As reported by ZDNet:
A decade ago, an idea was born in a laboratory at the University of California at Berkeley to create a lingua franca for computer chips, a set of instructions that would be used by all chipmakers and owned by none.
It wasn't supposed to be an impressive new technology, it was merely supposed to get the industry on the same page, to simplify chip-making in order to move things forward.
But a funny thing has happened on the way to a global chip standard: RISC-V, as the Berkeley effort is known, has begun to produce some technical breakthroughs in chip design.
As just one example, a recent microprocessor design using RISC-V has a clock speed of 5 gigahertz, well above a recent, top-of-the-line Intel Xeon server chip, E7, running at 3.2 gigahertz. Yet the novel RISC-V chip burns just 1 watt of power at 1.1 volts, less than one percent of the power burned by the Intel Xeon.
[...] The new 5-gigahertz processor, which is merely a prototype, is not the creation of a garage startup. It was made by Micro Magic Inc., a Silicon Valley intellectual property designer for chips that has been consulting to all the big Valley firms for twenty-five years. The ability of a small but seasoned crew of chip designers to accomplish such a task suggests a design renaissance that could be on the horizon.
(Score: 2) by eravnrekaree on Tuesday December 01 2020, @05:15PM (6 children)
The ISA is really about 0,0001% of what it takes to design a CPU. With computer assisted design of chips, it really shouldn't be too hard to design a new chip template either. The big engineering difficulty is in process technology. But you probably can just put together your own template and then farm out manufacturing to a fab who has the billions of dollars of equipment to make a chip and already does so for other companies.
(Score: 3, Interesting) by DannyB on Tuesday December 01 2020, @05:51PM (3 children)
It may not be hard, but doing an ISA is probably not easy either.
The ISA is not what is of major importance. It is that we may finally get an open instruction set that anyone can manufacture implementations of. Competition. Innovation. No vendor lock in.
To transfer files: right-click on file, pick Copy. Unplug mouse, plug mouse into other computer. Right-click, paste.
(Score: 0) by Anonymous Coward on Tuesday December 01 2020, @07:43PM (2 children)
From the article:
---------------------
The point, says Huang, is that because RISC-V is open, unlike CISC, the complex instruction-set architecture of Intel's chips, or even the version of RISC that's in ARM chips, things can be done in chip design to resolve that bottleneck in ways not possible if the chip's instructions were locked down.
---------------------
If the advantage is that the instruction set is easily changed, then this doesn't sound all that useful at all. It would be useful for a manufacturer inventing its own hardware and then writing all its own software. But without a standardized instruction set there's not much point other than that.
(Score: 2) by DannyB on Tuesday December 01 2020, @09:11PM
A common base instruction set is certainly important.
For exotic hardware, there might be special instructions. Once upon a time hardware to do floating point math was exotic and added instructions to the machine.
If Intel, AMD and ARM do not allow experimenting with additional instructions, then that kind of makes it impossible to fabricate hardware that might do something new and innovative. Having the ability to extend the instruction set, for your own purposes, is important to try new things.
But we don't want a hundred different variations of the instruction set. Common software probably isn't going to support them all. Standardization is important. But so is the ability to experiment without licensing restrictions from the instruction set licensor.
To transfer files: right-click on file, pick Copy. Unplug mouse, plug mouse into other computer. Right-click, paste.
(Score: 0) by Anonymous Coward on Wednesday December 02 2020, @04:53PM
Well it's RISC-V, I guess they'll submit a pull request. The whole point about RISC-V being open is just that.
(Score: 5, Touché) by jimtheowl on Tuesday December 01 2020, @05:52PM
99.99653% of statistics are made up.
(Score: 3, Funny) by mhajicek on Tuesday December 01 2020, @07:02PM
So we want CPUs that speak French? French cars are bad enough...
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 5, Insightful) by DannyB on Tuesday December 01 2020, @06:10PM (8 children)
Just imagine.
Suppose the first step of the instruction decode-execution pipeline were a decryption step! That's right, every individual processor manufactured has not only a serial number, but a unique instruction encryption key. Baked right in to the hardware.
The beauty of this is that no binary executable file from one machine can be moved to a different machine and executed. The processor cannot decrypt the instructions. No more software piracy.
But wait! There's more!
This would also buy us security! You could be sure that all your software was genuine Microsoft software that was encrypted by Microsoft, specifically for your microprocessor, using the secret key known only to Microsoft, the processor manufacturer, and the government. No more amateurs creating software. Only real, licensed software developers would be producing software. This would ensure that all your software would have the quality and security that you have come to expect from Microsoft.
This would also protect intellectual property secrets in software, because object code could no longer be disassembled to discover the inner workings of software. (if Microsoft used an O^2 bubble sort instead of a quick sort or heap sort, how could you possiblly tell?)
But wait! There's more!
Since we wouldn't have any more unauthorized software, we wouldn't have any more thieving piracy of copyright protected music and videos. The poor RIAA and MPAA would be rescued from the brink of starvation, finally.
In order to make this wonderful utopian idea a reality, it would be necessary to have world wide legislation that maintained the secrecy of the encryption keys of all microprocessors. They would be kept in an Excel spreadsheet locked in a vault whose lock is controlled by a secure IoT lock whose key is kept in the basement of the department of interior.
There would be no more talk of systemd.
To transfer files: right-click on file, pick Copy. Unplug mouse, plug mouse into other computer. Right-click, paste.
(Score: 3, Interesting) by RS3 on Tuesday December 01 2020, @06:57PM
Very good one. You should patent that. :)
In your 2nd to last line I was expecting "department of inferior."
(Score: 5, Interesting) by NateMich on Tuesday December 01 2020, @07:01PM (1 child)
That's great, but this seems far more likely to be something Apple would do at this point.
It would actually be trivially for them.
(Score: 1, Informative) by Anonymous Coward on Wednesday December 02 2020, @03:00AM
https://patents.google.com/patent/US7941640B1/ [google.com]
(Score: 5, Insightful) by pTamok on Tuesday December 01 2020, @07:09PM (1 child)
I'd rate this as funny, except it's actually scarily possible. I hope it is not prescient.
It reminds me of the Right to Read [gnu.org].
(Score: 3, Interesting) by DannyB on Tuesday December 01 2020, @09:12PM
It reminds me of: A Letter from 2020. (Which I've submitted as an SN story, which is still pending.) That letter was written 20 years ago in Sept 2000.
To transfer files: right-click on file, pick Copy. Unplug mouse, plug mouse into other computer. Right-click, paste.
(Score: 5, Funny) by The Mighty Buzzard on Tuesday December 01 2020, @07:10PM
You know, it would almost be worth it to get rid of systemd.
My rights don't end where your fear begins.
(Score: 0) by Anonymous Coward on Tuesday December 01 2020, @10:26PM
Wow, was getting excited there and then you lost me at "Excel spreadsheet".
Consider those keys pawned then?
(Score: 1, Informative) by Anonymous Coward on Wednesday December 02 2020, @02:58AM
https://patents.google.com/patent/US7941640B1/ [google.com]
(Score: 1, Insightful) by Anonymous Coward on Tuesday December 01 2020, @08:22PM (2 children)
From http://www.micromagic.com/ [micromagic.com]
Which puts it comparable to 8-12 year old processors from AMD as per https://www.eembc.org/coremark/scores.php [eembc.org] ... Still an accomplishment, especially at that level of power consumption, especially if it's stable and can eventually be purchased affordably, but 5GHz doesn't really mean anything on its own.
(Score: 2) by turgid on Tuesday December 01 2020, @08:27PM (1 child)
As intel demonstrated with the Pentium IV: more haste, less speed.
I refuse to engage in a battle of wits with an unarmed opponent [wikipedia.org].
(Score: 2, Insightful) by Anonymous Coward on Tuesday December 01 2020, @11:27PM
Depends though. The issue with the P4s was that they had very deep pipelines. That allowed Intel to push the clock speed, but required an overall longer latency for each instruction. It also made branch mispredictions very costly. The trade-off just wasn't with it. But if you can push the clock frequency without upping the latency or burning up the chip, it's a win.