Just found this:

http://portal.acm.org/ft_gateway.cfm?id=1400228&type=pdf&coll=portal&dl=ACM&CFID=76120516&CFTOKEN=63080657

Very interesting. Confirms that we made the right bet by not committing to STM.


Andrei

On Feb 2, 2010, at 9:44 AM, Andrei Alexandrescu wrote:

> Just found this:
> 
> http://portal.acm.org/ft_gateway.cfm?id=1400228&type=pdf&coll=portal&dl=ACM&CFID=76120516&CFTOKEN=63080657
> 
> Very interesting. Confirms that we made the right bet by not committing to STM.

I think I forwarded that article to you when it was in CACM a while back :-)  Definitely a good read.  I think the correct implementation of HTM may have a chance though.  That "Bulk Multicore Architecture" article I linked a while back seems incredibly promising.

Sorry. I'm indeed behind with my reading!

Andrei

Sean Kelly wrote:
> On Feb 2, 2010, at 9:44 AM, Andrei Alexandrescu wrote:
> 
>> Just found this:
>>
>> http://portal.acm.org/ft_gateway.cfm?id=1400228&type=pdf&coll=portal&dl=ACM&CFID=76120516&CFTOKEN=63080657
>>
>> Very interesting. Confirms that we made the right bet by not committing to STM.
> 
> I think I forwarded that article to you when it was in CACM a while back :-)  Definitely a good read.  I think the correct implementation of HTM may have a chance though.  That "Bulk Multicore Architecture" article I linked a while back seems incredibly promising.
> _______________________________________________
> dmd-concurrency mailing list
> dmd-concurrency at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/dmd-concurrency