A couple thoughts: Xiaomi probably builds an order of magnitude more phones than LG and Google. So there's volume discounting at play. Also, Xiaomi's affordable phones are missing important NA bands. I wonder how this omission influences the cost of the SOC.
865 is simply to small a performance jump to be worth it. Flagship phones are mostly about the latest features not the specs as relatively few people actually buy them.
Unfortunately it's a bandwagon effect. If you don't have the latest processor (and another major flagship does), you're behind, regardless of actual performance benefits.
The number of people outside of HN who buy a phone based on the processor is vanishingly small.
It's not like when Intel's advertising told everyone they should feel bad if they didn't have a Pentium in their desktop boxes. If you ask people what processor is in their current phone, 99.99999999999% will have no idea.
I would bet the number of people on HN who buy a phone based on the processor is vanishingly small. Myself included. Maybe remove a couple of 9's and your second statement would hold too.
I honestly would think HN would be the opposite. The only reason I got rid of my Nexus 5 was that the battery was dying, there isn't really anything compelling enough for me to even consider a new flagship phone.
> The number of people outside of HN who buy a phone based on the processor is vanishingly small.
I would say that the number of people on HN who buy a phone based on the processor is vanishingly small.
And this is not a trend that went unnoticed. A very expensive Digitimes market research papers tell that we had a very long and very pronounced trend of people who previously bought flaghship products downgrading to mid and low end products.
It's very visible in the Laptop market too. People now how are buying into a niche of laptops with weak CPU, but good large screens. Those were making killer sales for last 3-4 years.
Sure, and I'm not saying people do... I'm just saying that if one company decides to include a Snapdragon 865, then you heavily pressure other companies into including it. It's only when everyone collectively agrees to not include the latest that you see that happen.
Xiaomi has a very interesting pricing strategy. Generally speaking, they tend to be lower margin, but also they really cut supplier margin to the bone. They love pitting competitors against each other for the same socket (promising high volume, but ultimately you get beat out by someone else). It's hard to turn down their promised volume numbers. I wouldn't be surprised if they're getting the 865 cheaper than other players like Samsung.
Xiao is loosing money on most of its phones. Phones are a small net loss for them.
They recoup the cost big time on accessories, and other merchandise.
It didn't work in India (biggest market outside of China,) where BBK rules supreme: People will buy the best deal phone, but will not sign onto their overpriced smart vacuum cleaners.
Funny you mention smart vacuum cleaners. Xiaomi vacuums (the roborock ones) are some of the best value to feature robot vacuums out there. The only thing that beats them on the high end is the iRobot latest, which costs like 2-3x as much.
1. Qualcomm has invested in Xiaomi. It’s possible that influences the prices they can negotiate.
2. Like Samsung, Xiaomi has their own in-house SoC development program. Just the threat of an alternative gives them more ability to negotiate pricing.
I think you based this counter-argument on the title instead of the article. The article says:
> According to the report, both Google and LG are opting for the Snapdragon 765G, but not because it’s just saving them money. It’s Qualcomm’s first SoC to sport an embedded 5G modem saving manufacturers a ton of design hassles.
And that Xiaomi you linked doesn't have mmWave 5G. So Xiaomi avoided the design complication the 865 results in by just not having the full radio feature set.
This then is the other side of the problem with 5G being not the same thing as 5G. It's not just carriers that get to save money rebranding 4G LTE as 5G, but also phone manufacturers can now also claim 5G without actually supporting all of 5G. I wouldn't be surprised if mmWave 5G gets branded as 5G+ or something like that to differentiate itself from 5G-the-rebranded-4G-LTE.
Assuming that Apple releases the iPhone 9 with the latest A13 chip for $400 to $500 soon, what will be good Android alternative? Compact size, most powerful chip, affordable. I was eyeing S10e last year. Has anything better been released since?
I was really impressed with the Pixel 3A myself... I'm still holding on to my 2XL, I got a 3A for my daughter last fall after her Galaxy S6 bit the dust.
I don't think I'll convince myself to spend over $400-500 (or value equivalent) again... I didn't like spending as much as I did on my 2XL, and the features just don't seem worth it to me. I'm also holding on to the 3+ year mark this time around... my last couple phones really haven't been much of an improvement...
Nexus 4, Nexus 6p, Pixel 2XL... each has mostly been out of frustration with failures from the prior generation's battery failure (the Nexus 6p caught fire at the repair site for a battery replacement). TBH, I feel it's time to require user serviceable batteries for all electronic devices with a rechargeable battery, leaning into right to repair on this.
what's the story on updates for non-pixel phones these days? the main reason I keep buy google phones is for the monthly security updates that seem to go on for at least three years after release, but every year the pixel hardware seems like a worse value.
Can't comment on all vendors, but OnePlus at least are reasonably decent. My (two-three generations behind) OnePlus 6T got the February 2020 patches about two weeks ago if that's indication, and has been on Android 10 for a few months maybe?
Patches lapsed a bit while they were concentrating on Android 10 though iirc.
The solution to this I propose to anyone who cares about hardware choice, privacy and software independence is to buy a phone supported by LineageOS and install that on it.
You won't have any issues with a lack of updates for a long time, as they're automatically built in the CI for the model and shipped to you via OTA updates.
Here's the list of viable phones (anything not on this list I wouldn't consider, in the Android world, anyway): https://download.lineageos.org/
Bonus: There is a version of LineageOS that works with apps usually requiring proprietary Google binaries without them actually being installed: https://lineage.microg.org/
I had the One Plus One (first phone), only switched away when I went to Verizon for better coverage on the road. Is their band coverage for Verizon still partial? Haven't checked in the past year or so, last I checked you could use it with Verizon, but only with LTE bands.
Most powerful chip is hard to come by in small Android phones and often not really needed/warranted. I think you'll get poorer performance a couple years down the line (which is, I expect, the reason you're wanting the best chip) if you have to deal with Samsung touchwiz and bloatware vs. bare, unlocked Android.
The Motorola One lineup checks all my boxes right now: Stock Android with unlocked bootloader, uSD card, good displays, good battery, headphone jack, stays updated, and under $400. I'm mourning the loss of swappable batteries and small displays in the market, but it's the best option I've found.
They have atrocious support. Bricked a whole bunch of Moto G phones a couple of years ago - by breaking the charging functionality with a firmware update. Just straight up didn't fix it. Thousands of people affected. I'd never consider a Motorola device again.
There won't be a competitive Android alternative. The A13 will beat most x86 chips in terms of performance. No other ARM licensee has a chip out that's anywhere close.
That's interesting, is that peak performance or is it sustained performance?
I've given up on performance metrics for phones as 99% of them will just overheat within seconds and throttle if you actually try to use the advertised performance.
I've never owned any apple product so I'm curious if this is the case for them too.
It's based on one single benchmark (geekbench) that has widely been discredited when comparing platforms (like x86-64 vs ARM64).
iPhones also fare much worse in sustained / multiple run benchmarks due to thermal throttling. In some test run in a loop, iPhones are as much as 1/3rd slower in their 3rd run vs their 1st run due to throttling.
The age old question: does a benchmark that runs for 20 minutes before throttling really represent any real world gains or lack thereof? For example I want a few images/videos to compress quickly before I share them with someone, but I will pretty much never need my phone to chew through gigabytes of data all at the same time. I want a webpage to load within a few seconds but I don’t constantly refresh webpages every 5s for 30mins straight.
The last phone I bought was a one plus 6 and the next phone I buy will be a one plus. I don't plan on looking at anything else. Everything is fast and solid. Fast UI, good screen, fast charging, minimal bloat. Decent cameras decent speakers, usb C. No mobile processors stand up to apple's killer custom chips, but without bloated software android phones can get close enough.
Purely in terms of Performance, if iPhone 9, Apple's Entry level iPhone will get an A13, then no Android on the market, including flagship Qualcomm SD 865 will be as fast.
DHH and Jeff Atwood ( discourse ) had a rant about the performance of Android on JS Webview rendering.
I'm confused as to what you mean by "hit its limits", considering there's still a lot of room for improvement - just go anywhere that's not a major urban centre. Are we at a point where we can't possibly improve as the laws of physics won't allow it? I find that doubtful.
So you're ruling out a change in tech? We're stuck here because there's no reason we could conceive of something better due to the ROI? I'm confused as to why you believe now would be that moment, as opposed to something earlier.
Yeah, the laws of physics are not the limiting factor here. With faster compute, you can do better beamforming. There will always be lag between what is technically possible and what standards exist and are deployed, but I don't see any reason why wireless won't get faster over the coming years.
Channel capacity depends on signal power, noise power, and bandwidth. Bandwidth and signal power are bounded by government regulations, but if you can make your antenna more selective (rejecting more noise) you can transfer more bits per second over a channel. MIMO/beamforming is one way to do this, and it is mostly limited by DSP power. Compute/watt isn't really increasing exponentially anymore, but it is still getting better, so we can still squeeze more capacity out of available bandwidth.
I had an issue with the title but the article addresses the issue. It isn't just the higher price tag. It's the thermals and power management that comes with having a separate modem as well as the price tag.
tldr: "Google and LG are opting for the Snapdragon 765G, but not because it’s just saving them money. It’s Qualcomm’s first SoC to sport an embedded 5G modem saving manufacturers a ton of design hassles"
Good news to pick a lower power chip as Google has been putting wimpy batteries into their phones.
Maybe these low-priced phones with a top-end SoC are making some bad tradeoffs to get that low price, but there's at least some evidence that Qualcomm isn't forcing phone prices over $1,000.
The last time I check Xiaomi ASP in still below $150. That is basically saying they dont sell their flagship model in any volume and are likely used as loss leader or at break even.
And yes, WCCFtech deserve no place on HN. Hence I have flag it.
58 comments
[ 1.6 ms ] story [ 80.8 ms ] threadMaybe it would be better to link the original source.
Are they looking to make more affordable flagships or are they upset their margins are being hurt?
It's not like when Intel's advertising told everyone they should feel bad if they didn't have a Pentium in their desktop boxes. If you ask people what processor is in their current phone, 99.99999999999% will have no idea.
I would say that the number of people on HN who buy a phone based on the processor is vanishingly small.
And this is not a trend that went unnoticed. A very expensive Digitimes market research papers tell that we had a very long and very pronounced trend of people who previously bought flaghship products downgrading to mid and low end products.
It's very visible in the Laptop market too. People now how are buying into a niche of laptops with weak CPU, but good large screens. Those were making killer sales for last 3-4 years.
They recoup the cost big time on accessories, and other merchandise.
It didn't work in India (biggest market outside of China,) where BBK rules supreme: People will buy the best deal phone, but will not sign onto their overpriced smart vacuum cleaners.
1. Qualcomm has invested in Xiaomi. It’s possible that influences the prices they can negotiate.
2. Like Samsung, Xiaomi has their own in-house SoC development program. Just the threat of an alternative gives them more ability to negotiate pricing.
> According to the report, both Google and LG are opting for the Snapdragon 765G, but not because it’s just saving them money. It’s Qualcomm’s first SoC to sport an embedded 5G modem saving manufacturers a ton of design hassles.
And that Xiaomi you linked doesn't have mmWave 5G. So Xiaomi avoided the design complication the 865 results in by just not having the full radio feature set.
This then is the other side of the problem with 5G being not the same thing as 5G. It's not just carriers that get to save money rebranding 4G LTE as 5G, but also phone manufacturers can now also claim 5G without actually supporting all of 5G. I wouldn't be surprised if mmWave 5G gets branded as 5G+ or something like that to differentiate itself from 5G-the-rebranded-4G-LTE.
LTE-Advanced rebranded as "5Ge" (only AT&T for now AFAIK)
Low-band NR which is genuinely 5G but only moderately faster than LTE-A (T-Mobile)
High-band (mmwave) NR which is the fastest 5G flavor but it cuts out really easily (Verizon)
I don't think I'll convince myself to spend over $400-500 (or value equivalent) again... I didn't like spending as much as I did on my 2XL, and the features just don't seem worth it to me. I'm also holding on to the 3+ year mark this time around... my last couple phones really haven't been much of an improvement...
Nexus 4, Nexus 6p, Pixel 2XL... each has mostly been out of frustration with failures from the prior generation's battery failure (the Nexus 6p caught fire at the repair site for a battery replacement). TBH, I feel it's time to require user serviceable batteries for all electronic devices with a rechargeable battery, leaning into right to repair on this.
You won't have any issues with a lack of updates for a long time, as they're automatically built in the CI for the model and shipped to you via OTA updates.
Here's the list of viable phones (anything not on this list I wouldn't consider, in the Android world, anyway): https://download.lineageos.org/
Have been a happy user for a few months.
The Motorola One lineup checks all my boxes right now: Stock Android with unlocked bootloader, uSD card, good displays, good battery, headphone jack, stays updated, and under $400. I'm mourning the loss of swappable batteries and small displays in the market, but it's the best option I've found.
I've given up on performance metrics for phones as 99% of them will just overheat within seconds and throttle if you actually try to use the advertised performance.
I've never owned any apple product so I'm curious if this is the case for them too.
iPhones also fare much worse in sustained / multiple run benchmarks due to thermal throttling. In some test run in a loop, iPhones are as much as 1/3rd slower in their 3rd run vs their 1st run due to throttling.
DHH and Jeff Atwood ( discourse ) had a rant about the performance of Android on JS Webview rendering.
Think about it as your home wifi, it's essentially the same, how do you increase its capacity?
If the industry doesn't have to factor in ROI, then your argument is valid, but it's a business.
Channel capacity depends on signal power, noise power, and bandwidth. Bandwidth and signal power are bounded by government regulations, but if you can make your antenna more selective (rejecting more noise) you can transfer more bits per second over a channel. MIMO/beamforming is one way to do this, and it is mostly limited by DSP power. Compute/watt isn't really increasing exponentially anymore, but it is still getting better, so we can still squeeze more capacity out of available bandwidth.
Good news to pick a lower power chip as Google has been putting wimpy batteries into their phones.
$500: Xiaomi Mi 10
$500: Nubia Red Magic 5G
$500: iQOO 3
$550: OnePlus 8
$700: Sharp Aquos R5G
$800: Nokia 9.2
$800: LG V60 ThinQ 5G
https://arstechnica.com/gadgets/2020/03/reports-google-lg-do...
Maybe these low-priced phones with a top-end SoC are making some bad tradeoffs to get that low price, but there's at least some evidence that Qualcomm isn't forcing phone prices over $1,000.
(And yeah, don't post WCCFTech.)
And yes, WCCFtech deserve no place on HN. Hence I have flag it.