Actually the ZU3EG in the ultra96 is closer to an rpi3 than an rpi4. The rpi3 was a quad core A53 @1.2ghz w/1gb DDR2 the ZU3EG is a quad core A53 @1.5ghz (ultra96 has 2gb DDR4) The Rpi4 is a quad core A72 @ 1.5ghz w/ upto 4gb DDR4
The A72 cores are much better than the A53's. They are only slightly better than A57's but have better power efficiency. So this is basically an overclocked rpi3 with faster/more RAM. Which isn't bad for an socfpga, but the rpi4 will beat it outright all day. While the Mali GPU driver situation is not great at the moment, Mali GPU's are good, much better than videocore N's. Once the OS drivers are sorted (if they ever really are) Mali GPUs will be a blessing, not a curse. the ZU3EG also has 2 ARM R-5 "realtime" cores which sound like something fun for interfacing the FPGA, That might make it even better I/O wise but I doubt they have any use in application stuff. On Thu, Oct 3, 2019 at 4:29 AM Michael Brown <[email protected]> wrote: > Hi Steve > You forgot reply all and your message went off the list, So I have pasted > it below. > > Excuse me, I am still learning, but how do you compare a Zynq 7010 to a >> Ultrascale+. ? >> >> Steve C >> > > Well you don't that'a the point you can compare a the Zynq to a > Beaglebone or a rpi3. > > The Ultra96 is mpsoc (Ultrascale+ with mali 400) so that why I point at > the ZCU104 (similar cpu as rpi4). > Don't let the price fool you. > > On Tuesday, 1 October 2019 00:48:18 UTC+2, Michael Brown wrote: >> >> Firstly you can just run the U96 headless with a console image and ignore >> the mali gpu with Arms crappy non opensource driver support >> until mesa lima wayland support matures into being production ready. >> >> Secondly you have to compare this unique bargain in the right direction >> (not with a BB AI / rRPI4 or the like), but with >> something that has the same chip like this: >> >> https://www.xilinx.com/products/boards-and-kits/zcu104.html >> >> --- >> >> >> On Monday, 30 September 2019 23:47:03 UTC+2, justin White wrote: >>> >>> I want to like the ultra96 but for the price......I dunno. >>> >>> Not that it's bad, it's just the DE10-Nano fits the mksocfpga idea I had >>> in mind better. The whole Idea was to keep the graphics and everything off >>> the cpu running HAL and hal as close as possible to the FPGA. I do like the >>> framebuffer of the DE10 as it can serve as an all in one HMI, but where >>> performance matters I like the thought of a remote mk-cnc <=> mk-hal setup. >>> You'd be better off with this board (other than the form factor) either >>> way, but for $250 I dunno. >>> >>> >>> On Monday, September 30, 2019 at 11:34:03 AM UTC-4, Michael Brown wrote: >>>> >>>> >>>> On Monday, 30 September 2019 03:19:06 UTC+2, [email protected] wrote: >>>>> >>>>> What about the ultracheap ZYNQ boards coming from China like this one: >>>>> https://www.aliexpress.com/item/4000042572307.html?spm=a2g0o.productlist.0.0.ffa06f58V00L7M&algo_pvid=1673d04d-c50f-4962-9448-92686905272a&algo_expid=1673d04d-c50f-4962-9448-92686905272a-0&btsid=24d1b06e-03fd-466c-a7b1-dd52368fa2ad&ws_ab_test=searchweb0_0,searchweb201602_,searchweb201603_53 >>>>> < >>>>> https://www.aliexpress.com/item/4000042572307.html?spm=a2g0o.productlist.0.0.ffa06f58V00L7M&algo_pvid=1673d04d-c50f-4962-9448-92686905272a&algo_expid=1673d04d-c50f-4962-9448-92686905272a-0&btsid=24d1b06e-03fd-466c-a7b1-dd52368fa2ad&ws_ab_test=searchweb0_0,searchweb201602_,searchweb201603_53> >>>>> ? >>>>> >>>>> Specifications are low, but the price is very interesting. And for >>>>> Machinekit-HAL it could be enough. >>>>> >>>>> Cern. >>>>> >>>>> >>>>> Yes the price is alluring and that's about the only thing for what >>>> seems like a "cat in the box" board. >>>> Personally I would not go for a board without at least a template >>>> project (and/or a board definition file) and a viewable/readable schematic, >>>> downloadable up front. >>>> I these things some how are missing or faulty upon delivery or the >>>> board has (hidden)design flaws/issues, >>>> you are better off throwing your cheap money out of the window and >>>> spending your time on something else.... >>>> >>>> Michael B. >>>> >>> -- > website: http://www.machinekit.io blog: http://blog.machinekit.io github: > https://github.com/machinekit > --- > You received this message because you are subscribed to the Google Groups > "Machinekit" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/machinekit/1aa7d4f5-22c5-4503-ab95-fc53914b2ea1%40googlegroups.com > <https://groups.google.com/d/msgid/machinekit/1aa7d4f5-22c5-4503-ab95-fc53914b2ea1%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit --- You received this message because you are subscribed to the Google Groups "Machinekit" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/machinekit/CA%2BQ02MMkN78F8M8QHmMsjCq%3DwbpBDV%2BteEXRF-2Wc4R6JbEZ9g%40mail.gmail.com.
