it's a radical departure in architecture worth looking into