The Lounge is rated Safe For Work. If you're about to post something inappropriate for a shared office environment, then don't post it. No ads, no abuse, and no programming questions. Trolling, (political, climate, religious or whatever) will result in your account being removed.
The world's largest chip, weighing in at an unbelievable 400,000 cores, 1.2 trillion transistors, 46,225 square millimeters of silicon, and 18 GB of on-chip memory, all in one chip that is as large as an entire wafer. Add in that the chip sucks 15kW of power and features 9 PB/s of memory bandwidth:
The system measures a mere 15 Rack Units tall, or roughly 26", so three can fit in a single rack. Given the performance, that's an incredibly compact package: It would take a 1,000-GPU cluster, which consumes 15 entire racks and half a megawatt of power, to match the performance of one CS-1 system. That's because a single Cerebras chip has 78 times more cores, 3,000 times more memory, and 10,000 times more memory bandwidth than a single GPU. It also has 33,000 times more bandwidth (PB/s).
«One day it will have to be officially admitted that what we have christened reality is an even greater illusion than the world of dreams.» Salvador Dali
They compared its 400k cores to a GPU with 5k cores, so they are probably continuing the bad tradition of calling SIMD lanes "cores", IIRC that's something GPU vendors started (specifically nVidia and their "CUDA processors", AMD likes to call them "stream processors" which is fine I guess?). Or maybe it's even counting some more specialized component such as 8-bit multipliers (important in deep learning) or whatever, who knows, actually if someone knows please explain.
Either way it's a beastly processor, actually either way it's the same processor, I just think their marketing material is deliberately misleading about its internal organization.
I pride myself that I don't scare too easily, with one exception:
After a recent clean install on my machine (A Dell desktop), I thought it's a good time to run Dell's update utility. It gave me the dreaded message: "Your BIOS needs updating"! I do understand a little about BIOS systems and what can happen if you write a new BIOS to the main board and the process gets interrupted or fails for any reason. In many instances you can kiss your main board goodbye!
So with trembling fingers I told Dell Update to proceed, eyes glued to the main monitor. I blush to disclose that I may even have uttered a silent prayer - my first in a long time. Things did not seem to go very smoothly. The machine must have restarted about 5 times. At this stage my nails were gone!
Finally after many minutes, the machine restarted a final time and I was greeted with the running cursor as Windows loaded. I breathed out and collapsed behind the keyboard. What a hair-raising experience!
In the old days, nine out of ten program installs required a reboot (ignoring those where "installing" was nothing but copying an .exe). This carried over from DOS to the early Windows versions, and five reboots for an entire, complex update is nothing, by those standards!
It annoyed people so much that MS introduced some mechanisms that made reboots unnecessary (e.g. a program already running could continue using the old version of a DLL, while programs started after the installation would use the new version, and the old version would be flagged for deletion at the next reboot). If you wanted a "Designed for Windows 95" sticker on your product, one of the requirements was no reboot required for installation. It turned out to be too strict for some classes of software, so the requirement was later dropped.
I agree that five reboots is a little on the high side, but I do not blink over two or three. Maybe that is because I have seen far more reboots in the pre-Win95-days. (Admittedly: Rebooting DOS and early Windows was a much faster operation than rebooting a modern Windows, especially when you consider the speed of the hardware.)
The strange part is: The old BIOS seemed to work fine since I got the machine a year ago. And: I don't see any difference with the new version. I wonder what can be different, but I assume Dell know what they're doing.
Does anyone, ever? Nowadays, these updates usually take care of rather obscure scenarios only. I'm sure Dell has a page that describes what the update is supposed to do, but I'll bet even if you read it you'd wonder why you bothered. Not that I ignore firmware updates.
Last Visit: 22-Nov-19 16:35 Last Update: 22-Nov-19 16:35