The final twenty projects for the five categories in the Hackaday Prize have been announced, for Open Hardware, Robotics, Power Harvesting, Interfacing and Music. Those topics cover a gamut of projects, from building your own motion tracking system through running electronics off of the energy leaked from your microwave to setting up a semiconductor lab in your garage. As these designs are all open source and part of the competition was to create detailed build instructions you can look through all the submissions for ideas of your own, or a useful project to build for yourself.
"Over the last eight months, we’ve been deep in the weeds with this year’s Hackaday Prize. It’s five challenges, with twenty winners per challenge. That’s one hundred projects that will make it to the semifinals in the hopes of becoming the greatest project this year."
Here is some more Tech News from around the web:
- Intel CPUs Impacted by New PortSmash Side-Channel Vulnerability @ Slashdot
- Cisco firewalls under attack – and there's no patch: Too many SIPs and they drown in data @ The Register
- Diablo Immortal on phones, WarCraft 3: Reforged top BlizzCon’s announcements @ Ars Technica
- Swedes grumbling about Apple Store in their park are lucky – in Toronto, Google eats all your data @ The Register
PortSmash is another example
PortSmash is another example that Intel Traded Security for Speed in Intel’s Implementation of SMT(Branded as HyperThreading). But that proof of concept code only works on Intel’s CPUs currently with other CPUs from other makers yet to be found vulnerable.
You just Know Intel would in the short term galdly give up on its Hyperthreading(TM) version of SMT just to sell more CPUs wihtout SMT enabled and maybe that’s the whole Idea of these Security Researchers if they can find any other maker’s SMT to be affected.
AMD’s implementation of SMT may be found to be a little more secure and Intel would have to invest more time in fixing its current and future CPU micro-archs than Intel would like to have to spend. So having the entire server market have to give up on SMT would benefit Intel some more against AMD for any server workloads that like SMT enabled.
I guess that Intel’s fear of Epyc/Rome is palpable and Intel really will try any dirty tricks in the book, and even others that have yet to make it into the dirty tricks book, to keep its doninance in the server market.
If Intel’s version of SMT, Hyperthreading(TM), is found to not be easily fixed in the short term and with Intel’s CPUs having some many other security issues maybe look to Intel/co-conspirators to favor turning off SMT Industry Wide to favor any of Intel’s few non SMT advantages. But that’s Only until Intel can get more time to actually fix its x86 micro-arch and then Intel will not care again.
Intel is really afraid of AMD’s Epyc/Rome and even more cores/threads than Intel will have the time in the short term to deal with. And we all know that Intel will do anything to keep market share via the best friends that Intel’s money can buy! AMD’s first generation Epyc/Naples is still bringing the design wins in lately with its better price/performance metrics.
And this Intel CPU shortage is being worked by Intel to Intel’s advantage in retaining Intel’s higher margins in the short term. But that will only help Intel’s stock value to keep it from shrinking in the short term and cost Intel more market share to AMD/others. Intel will be forced to reduce its markups in the longer term if AMD continues to take more server market share because Epyc/Naples is pretty damn close to Intel in performance with Epyc/MB Platform having better PCIe lane counts(128) and total memory channels(8 channels per socket) standard across all of the Epyc 7000 series of SKUs.
Intel will have to forgo those mid 60% gross margins at some point if Epyc/Naples continues to take market share and then Epyc/Rome will begin to ramp 2H 2019 into 2020 and beyond. Intel will also have to contend with Power9 and Power10 on the high end as well as the ARM server makers products on the lower end also. So that’s more downward pressure on Intel’s very high gross margins to keep more of that server market share with Intel’s Pricy kit inside.
Intel’s Management is currently loving that chip shortage because it keeps Intel’s quarterly qross margin figures looking all rosy. But there is that other server market share metric than will be of more concern to Intel and the investment community(Wall Street) and the ony way to retain market share in a very competative market is to lower the markups to match the competition’s price/performance metric more closely! High Margins in the face of lower cost competition will result in lost market share in the longer run and more potential lost revenues with every percantage point lost in srever market share.