[Arm-netbook] Have you checked out this Open Source RISC-V-based 32-bit μC?

Luke Kenneth Casson Leighton lkcl at lkcl.net
Tue Nov 22 10:33:27 GMT 2016


http://www.electronics-eetimes.com/news/codasip-baysand-bring-risc-v-processor-ip-market?news_id=88521

may be related, may be not - this is 65nm (around 700mhz top speed)
and 40nm (around... 1.2ghz top speed if you're very lucky and have a
decent design).

"multivendor" program, basically you chuck a whole bunch of customer's
designs onto a single wafer, absolutely nobody expects to get very
many actual working ICs out of it, but the cost of a single wafer
(bear in mind that 16 wafers have to be made simultaneously, total
cost around $250k or that order, for 40nm) is subdivided amongst many
many people.  end result: relatively cheap prototype grade ICs.

the bit about conversion from FPGA... *shudder* that's truly dreadful,
to not bother with the conversion from FPGA layout to a proper IC
layout, just do like a "hard copy" of an FPGA, but given that the
tools can cost $250k *PER WEEK* to rent (not buy) that's hardly
surprising.

anyway all looks very interesting.

l.

---
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68


On Fri, Nov 18, 2016 at 4:57 AM, Luke Kenneth Casson Leighton
<lkcl at lkcl.net> wrote:
> ---
> crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
>
>
> On Thu, Nov 17, 2016 at 9:09 PM, Paul Boddie <paul at boddie.org.uk> wrote:
>> On Thursday 3. November 2016 15.59.16 Luke Kenneth Casson Leighton wrote:
>>> what's the name of the group behind it, and who is in that group?
>>> there's been a lot of publications and discussions online about risc-v
>>> - it should not be hard to find what they're planning as it is
>>> extremely likely that it's been discussed online for several months if
>>> not years.
>>
>> We at least know who is fabricating their product:
>>
>> https://www.crowdsupply.com/providers/electronics-fabrication/taiwan-
>> semiconductor
>>
>> That's rather interesting because I had the impression that TSMC was really
>> for people committing to huge volumes, with orders covered by highly secretive
>> contracts, which isn't what you'd normally associate with a crowdfunding
>> campaign. But maybe they like to keep their older foundries busy with smaller-
>> volume orders like this.
>
>  yeah i investigated TSMC some time ago, iirc they still have down to
> .36 micron, typically used for power transistors and so on.  they'll
> almost certainly have .18 micron and 65nm - all of these will be
> highly cost-effective (lots of competition), not in huge demand, and
> would also be a good proving-ground for an effort such as a RISC-V
> slow-speed 32-bit processor.
>
> l.



More information about the arm-netbook mailing list