|
Perhaps there is a better thread for this, but is there any news on general availability of Knight's Corner/Intel MIC?
|
# ¿ Jan 26, 2012 20:51 |
|
|
# ¿ May 9, 2024 13:42 |
|
Star War Sex Parrot posted:Probably not any time in the near future, if I had to guess. It's dependent on Intel getting their 22nm 3D tri-gate process perfected, which Ivy Bridge seems to be the first commercial guinea pig for. Thats unfortunate, was hoping for a 75W supercomputer. Video cards take so much power... edit: Upon looking that up again, did I imagine 75W?
|
# ¿ Jan 26, 2012 22:32 |
|
theclaw posted:Late 2012. Expect power consumption and theoretical floating-point performance similar to Kepler. If nVidia's roadmaps are to be believed then Maxwell should blow that out of the water a few months later? IPP/MKL integration would be much easier to deal with than CUDA however.
|
# ¿ Feb 4, 2012 14:22 |
|
Standish posted:This is really cool, more of a big deal for servers than desktops though. FMA will also be pretty cool for some applications.
|
# ¿ Feb 19, 2012 04:42 |
|
calcio posted:And really, a ps2 port still!? No PS2 port means enabled USB which means no secure computers and no DoD sales.
|
# ¿ Apr 30, 2012 05:24 |
|
dpbjinc posted:I would like to point out that Thunderbolt does suffer from the same security problems as FireWire and ExpressCard, in that any computer that has the ports enabled and automatically loads drivers for new hardware can have its memory read and/or altered by a malicious device. It's not a major issue, since it only affects non-server systems that need full disk encryption, which are a vast minority of systems out there, but it is something to consider for organizations where that would be an issue (i.e. DoD and friends). In other news if you open up a computer and plug in a PCIe card you also get DMA access. And if you solder a different bios flash chip in you can bypass a boot password. DMA and guaranteed bandwidth is the entire point of those protocols.
|
# ¿ May 13, 2012 21:32 |
|
Install Gentoo posted:Firewire did this too. Also I'm pretty sure this could be done with PCMCIA and ExpressCard on laptops! Just imagine the security vulnerabilities of my currently set up MBP with a thunderbolt to expresscard bridge and two firewire expresscards installed! (all this to run 3 machine vision cameras off a laptop - someone needs to release a thunderbolt to PCIe bridge yesterday)
|
# ¿ May 14, 2012 01:34 |
|
Alereon posted:-E series CPUs exist solely for niche applications that are heavily multi-threaded or dependent on memory bandwidth, which really isn't a lot of things. Those are he applications that make the world go round though. Intel probably figures Sandy bridge EN/EP will be fine until haswell so there won't be anything to downgrade to the regular E part. Or something.
|
# ¿ Jun 30, 2012 06:35 |
|
Alereon posted:Almost nobody runs those applications on their desktop computer though, which is why it's a pretty tiny niche. Which is why I mentioned the EN/EP xeons. They won't hit until around the release of Haswell, right? That'd be around when we would see ivy bridge e; if ever.
|
# ¿ Jun 30, 2012 06:48 |
|
Doesn't seem unreasonable that each tick-tock set would have a new socket.
|
# ¿ Jul 2, 2012 03:44 |
|
HalloKitty posted:Isn't that tock-tick? Clocks go tick tock so I will write it that way. (you're of course right)
|
# ¿ Jul 2, 2012 19:46 |
|
movax posted:The consumer SKUs fall off relatively quickly, but they guarantee availability for products coming out of their embedded group, and I imagine the server SKUs may enjoy a little longer longetivity as well. For example, Intel told us we can buy embedded SKUs of say the i7-620 for at least the next eight years. And SNB-E is still the highest end consumer hardware. As discussed earlier IVB-E and related xeons won't be out for quite a while so there's plenty of SNB related processors still being sold. As for embedded, Intel will still sell you Pentium 3s if you want them.
|
# ¿ Jul 5, 2012 15:10 |
|
movax posted:Look at this Lexus of Z77 mobos. Tons of add-on goodies and a PCIe switch for x8/x8/x8/x8 graphics That little SSD is the most useless thing...
|
# ¿ Aug 15, 2012 02:02 |
|
HalloKitty posted:Surely DirectCompute and OpenCL can provide? I guess you mean a physics-specific framework based open one If theres anything the world needs it's more GPU frameworks.
|
# ¿ Sep 14, 2012 23:25 |
|
ohgodwhat posted:Have you ever considered it's possibly faster to incorporate those things on one chip and your concerns are really outmoded? The biggest issue these days really is memory bandwidth and moving everything as close as possible helps a lot.
|
# ¿ Nov 3, 2012 22:19 |
|
Shimrra Jamaane posted:So anyone who desires to build/upgrade a gaming PC is all set. I don't think that many gamer machines are Sandy Bridge E.
|
# ¿ Mar 22, 2013 04:29 |
|
Factory Factory posted:They'll also probably win a Nobel prize for physics, what with figuring out room-temperature superconductors to cool a 47W chip in a form factor that normally struggles with a third of that. 13" rMBP can have a 35W tdp processor in it. Thats probably about as close as you'll get.
|
# ¿ Mar 27, 2013 21:42 |
|
cstine posted:In 2007, we were talking junk like windows mobile 6.5, blackberry os 4, symbian, and still ancient junk like palmos. But that's all smartphones were back then, a phone with a terrible browser and terrible email support.
|
# ¿ May 17, 2013 20:41 |
|
Is that where intel landed on that one?
|
# ¿ Jun 10, 2013 13:56 |
|
InstantInfidel posted:You're right, there's nothing wrong with it. They just chose some very esoteric units for that and a couple other charts they had, and I definitely believe that they could have gotten across the same point much more succinctly and equally as effectively by linking to some of those charts as sources rather than dropping one in every couple of sentences. Regardless, it's a better article than I gave it credit for. Watt hours are not an esoteric unit, have you ever looked at a power bill?
|
# ¿ Jun 13, 2013 00:01 |
|
WHERE MY HAT IS AT posted:Where's a good starting point to learning about various CPU architectures and how they work? My program is more focused on the software side, and we learn the basics of CPUs and how they work (like I know about logic gates and binary adders and APUs and such), but I find it all very interesting and would love to learn more. Any good resources besides signing up for summer courses at school? Whats your major? If you have a computer engineering program there should be a juniorish level computer architecture class that should require you to write your own CPUs in verilog or something along those lines. If you're CS you might get to count it as an elective.
|
# ¿ Jul 18, 2013 17:52 |
|
Is that thing the target for star citizen?
|
# ¿ Jun 6, 2018 18:29 |
|
BIG HEADLINE posted:The new rumor is eight-core Coffee Lake by September: http://www.guru3d.com/news-story/intel-8-core-coffee-lake-to-be-released-in-this-fall.html Is it time to reverse Intel/AMD on the cores/cores comic?
|
# ¿ Jun 13, 2018 12:48 |
|
mewse posted:I was watching old youtubes and apparently David Letterman not only banged one of his staff, when a dude tried to blackmail him he spoke about it on his show and the blackmailer was convicted Well yeah, asking them for money is illegal, asking a paper for money is however perfectly legal. It’s clear what the law wants you to do!
|
# ¿ Jun 21, 2018 14:30 |
|
LRADIKAL posted:Believe it or not 90 mv is .09 volts for everyone. he didn't give units so clearly he meant decivolts
|
# ¿ Jun 26, 2018 21:35 |
|
SourKraut posted:Nokia possibly? Ericsson is the other possibility. Nokia market cap appears to be around 31b, Ericsson 25b. Siemens, Qualcomm and Samsung are significantly larger so are right out. (In addition to more obvious reasons why it wouldn’t be them) hobbesmaster fucked around with this message at 21:40 on Jul 4, 2018 |
# ¿ Jul 4, 2018 21:38 |
|
suck my woke dick posted:"don't you think it's a bit risky betting the entire business on intel getting 10nm out roughly in time" Whats curious is that they're not concerned about intel 5G chipsets in general. https://newsroom.intel.com/news/intel-introduces-portfolio-new-commercial-5g-new-radio-modem-family/
|
# ¿ Jul 5, 2018 02:19 |
|
evilweasel posted:that is, uh, not a statement that fills me to bursting with confidence "we managed to successfully complete one telephone call without the modem breaking" Its not an uncommon thing to see in cell phone industry publications. It means that they have working phones and base stations for the new technology so its ready for pilot deployments. The real question is what is intel's yield...
|
# ¿ Jul 5, 2018 15:50 |
|
wargames posted:these things require silicon right? So probably using their 14nm+++++++++++ node so might not be terrible. Yeah current LTE-A modems use 14nm: https://www.intel.com/content/www/us/en/wireless-products/mobile-communications/xmm-7560-brief.html Cellular modems usually use older nodes so I'm not sure why delays in 10nm screw them over. Unless it was designed for 14nm equipment that is not going to be available because the 10nm stuff isn't online?
|
# ¿ Jul 5, 2018 18:33 |
|
Methylethylaldehyde posted:5g stuff needs 10nm or better because of the hilarious power requirements needed to drive the system. A 5g SOC is like twice as power hungry as current SOCs, and almost all of it is in the actual modem. Same with the base station stuff, a LOT of the gear is so close to cutting edge they desperately need the extra 40% power savings in order to avoid having the telco racks sound like the mid 90s 1U pizza box servers. Its infrastructure stuff so mid 90s 1 U pizza box server should be ok right I didn't realize the power density was that much higher for 5g, but I work with the slower side of the cell industry.
|
# ¿ Jul 5, 2018 19:59 |
|
Combat Pretzel posted:Intel prohibiting benchmarks and comparisons on their newest microcode release fixing various of the security issues? This is in the real release, not just engineering samples?
|
# ¿ Aug 23, 2018 16:59 |
|
Paul MaudDib posted:It's kind of annoying that people talk about resolutions being easy or difficult for a CPU to drive, when it's really the framerate that matters. 120 Hz 3440x1440 is not any easier to drive than 120 Hz 1920x1080, what matters is the refresh rate you're trying to drive. 3440*1440*120 > 1920*1080*120, I'm not sure what you're getting at?
|
# ¿ Sep 18, 2018 17:40 |
|
YOY revenue growth is bigger but esports are “only” a billion dollar industry That’s a couple NFL teams
|
# ¿ Oct 8, 2018 15:21 |
|
VulgarandStupid posted:Who buys an SA account in 2017? People that were banned. Maybe AMD could hit that 30% a year from now if Intel continues doing terribly. Maybe.
|
# ¿ Oct 15, 2018 15:52 |
|
jisforjosh posted:Mesh on low is actually the "try hard" help you see people thing apparently. Mesh affects rendering distance of certain objects but they supposedly changed the way the setting works from BF1->BFV. I think on Low, certain buildable coverage is not rendered but players and vehicle models are. Ultra settings for pve, lowest usable settings for pvp. For another example; until a month or two ago in overwatch running below 100% render scale and on low settings would increase the size of the team color outline on players which can help a lot with target recognition. Overwatch also renders a lot of extra “stuff” on higher settings that isn’t actually there; usually it’s just in spawns but sometimes you can see through a plant on low settings but not on high. Tournaments often (and LANs almost always) enforce uniform graphical settings for this reason. For example epic requires max settings to be used for fortnite tournaments.
|
# ¿ Nov 25, 2018 04:16 |
|
priznat posted:The newest Jetson (Xavier) has a Cortex A55 with 8 cores and it is pretty sweet. Have a couple dev kits at work, Gen4 x8 PCIe slot on it too! A busy box based user land is “full” Linux. You’re free to waste flash and RAM on a “real” user land if you want. All you need for a full Linux is a mmu.
|
# ¿ Feb 9, 2019 01:38 |
|
KKKLIP ART posted:I remember buying a Soyo Dragon for my socket a machine and thinking It was the hottest poo poo around. I am actually surprised that Soyo is still around today That iteration was liquidated in 2009 apparently. https://en.wikipedia.org/wiki/Soyo_Group
|
# ¿ Feb 9, 2019 23:45 |
|
craig588 posted:It might be caching or something. The Windows file transfer box regularly reads around 100MB/s so it's at least tricking whatever that measures. How are you connecting these drives to the rpi? This doesn’t make much sense
|
# ¿ Mar 23, 2019 01:57 |
|
craig588 posted:I have some generic USB adapters for 2 WD Reds and 1 Seagate Ironwolf. Nothing name brand, but it works for me with the type of consumer workloads I have. Of course if you have something specific in mind a dedicated box will be better, but to just appear as one big drive that doesn't know anything that I move single multi gigabyte files to it works great. A rpi doesn’t even have gigabit Ethernet so uh yeah that’s a pretty massive mistake. hobbesmaster fucked around with this message at 03:17 on Mar 24, 2019 |
# ¿ Mar 23, 2019 19:10 |
|
|
# ¿ May 9, 2024 13:42 |
|
Sidesaddle Cavalry posted:Hidden takeaway from this is that 6-core gets cheaper since it's no longer top of the consumer stack. That gets people into the multitasking game a lot easier AMD must be really excited after seeing the contents of these leaks. Unless zen 2 is a complete flop, but if Intel is standing still and the current ryzens are beating Intel's mid range...
|
# ¿ Apr 8, 2019 15:24 |