00:17 <gruetzkopf> i still need a location to run the onyx2 contiuosly

00:17 <gruetzkopf> and hand out shells on irix6.5.30 (with licensed compiler suite)

01:01 <TD-Linux> >.sv

01:01 <TD-Linux> this kills the yosys

01:02 <TD-Linux> sorear, I don't think the R4000 is dual issue?

01:02 <TD-Linux> I guess if you count integer+float it is

01:03 <sorear> no, it's not, I misremembered the order

01:04 <TD-Linux> looks like swerv actually has 4 pipes

01:04 <TD-Linux> 2 integer, 1 load/store, 1 mult

01:05 <TD-Linux> but oly the integer pipes can write back so that's why they only call it dual issue

01:06 <sorear> I'd interpret "dual issue" as "two instructions can start in any cycle"

03:17 <cr1901_modern> sorear: What do you mean by "misremembered the order"?

03:17 <cr1901_modern> And why doesn't superscalar imply dual issue?

03:22 <sorear> cr1901_modern: TD-Linux last comment was about dual issue versus quad issue, both are superscalar

03:23 <cr1901_modern> oh oops

04:33 <azonenberg> yeah dual issue is a subset of superscalar

04:33 <azonenberg> i guess technically you could have a dual issue VLIW

04:33 <azonenberg> which is not superscalar because the instructions are scheduled at compile time vs run time

04:36 <sorear> I don't normally hear the word "issue" in that context, but my sample size of dsps is lacking

04:40 <azonenberg> yeah i just mean, a 2-lane-wide VLIW is clearly dual issue, it executes two instructions per clock

04:41 <azonenberg> But it's not superscalar because there's no hazard checking etc

04:45 <Bob_Dole> I started reading up, from this discussion mentioning VLIW, and at first thought the R4000 would be in reference to Radeon (HD) 4000.. and then read up to see, no, the MIPS R4000, and not RV770.