For sale: Cheyenne supercomputer with 8,064 Xeon CPUs and 306TB of DDR4 memory — some assembly and maintenance required
Buyer is responsible for relocation of 26,000 pounds of equipment.
What was once the 21st most powerful supercomputer in the world is now available to the highest bidder — well, maybe, as the current bid of under $30,000 has not met the required amount. The U.S. General Services Administration opted to put the Cheyenne supercomputer, deployed in 2016, up for auction, in part due to ongoing repair and maintenance problems.
The retired supercomputer is, as the name suggests, a monster. It’s a 5.34 petaflops system, one of the last deployed by Silicon Graphics International after its acquisition by Hewlett-Packard. Since then, it's been a cornerstone of operations at the NCAR-Wyoming Supercomputing Center in Cheyenne, Wyoming.
The Cheyenne supercomputer is a water-controlled installation made up of SGI ICE XA modules with 28 racks holding 8,064 Intel E5-2697v4 CPUs. That totals 145,152 cores, for those keeping count. The main system is spread across 4,032 dual-socket nodes. Here are the specs of the primary components:
Component | Quantity |
---|---|
E-cells (1500 lbs. each) | 14 |
E-racks (water-cooled) | 28 |
Intel E5-2697v4 CPUs | 8,064 |
DDR4-2400 ECC single-rank memory | 313,344GB |
IB Switches | 224 |
Air-cooled management racks | 2 (see below for specs) |
Each E-Cell weighs in at 1,500 pounds, and shipping is not included in the winning bid. The purchaser will need to hire a professional moving company to transport the supercomputer from the facility to its new home. The auction notes also state that the supercomputer will be sold as-is, and that it "is currently experiencing maintenance limitations due to faulty quick disconnects causing water spray." Not exactly the pinnacle of supercomputing achievements, then.
Beyond the above hardware, the supercomputer also includes two air-cooled management racks. These consist of 26 1U servers each, 20 of which have 128GB of memory and six with 256GB of memory. That's an additional 8TB of DRAM, if you're wondering. The management racks also include 10 Extreme Switches, and two Extreme Switch power units, and each rack weighs 2,500 pounds.
While the Cheyenne supercomputer has been in operation for the past seven years, the auction notes says the "expense and downtime associated with" fixing the current cooling problems makes it unworthy of continued maintenance. And of course, even though this was a lightning-fast supercomputer when it first launched, it would be considered sluggish by 2024 computing standards. This is a fate shared by many supercomputers, even some of NASA's most powerful ones.
Cheyenne peaked at number 21 on the Top500 list of the most powerful supercomputers back when it launched. Today, it sits at number 160 — based on an Rmax score of 4.79 petaflops. The paradigm shift to GPU-powered supercomputers over the past decade means that, as an example, you could potentially exceed that level of performance with around 23 Nvidia DGX H100 systems sporting 46 CPUs and 184 GPUs.
Even so, the auction comes with a treasure trove of parts and components for whoever is willing to pony up the cash. The supercomputer will be drained for removal, and it seems it won't necessarily include all the necessary cabling. However, it does include a whopping 313,344GB of DDR4-2400 ECC RAM. That alone could be worth more than $350,000 — not to mention an unspecified amount of storage.
Also of interest is that the supercomputer uses around 1.727 MW of power when fully assembled. Which means that if you want to power it up and run complex simulations on it, the power requirements could cost over $4,000 per day (depending on the price of electricity, naturally).
We presume most bidders would be more interested in parting out the system rather than attempting to get it running again. Besides the missing Ethernet and optical cabling that you'd need to acquire, there's the apparently unresolved issue of the leaking quick-connect liquid cooling components. But who knows? Maybe some enterprising business will find a way to bring Cheyenne back into service, like a phoenix rising from the water-logged ashes.
Updated: Bidding is scheduled to end on May 3, 2024 and is up to $28,085 $120,085 now. There's no longer a "reserve not met" disclaimer, which apparently means someone is going hope with several truckloads of old supercomputer parts. Whether the buyer will try to fix up the system and get it running, sell it for parts, or just grind it up for the raw materials remains to be seen.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
Jeff Butts has been covering tech news for more than a decade, and his IT experience predates the internet. Yes, he remembers when 9600 baud was “fast.” He especially enjoys covering DIY and Maker topics, along with anything on the bleeding edge of technology.
-
jp7189 I know the main article mentions leaky water connections, but "water spray" sounds a lot worse than I was thinking. From thr listing:Reply
"However, the system is currently experiencing maintenance limitations due to faulty quick disconnects causing water spray." -
Avro Arrow It's not even worth the power it uses at this point. The power:performance ratio is what really kills older supercomputers compared to new ones because the new ones are so much less expensive to operate.Reply -
punkncat I also found it interesting the disclosure of how heavy this thing is, as well as the need for specialized personnel and equipment required to move it. Optical and cabing not included....and how about the aspect that 1% of this gigantic pool of RAM is bad. They mention ECC RAM and one could do the math as to how many sticks that compromises and wonder if it can even be found at this point?Reply
It seems to me a lot like the current owner has found a way to have someone else pay to move a large bit of equipment that is facing operational obsolescence. -
JarredWaltonGPU
$50,000 bid right now and still "reserve not met." It's pretty clear the US gov't doesn't really want to try to fix the problems with Cheyenne, but it also looks like it's not just trying to offload the movement of the hardware to someone else. Otherwise, there wouldn't be a reserve.punkncat said:I also found it interesting the disclosure of how heavy this thing is, as well as the need for specialized personnel and equipment required to move it. Optical and cabing not included....and how about the aspect that 1% of this gigantic pool of RAM is bad. They mention ECC RAM and one could do the math as to how many sticks that compromises and wonder if it can even be found at this point?
It seems to me a lot like the current owner has found a way to have someone else pay to move a large bit of equipment that is facing operational obsolescence.
Though I suspect when all is said and done, they might attempt to work out some agreement with the top bidder, whether or not the reserve has been met. 1.7MW of power is a lot, to say the least. Sure, there are supercomputers that can use 60MW out there... but those are also not leaking and failing in various ways, and they're about 150X faster than this older supercomputer.
The real difficulty is that, even if you want to part this sucker out, that will take a long time and it's very niche hardware considering its age. You can often find used servers from 10 years back selling for a song, relatively speaking. Because data centers don't want them, and neither do most businesses or even server enthusiasts. Anyone who has been in a data center can tell you how much fun it is to sit next to even one server with all those whirring fans!
I'm betting the top bidder will end up being some recycling / parting out business, though. I just can't imagine anyone really wanting to try to "fix" a leaky supercomputer that the US government is unwilling to deal with. -
PEnns Well, maybe somebody could salvage some of the components and use it to build something that will run ...........Crysis smoothly.:cool:Reply -
brandonjclark
There should be at least 30k in raw materials in that thing, right?Avro Arrow said:It's not even worth the power it uses at this point. The power:performance ratio is what really kills older supercomputers compared to new ones because the new ones are so much less expensive to operate. -
Co BIY A university or government agency in a developing country seems like one of the few likely markets for something like this. Where you could have a pool of skilled but low cost labor for the upkeep and some need for the computing power on the cheap.Reply
The nice thing about the Cray-1 is that after it was no longer useful as a computer it was still functional as furniture.
https://www.cisl.ucar.edu/sites/default/files/styles/extra_large/public/2021-10/cray1.jpeg?itok=fegEQU0V -
Amdlova The software costs and additional hardware and engineers to run that will be insane.Reply
Better scrap it -
jp7189
I did some screwing around on ebay last night, and it seems you could get ~$500-700k from CPUs and RAM..theoretically. Actually selling 8000+ CPUs might be kinda hard.brandonjclark said:There should be at least 30k in raw materials in that thing, right?