2,000 Retired Google Pixel Phones Get a Second Life As a Private Cloud (theregister.com) 26
UC San Diego researchers are working with Google to build a private cloud from 2,000 retired Pixel Fold motherboards, demonstrating how discarded smartphones could provide useful, low-cost computing capacity. "The full smartphone cluster is expected to launch this fall," reports The Register. "Depending on how well the initial phase goes, we're told the cluster could grow even larger." From the report Once the phone's motherboards have been extracted from their shells, the researchers say that the chips hiding within remain more than potent enough to be useful for a variety of tasks. In many cases, the single-threaded performance of these chips is as good as, if not better than, what you'd find from a many-cored datacenter chip. The Pixel Fold smartphones, which will form the basis of the cluster, are powered by a Google Tensor G2 processor with two 2.85 GHz Cortex-X1, two 2.35 GHz Cortex-A78 and four 1.80 GHz Cortex-A55 Arm cores, a Mali-G710 MP7 GPU, and 12 GB of system memory. Early benchmarking using the SPEC suite suggests that 25-50 phones should deliver performance similar to that of a conventional server.
The major challenge, instead, is distributing workloads across multiple devices, each of which has a handful of cores of one or more varieties, and most have 8-12 GB of memory. UCSD researchers are approaching this challenge from a couple of different angles. The first is by targeting applications that can easily fit within a single device. The second is using Kubernetes to orchestrate container deployments across clusters of 25-50 phones. For this to work, the devices first need to be flashed with a Linux operating system suitable for the job. While Android makes for a great handheld experience, it is not intended for server duty. In the blog post, researchers note that Android includes functionality intended to stop rogue applications from chewing up excessive amounts of memory and draining your battery. In server context, these safety mechanisms are no longer necessary.
[Ryan Kastner, an associate professor of computer science at UCSD] told us this was by no means an easy task, but the team has made steady progress toward getting Linux running smoothly on these devices, including support for the phone's onboard GPUs. Access to some functionality, like the chip's integrated tensor processing unit, remains elusive. Clustering these devices will require networking the phones together. Normally these devices would connect over cellular or Wi-Fi, but at this scale, this not only isn't practical, but also has implications for security, he explained. Instead, the team will employ PCBs that both supply power and break out wired Ethernet networking.
The researchers suggest that many EdTech, grading, and research workloads commonly run by universities in the cloud are small enough to run on the cluster without issue. "The vast majority of these applications are within the capabilities of a single smartphone to host, with the standard grading backend running on small cloud instances," a blog post detailing the planned deployment reads. "Early experiments show that even a moderately-sized cluster of 20 phones is capable of supporting peak submission rates for a 75+ student class."
The major challenge, instead, is distributing workloads across multiple devices, each of which has a handful of cores of one or more varieties, and most have 8-12 GB of memory. UCSD researchers are approaching this challenge from a couple of different angles. The first is by targeting applications that can easily fit within a single device. The second is using Kubernetes to orchestrate container deployments across clusters of 25-50 phones. For this to work, the devices first need to be flashed with a Linux operating system suitable for the job. While Android makes for a great handheld experience, it is not intended for server duty. In the blog post, researchers note that Android includes functionality intended to stop rogue applications from chewing up excessive amounts of memory and draining your battery. In server context, these safety mechanisms are no longer necessary.
[Ryan Kastner, an associate professor of computer science at UCSD] told us this was by no means an easy task, but the team has made steady progress toward getting Linux running smoothly on these devices, including support for the phone's onboard GPUs. Access to some functionality, like the chip's integrated tensor processing unit, remains elusive. Clustering these devices will require networking the phones together. Normally these devices would connect over cellular or Wi-Fi, but at this scale, this not only isn't practical, but also has implications for security, he explained. Instead, the team will employ PCBs that both supply power and break out wired Ethernet networking.
The researchers suggest that many EdTech, grading, and research workloads commonly run by universities in the cloud are small enough to run on the cluster without issue. "The vast majority of these applications are within the capabilities of a single smartphone to host, with the standard grading backend running on small cloud instances," a blog post detailing the planned deployment reads. "Early experiments show that even a moderately-sized cluster of 20 phones is capable of supporting peak submission rates for a 75+ student class."
Memory prices (Score:4, Interesting)
Given that the price of memory is so high that cell phone production is dropping like a stone, it is clear to me that used cellphones will end up being worth a lot more - both for this kind of networked computing and merely to remove and reuse parts
Re: Memory prices (Score:2)
What would really make them worth something is an easy upgrade path to an operating system that was still getting security updates.
Google, Apple, and the major phone vendors could score big PR points be extending security updates to 10 years on products introduced since 2016. In the long run PR points can translate into customer loyalty which can translate into "Step 4: PROFIIT!" in a non-sarcastic way.
Re: Memory prices (Score:2)
My iPhone 7+ got updates for 8 years.
Re: (Score:3)
What would really make them worth something is an easy upgrade path to an operating system that was still getting security updates.
Google, Apple, and the major phone vendors could score big PR points be extending security updates to 10 years on products introduced since 2016. In the long run PR points can translate into customer loyalty which can translate into "Step 4: PROFIIT!" in a non-sarcastic way.
The iPhone 6s (released in 2015) got a security update last month. So that's almost 11 years and counting.
Re: (Score:2)
I think it would be interesting to see companies adopt designs where chips could be taken from phone logic boards and
Re: (Score:2)
It'd be nice if people would use all this excess coding capacity as provided by AI to clean up the human written bloat slop they've been churning out for the past 20 years...
Re: (Score:1)
Somehow that seems even less likely to happen. The amount of vibecoded bloat slop will shock and horrify us. We will wish for the days of human-coded bloat slop.
Re: (Score:2)
Could be an interesting use for old phones. I've got a couple. I looked at using them as CCTV cameras, but the software is kinda crap. If they could run a more generic Linux distro they might be useful, low power servers.
Re: (Score:2)
With the way memory prices have gone with this latest generation, I want the next gen of CPU's to have a memory interface that can support mulitple different types of memory.
EG, the boat doesn't have memory slots - it has a connector that allows you to connect memory slots of whatever type you want (DDR6, DDR5, DDR4, etc).
These damned memory sticks have gotten too expensive to replace every time I upgrade my CPU.
I wouldn't mind seeing GPU's with socketed VRAM either. Given how much of the price is tied to
Re: (Score:2)
So, our dumb phones and outdated smartphones (e.g., iPhone 6+) will be useful again? :P
Obligatory (Score:5, Insightful)
Re:Obligatory (Score:5, Funny)
Imagine a beowulf cluster...
came to say the same, but TFA DID imagine it!
and it's some sort of /. crime that the line didn't make it into TFS.
Alongside traditional IT applications, the cluster will also support exploration into parallel computing and systems programming, which sounds an awful lot like the smartphone equivalent of the Beowulf clusters of the ‘90s, which saw researchers cobble together supercomputers from consumer PCs.
Don't Bot Farms... (Score:1)
Re: (Score:2)
The one you saw is probably the fake one [reuters.com]. However, Chinese phone farms do exist and they advertise [cxtfactory.com] in the [niaozun.shop] open [alibaba.com].
What about other parts? (Score:1)
There is a market for parts that aren't crypto-locked to or residing on the motherboard.
Cheap small usable high-quality camera for my next project? Yes please.
Meaningless stunt is meaningless (Score:2)
Seriously, were are we? In kindergarten?
Re: (Score:1)
nerds doing cool stuff with tech is what slashdot is for
Acurast project (Score:2)
I came across a project a while ago, Acurast [acurast.com], that is a distributed system of a quarter million smartphones, rented out as compute capacity. IIRC, the most common workloads are pretty lightweight, like web scraping. It's a good way to keep old phones (or phones with broken screens) out of the landfill, but I wasn't convinced that it would be competitive against what you can rent in the cloud.
And then chip prices went crazy.... maybe there is something to this phones-as-compute-resources idea after all.
How much do they consume? (Score:2)
What you save in scrapped hardware you will lose in energy waste.
Nice idea though, but not really doable.
Re: (Score:2)
Although, if your apartment building doesn't have you pay for power (like mine)... mining coin is worth it, and if I could use the old phones I have to make some money, that'd be nice.
Re: How much do they consume? (Score:2)
That's stealing electricity.
Just cause it can be done, use common sense (Score:2)