Gadgetory


All Cool Mind-blowing Gadgets You Love in One Place

HW News - Intel Shortage Ending, Ryzen 9 16-Core, & AMD Supercomputer

2019-05-14
everyone today we're doing a hardware news recap for the last week of computer hardware and some notes on the upcoming Computex trade show which is the biggest for our industry every year it is less than less than well it's a couple weeks away at this point and also recapping on the stream but beyond that rumors on risin because the new ones going up so those are going to be nonstop for a while Intel debuting seven nanometer process at some points to try and combat TSM sees five nanometer and then until also expecting a rebound in CPU shortages for a second half of 2019 before that this video is brought to you by EB J's a new audio sound card engineered by audio note EVGA a CEO knows high quality audio and has begun bringing sound cards back the new audio sound card is capable of delivering hair-raising audio superior to onboard sound the card includes a line in headphone line out and Mike in and a Sony Philips digital interface new audio also leverages the EVGA specv design experience has upgradeable op amps and uses AKM premium components for its DAC and ADC learn more at the link in the description below so for the quick GN store news item the media mod mats have actually been shipping out and the large as well since we got them back in a lot of you have bought the mediums thank you for picking those up we have a decent amount of stock on medium we're very low on large at this point but they aren't on stored on cameras access denied if you're interested the livestream results and Computex if you missed our about 8 hours of live streaming this past weekend we overclocked a risin 720 700 X with Joe Stefan Z aka bearded hardware and overclocked a kingpin EVGA 28 ET i card we have recaps on those streams on the channel for more detail but the very very short version is that the Rison 7 2700 CPU we got up to about 5.6 gigahertz nearly and scored pretty high in Cinebench as well we were approaching not quite there we were approaching low and thread Ripper like 1950 X last gen results so we were something like 500 points away from thread repair and then for the kingpin card the the 20 atti cane pin we ended up at almost 2,600 megahertz and for point of reference the base overclocks on most 2082 we'll end at about 20 52 maybe 21 15 if you're lucky so 2600 is quite high that's that's why we still have a lot of Ellen to because we're using it for those streams so if you do want to catch the short recaps with more detail on what we did what types of settings we applied you can find those on the channel but that's the short version if you wanted one so rumor up first there's actually been a lot of these last few days that's all about rise and first the the note on rumors we try not to pay too much attention to rumors or leaks of specs because a lot of the time they're wrong we try not to speculate especially but if it looks like it's credible we'll talk about it one thing that is not really at this point super crowd of although maybe possible is a it seems like so people are talking a lot about 5 gigahertz and rise in 3000 and there's kind of its nebulous what that means 5 gigahertz doesn't maybe like super XFR on who knows what future chip but realistically you're probably almost certainly not going to see 5 gigahertz all core so get that out idea out of your head now 5 gigahertz all core stock out of the box almost certainly not happening I just wanted to address that one right away because that one's that one's pretty easy to tackle and a lot of people do seem to be expecting that so expect smaller gains than that but beyond that we can talk about some of the other things here so rumors and speculation running rampant we know that there's going to be 16 corner 12 core GN talked about that after CES we were able to confirm it at CES the release date for those is to be determined there's going to be an 8 core rise in 7 SKU that is known fact and 12 and 16 core we know with pretty high certainty for several months now so those are happening but that was a I guess leaked a bit again and T um API sak on Twitter is the one who has been claiming that there's a 16 core engineering sample out in the wild and reporting a an alleged base clock of 3.3 gigahertz and an alleged abuse clock of 4.2 gigahertz so this coincides with a lot of what we've heard we know there's 16 cores at some point we've also heard well I'm not gonna give the frequency numbers we've heard because I don't know if they're fully accurate but not 5 gigahertz we have not heard that yet so it may be XFR on like a lower end lower core count chip or something maybe but anyway that'll be on X 570 we have information on those boards coming up at Computex for sure the motherboard event manufactures will almost certainly all have X 5 somebody there will be reporting on it to make sure you say stay subscribed you get subscribe to see our Computex coverage we will get there like May 22nd or something and then the show officially starts roughly the 27th and after Andy's keynote is when you can expect to see a lot of X 570 coverage as for the rest well stay tuned I guess the the biggest leak that we have right now is potentially 2.3 gigahertz base 4.2 boost and to be determined on how many cores that effects and that's for the 16 core it's also going to be a 12 core at some point okay NVIDIA is apparently going to stop dividing taurine into the a and non-ace queues this is something we reported on when dissecting the EVGA r-tx 2070 XE ultra we compared the TU 106 400 - a 1 and the T 106 400 a - a one dies and there was a pretty significant difference between those the separate Turing die is allowed and video to cherry-pick the best of the best out of the yields and offer them in the form of a dies for partners flagship and overclocked cards for an extra cost meanwhile Nvidia could offer the lower quality non aid thighs which Nvidia forbade pre overclocking on for lower priced cards down the stack users could of course I overclocked these cards themselves but AI be partners were not allowed to pre overclock them and it's highly unlikely that a non a die with overclock as well as it's been a die counterpart according to the Tom's Hardware report at the end of May the r-tx 2080 and 2070 cards will use only one variant of Turing and that's going to be the tu-104 410 no more a and there in tu 106 410 respectively this implies that TSM C is 12 nanometer FF n processes that the GPS are built onto is mature enough for optimum yield and in video no longer needs to separate the silicon out this is theoretically good news as new silicon should level the playing field in terms of price relative to the GPU users are buying it should also mean that we can expect higher quality taurine silicon in the future we'll see if nvidia plans to do the same thing with the 2080 TI and the TU 102 die in intel news the company revealed that it's 2019 major investor medium that it had was like a 4-hour thing on the phone that it will ship its first 10 nanometer CPUs in volume finally i slake this june and this also aligns with intel's plans to have ice lake based devices on shelves for a holiday of 2019 intel also anticipates a steady cadence of 10 nanometer products throughout 2019 and 2020 and then 10 nanometer agile x fpga is a 10 nanometer GPU at some point tiger lake and 10 nanometers norwich SRC aimed at 5g until will also be debuting at 7 nanometer process in 2021 competing with TSM C's 5 nanometer process and Tom's first seven animated product will be an Intel X Y or collector--base GPU following its first discrete GPU in 2020 Intel 7 animated process will be the first time they come into use is EUV or extreme ultraviolet lithography and Intel expects a two times improvement in terms of density over 10 nanometer Intel will also lean heavily into its EMI B and FiOS technology as 7 nanometre Intel finally also it plans to focus on intra node optimizations so this would be more plus steps we saw a lot of 49 or plus plus plus plus plus plus plus plus and despite getting kind of mocked to intel's credit driving 14 nanometer as far as it has is genuinely impressive they got a lot out of that process and it sounds like this will continue going forward so rather than just being a stopgap for 10 nanometer this will this will go into perhaps a 10 nanometer plus at some point in the future and it also helps avoid taking too much risk with new nodes going forward while also improving the scaling between generations and how much can be gotten out of each step in the process so Intel expects to deliver one Moore's Law game at the beginning of a node and then one at the final revision of a node am dnews now the worlds at current fastest supercomputers called sumit and it is set to be dethroned in 2021 with a joint venture between AMD and cray cray if you don't know was one of the original one of the og high-end or super computer manufacturers and we actually have some coverage of them from years ago when we visited the Mountain View Computer History Museum in California and we have that coverage on the channel if you just search for a computer history museum look for the one with Jim Vincent joining me we show the Cray supercomputers ants with the other Cray computers at that time so frontier is what the system's supposed to be called it is expected to be delivered to the Oak Ridge National Laboratory in 2021 and will be the world's first exascale supercomputer targeting 1.5 x of flops of compute performance this is it roughly five times some its current 200 petaflop some performance and the exact details haven't yet been confirmed but frontier will be leveraging and these epic CPUs Radeon Instinct GPUs and Andy's infinity fabric and craze slingshot technology it's an interconnect each node will feature one AMD FX CPU and then four Radeon Instinct GPUs so the quote from Steve Scott senior vice president and CTO at Cray is as follows we are excited to work with the team at and eats deliver the front tier system to Oak Ridge National Laboratory crazy Shasta supercomputers are designed to support leading-edge processor technologies and high performance storage all tightly interconnected by crazy new slingshot Network the combination of cray and AMD technology in the frontier system will dramatically enhance performance at scale for AI analytics and simulation enabling do-e to further push the boundaries of scientific discovery frontier will be deployed to study whether subatomic structures genomics physics and more different times has been reporting recently that's Intel and its supply chain vendors for notebooks are signaling Intel CB shortage will ease in June of 2019 the first time we've really heard anything of that sort for about a year now and this is as Intel expects to increase its shipments of entry-level processors so Intel has been aggressively focusing on the higher margin products it's also shifted some of its DIY processors to sis or system integrators and said the company has been focusing on top customers as well during the CPU drought and this change if they do overcome the shortage will be significant for pricing and for availability digitized reports that Intel has informed Notebook partners that it will begin shipping entry-level chips in June and while there will still be some shortages the shortages will greatly narrow notebook shipments are expected to rebound in the second quarter after a sluggish first quarter primarily riding on the hype for Intel and Nvidia's newest mobile chips install CPU shortage has driven many vendors and OEMs to put in orders with AMD however according to digitized sources vendors such as Dell HP and Lenovo are expected to be putting in more orders with Intel instead of AMD Intel CEO Bob Swan recently acknowledged Intel CPU troubles in an earnings call stating that the CP shortage would not be fully rectified until quarter three SWANA also vowed quote never again to be a constraint in Intel's customers growth last year Dell EMC CTO John Rose didn't seem to be overly impressed with Andy's epic platform and we have a quote they're saying AMD is doing some interesting things and by adding them to the portfolio we pick up a few extra areas but let's be very clear there's a huge dominant player in compute semiconductors and then there is a challenger which is doing some very good innovative work called AMD but the gap between them is quite large in terms of market share and use cases so our portfolio is not going to change in a meaningful way and Roche said this in the interview last year and then said don't expect it to be a duopoly any time soon now in a very contrasting statement dolls at Dominique Van Ham states that the company will essentially triple its epic server offerings and support the upcoming seven nanometer epic Rome products and this is got a quote as well that says out of let's say 50 or so platforms that we have today three of them are AMD and we'll probably triple that by the end of this year going up to nine in that instance so aside from the promises that seven nanometer brains and ham sights a high demand for epic from customers particularly in general-purpose markets still and these representation in dolls server portfolio will pale in comparison to Intel's again nine out of in that instance but it is growing and challenges aside and these epic adoption continue steadily as the company has been picking its shots wisely with cloud service providers we talked about that last week with AWS that's it for this one thank you for watching subscribe for more make sure you check back for copy tax coverage as we ramp into it we do have lots of GPU coverage and some reviews coming up this week you can subscribe to those go to stored on camera access dotnet support us directly by buying something like one of our mod mats or one of our shirts this is the GPU artifact insurance and patreon.com/scishow Saxons for behind the scenes videos thank you for watching I'll see you all next time
We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.