The 4G patents were also under FRAND. Didn’t stop Qualcomm from failing to actually license under those terms. The judge specifically found so in the case.
That is not strictly true, Qualcomm offered a price for all Wireless patents, and I hardly call it unreasonable given the contribution of Qualcomm in 4G and 3G.
Not to mentioned having read the 233 pages I found the Judge to be... one sided? She dismiss all the defendants' testimony because they were speaking too quick.
I didn’t realize so many companies had 5G related patents. Are they part of patent pool like MP3 (was) or h.264? Where you can pay one entity and it gets distributed to all of the holders?
>I didn’t realize so many companies had 5G related patents.
Which is part of the reason why you are downvoted. I upvoted it so it just stay visible and not lost in the conversation.
Nope, there isn't a patent pool. Which is part of the problem. But most Manufacturers and Assembly take care of it. So in this case, Foxconn would have everything sorted out for you. And if you are big enough, you can negotiate with each individual patents owner.
And there are lots of standard patents that aren't really useful in mobile, like lots of IoT and V2X use cases. Some of them I think are complete garbage.
I'm going to be honest, at this point, after blatantly stealing tech from both Cisco and Samsung, how can you even claim Huawei developed the technical know how and didn't just steal it?
Because they have a lot of patents. Either they developed the tech in the patents independently or they stole tech which other companies declined to patent. The latter option doesn't make sense since everyone in the wireless industry makes money from patent licensing. (Opposed to something like DRAM manufacturing tech where you generally aren't licensing patents and your tech is your competitive advantage)
But to be fair Huawei is spending a large amount of their revenue on R&D so not surprised they are inventing new technologies. That said we also know that they have stolen a lot of IP in the past (and maybe now).
While the stealing part is likely true, everytime a Westerner has dismissed the innovation ability of China, they've typically been wrong. China has long passed the phase of being an imitator with low cost manufacturing abilities, and matured into an innovator.
There probably is still a cultural tolerance of IP theft, but that too will pass in a decade
A more appropriate example would be if every engine was the same right down to the Daimler logo still plastered on the engine blocks. That company has no shame.
From everything I know,MediaTek basically caters to the low end. It will be years before catering to the low end will be a viable business model. The only companies selling high end phones at scale are Apple and Samsung and even Samsung sells mostly low end phones -their average selling price is below $300.
That’s also the problem with Android watches. The Android market for Android Wear is minuscule because the processors suck and the processors suck because the market is too small to be worth the investment.
Almost everyone. Radio nowadays is just a RF frontend + ADC/DAC block + DSP. RF is the magic grey beards part, but can be quickly solved with off the shelf parts. Conversion is also easy off the shelf, or even integrated into rf chip. DSP is where patents come into play and dragons live during implementation phase, but its more or less all software, Apple is good at software.
This is completely out of my area of expertise. But if that’s the case, why is it so much easier (relatively) for Apple to design processors with better thermals and lower power requirements than anyone in the world, soon will be able to design better GPUs for mobile (not high end GPUs) than any other company in the world, but can’t design cellular chips in house just as “easily”?
My limited understanding is that Apple A series chops benefited initially from expensive manual layout while many ARM chips of that time were using more machine driven layouts. Also Apple, as Apple does, city out a lot of legacy bits and pieces that Wintel are expected to support.
So leaner more rigorously designed chips. Of course I could be off the mark here.
As usual Anandtech has excellent deep dives into various SoCs:
> RF is the magic grey beards part, but can be quickly solved with off the shelf parts.
What do you mean by "off the shelf parts"? A competitive phone chipset is going to have integrated RF and digital baseband. Because of size constraints, you can't mix and match ICs and have any hope of fitting into a small form factor.
I mean RF is a solved problem, there are no secrets in that part of modem design anymore. You can pick an off the shelf cable modem RF frontend (AD9866 etc) and implement 4G baseband. What you will need is experienced software people and IP rights.
On a tangential note Intel totally screwed DOCSIS modem chipsets, after taking over TI's Puma product line, by stuffing it with x86 cores and assigning work experience students to write firmware, after all everyone can code x86, right? :) Puma 6/7 can be DOSed by as little as 10KB/s stream. Supposedly there are firmware fixes, but from what I gather they just move critical processing tasks from integrated x86 cores to external ARM SoC - prevents DOSing, but latency issues persist.
I think he means that there will be a few components that will be able to develop that IP for you. That is certainly the case for PLLs and other analog bits of digital designs.
