Qualcomm benchmarking controversy: What's happening?
Qualcomm benchmarking controversy: What's happening?
Right now, Qualcomm is facing some controversy surrounding its benchmarking claims for Snapdragon X series chips. Here's what's happening.
Both reported numbers that were nowhere close to what Qualcomm promised. How not close? Above 50% this time but one used the term “Celeron” to describe performance.
There is no harsher way to describe the performance of a CPU. Ouch.
"Atom" is one step further beyond
I remember saying months ago that Qualcomm's offerings often look excellent in synthetic workloads but fall apart in real-world usecases, and I got downvoted to oblivion and called an x86 shill for it.
I never found it likely that Qualcomm would be able to compete with AMD, Apple, or even Intel in the short term.
Wow, how disappointing. One of the quotes say they got less than half as much performance as was promised, I'm hoping it's a driver issue and not just outright scam marketing.
Yeah, it sounds like Qualcomm is driving straight at a cliff's edge in terms of their reputation..
We must wait until the dust settles and real products are released. Then, and only then, can we be certain. But it does not look promising.
If this is true it will be devastating for Qualcomm ... they hyped the new chips big time, and some huge manufacturers happily announced products with the new chip. So if this falls flat, a few influential players will not be happy at all...
Not to mention the rest of the tech media... they will happily sell the pitchforks and torches.
It’s been known for a long time that they cheat on benchmarks
https://www.anandtech.com/show/7384/state-of-cheating-in-android-benchmarks
They specifically detect benchmarking tools and run at a frequency they can’t maintain for the length of the benchmark
They lack competition so unfortunately it won’t affect them too much
I really can't imagine this going well for Qualcomm, if the article is true.
It smells like there's a lot of smoke around their implementation of the benchmarks, and consumers will immediately discover the fire when the time arrives and devices are in-hand.
I don't think Qualcomm can outperform Apple at CPU game out of the blue. They were always trailing behind and Apple has now three generations of M CPUs behind. It'd be nice, though.
Edit: fixed stupid typo - can't outperform -> can outperform
can’t outperform Apple at CPU game out of the blue.
To be fair their first ARM release was 2007. They’ve been at it for almost 2 decades.
Damn it man. I want arm laptops to be good. I can't justify buying anything apple even with Asahi as good as it is now.
Apple arm chips still have good days ahead of it. It’s been years since the MacBook m1 was released, how can competition not catch up already ?
I'm not sure i rate this particular article.
They seem to sort of hint at the importance of power and energy efficiency
But why did they then "ask about TDP" ? Surely they they need to know the actual input power(or energy) to achieve the benchmark, not TDP which is itself a wierd thing for chips that self regulate temperature by throttling.I'm not inclined to pay attention to this journo.
Power draw is heavily varied from fraction of a second to fraction of a second. TDP is more or less the target for average power draw over a longer period of time, and probably more importantly, something you're more likely to get a representative to answer. They're probably not going to give you the peak momentary power draw because it doesn't mean anything, and they probably won't give you average either. TDP is as much as you're likely to get.