Not for rocketry, but still cool electronics

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I once worked where they ran 4 Cray 90 super computers. What amazed me even more than the computers was the amount of support equipment required to run those things. The power was run on copper I-beams and the cooling units were twice the size of the computers. IIRC they had 8 or 10" chilled water lines feeding the coolers.

The amazing thing is today they can put in a rack of linked CPUs, get more computing power, and just use some fans to blow outside air at any temp threw the rack to cool it.
 
I once worked where they ran 4 Cray 90 super computers. What amazed me even more than the computers was the amount of support equipment required to run those things. The power was run on copper I-beams and the cooling units were twice the size of the computers. IIRC they had 8 or 10" chilled water lines feeding the coolers.

The amazing thing is today they can put in a rack of linked CPUs, get more computing power, and just use some fans to blow outside air at any temp threw the rack to cool it.
But no matter how crude they may seem now, they were ground-breaking machines. I'm amazed at the clock speed which was possible in something that spread out versus the tiny distances found on IC chips where it is far easier. Of course, fighting that (sub)speed of light velocity factor limited distance delay was a reason for the circular layout.
 
Great channel.

Mailbag (Displays, LoRa, Pi & Pi Pico, Watches, Goodies)

 
Groundbreaking 16-bit HP CPU used only internally by HP.



The CPD NMOS II Hybrid Microprocessor

http://www.hp9825.com/html/hybrid_microprocessor.html
Starts on page 15:

https://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1976-06.pdf
Click to enlarge:

51120497350_42e6764d6b_o.jpg
 
How NASA is Adapting to Busier and Noisier Communications With Mars - Perseverance’s lead telecom engineer explains how to handle interplanetary radio congestion

https://spectrum.ieee.org/tech-talk/telecom/wireless/how-nasa-is-adapting-radios-to-a-noisier-mars
Mars is more crowded than ever. With a successful arrival by the Perseverance rover in February, NASA alone is operating three orbiters, two rovers, and a lander on or around the planet. Other orbiters above the planet are being operated by the European Space Agency, Roscosmos, the Indian Space Research Organization, and others. The China National Space Administration hopes to successfully land its first rover, Tianwen-1, later this year. Radio spectrum, which never seemed in short supply on the barren world, is suddenly facing some of the same interference concerns that satellites around Earth already deal with.

Peter Ilott is the lead telecommunications system engineer for the Mars 2020 mission at the Jet Propulsion Laboratory. After more than a decade in the commercial spacecraft industry, he came to JPL in 2000. In the subsequent two decades, he’s had a hand in every Mars mission undertaken by NASA and JPL. Ilott recently took some time to talk with IEEE Spectrum about how interplanetary communications work, what happens when Mars gets more crowded, and how NASA will have to handle the consequences of that crowding.
 
I once worked where they ran 4 Cray 90 super computers. What amazed me even more than the computers was the amount of support equipment required to run those things. The power was run on copper I-beams and the cooling units were twice the size of the computers. IIRC they had 8 or 10" chilled water lines feeding the coolers.

The amazing thing is today they can put in a rack of linked CPUs, get more computing power, and just use some fans to blow outside air at any temp threw the rack to cool it.
Supposedly (this may be an Urban/Internet Legend), Dr. Seymour Cray once stated that "Memory is like an orgasm: it's better if you don't have to fake it."

Cray disliked Virtual Memory. Every machine that left his factory had the maximum memory the processor was capable of addressing installed in it. No swapping, paging or other nonsense like that. If the CPU could address a Terabyte of RAM, then that's how much RAM would be installed before it got to the customer. (Not that any of his machines could address a Terabyte, but you get the picture.)
 
20 Apr 2021
Cerebras’ New Monster AI Chip Adds 1.4 Trillion Transistors
Shift to 7-nanometer process boosts the second-generation chip’s transistor count to a mind boggling 2.6-trillion

https://spectrum.ieee.org/tech-talk...-ai-chip-now-has-a-trillions-more-transistors
The 7nm, 2.6 trillion transistor version:

Cores: 850,000
On-chip memory: 40 gigabytes
Memory bandwidth: 20 petabytes/s

For fairly obvious reasons, the size of the chip itself hasn’t changed. 300-millimeters is still the maximum wafer size in mass production, so the chip’s outer dimensions can’t change. And despite having twice as many AI cores, the WSE-2 looks just like WSE to the naked eye. It’s still divided into a 7 x 12 grid of rectangles, but that’s just an artifact of the chip-making process.

The computer system that hosts the WSE 2, called the CS-2, hasn’t really changed much either. “We were able to carry forward significant portions of the physical design,” says Feldman.

The CS-2 still takes up one-third of a standard rack, consumes about 20 kilowatts, relies on a closed-loop liquid cooling system, and has some pretty big cooling fans. Heat had been one of the biggest issues when developing a host system for the original WSE. That chip needed some 20,000 amps of current fed to it from one million copper connections to a fiberglass circuit board atop the wafer. Keeping all that aligned as heat expanded the wafer and the circuit board meant inventing new materials and took more than a year of development. While the CS-2 required some new engineering, it didn’t need that degree of wholesale invention, according to Feldman.


MzgwNTgzNA.jpeg


MzgwNTgxMw.jpeg
 
I had a 64 node SP2 cluster of RS/6000 604e high nodes in the late 90s. Each node had 8 CPUs. 56 minute boot times SUCKED!!!
You haven't lived until you've had to wait through a DEC Vax 11/750 booting from a TU-58 cartridge tape... it was like a little cassette tape. Took forever... with no verification on the console until you got the monitor prompt, just a blinky light on the tape drive.
 
You haven't lived until you've had to wait through a DEC Vax 11/750 booting from a TU-58 cartridge tape... it was like a little cassette tape. Took forever... with no verification on the console until you got the monitor prompt, just a blinky light on the tape drive.
I'll stick with my sub-minute boot times now. Had fun and learned alot back then but nothing was fast. We still had 9-track reel to reel when I retired that IBM hunk of metal.
 
Ahh, the good old days :rolleyes:. I used to write the software that controls the trains that run in Melbourne. The system is a gaggle of DEC PDP11/84 microcomputers (IBM Power PCs nowadays, emulating the PDP processors) and the compile time was 7.5 hours.
 
Last edited:
You haven't lived until you've had to wait through a DEC Vax 11/750 booting from a TU-58 cartridge tape... it was like a little cassette tape. Took forever... with no verification on the console until you got the monitor prompt, just a blinky light on the tape drive.
Cris, I "grew up" on a PDP-8e with timesharing hardware running TSS-8 (largely written by Dave Cutler). To backup the machine, there was a short sequence that had to be hand-toggled into the core memory, which read the bootloader from paper tape which, in turn, would allow you (among a couple of other things) to backup the hard drive to DECtape which I did nightly. Ahhhh... back when life was simple. ;)
 
Last edited:
You haven't lived until you've had to wait through a DEC Vax 11/750 booting from a TU-58 cartridge tape... it was like a little cassette tape. Took forever... with no verification on the console until you got the monitor prompt, just a blinky light on the tape drive.
Hadn't done that, but did regularly deal with the boot of a VAX 6000/430 with over 500 LAT printer devices to configure during boot. Over 45 minutes to boot.
 
We had a funky boot-rom on our PDP-11, it didn't always work. When it didn't I had to toggle in the bootloader code from the switches. There's a reason why all that stuff was in octal, instead of hex... it's a lot easier to mess up hex.

Kids just don't understand this stuff nowadays... the reality is that they don't really know how computers actually work. They don't understand printers, either... I'm rapidly getting the reputation for being "The Printer Guy" at my new job, because the younger techies can't get beyond installing a driver in Windows and setting an IP address.
 
I loved our line printer. Had to wear hearing protection around it tho. And I got some nasty shocks when I worked on the burster without grounding myself first.

And I wanna shoot the guy who decided to put a nylon gear in the Oki 321. Those things are still in use today but you gotta change that gear every year or so.
 
Last edited:
28 Apr 2021
The Radio We Could Send to Hell
Silicon carbide radio circuits can take the volcanic heat of Venus

https://spectrum.ieee.org/semiconductors/materials/the-radio-we-could-send-to-hell
The Vulcan II is a chip with multiple silicon carbide analog and digital circuits for testing at 500°C. We’ve made more than 40 circuits so far with Vulcan II and its predecessor.

[1] Ring oscillator [2] 8-bit successive approximation analog-to-digital converters and 4-bit ramp analog-to-digital converters [3] RS 485 receiver [4] 8-bit adders and 4-bit multipliers [5] 555 timer [6] 3-stage operational amplifier [7] DC-to-DC converter [8] Integrated gate drivers


MzgxMDUyMA.jpeg


MzgwOTgxOQ.jpeg
 
300,000 transistors within the cross sectional area of a human hair. Probe needles positioned using electron microscope.

Unreal Precision - Analyzing a Single TSMC 7nm Transistor

 
The Transistor: a 1953 documentary, anticipating its coming impact on technology

 
Back
Top