|
A lot of it depends on volume. A wafer of desktop processors is worth a hell of a lot less than a wafer filled with POWER8 chips. Which is, in turn, worth less than a wafer of one-run engineering or boutique chips.
|
# ? Sep 4, 2014 19:18 |
|
|
# ? May 9, 2024 00:36 |
|
These are also from the early 80's probably and certainly worthless, I was just curious after seeing that box of them drop in the factory heh
|
# ? Sep 4, 2014 19:23 |
|
1gnoirents posted:These are also from the early 80's probably and certainly worthless, I was just curious after seeing that box of them drop in the factory heh Well, if we're assuming they magically have modern circuitry on them, they still wouldn't be worth as much - wafer sizes have increased massively since then. Larger wafers are worth more.
|
# ? Sep 4, 2014 19:32 |
|
Yeah they are certainly smaller. Way smaller, probably the diameter of a bagel
|
# ? Sep 4, 2014 19:35 |
|
roadhead posted:Thats the shape the crystal (yes, its a single large crystal) grows in. Then they slice it really thin. Its like a big crayon shape almost. Those interested in more should google/wiki the Czochralski process. I designed an industrial controller for one of these machines once and the wiki article seems reasonable to me.
|
# ? Sep 5, 2014 01:37 |
|
1gnoirents posted:lol, the people walking up, stopping, staring... Keep in mind that they're not just reacting to losing the contents of the FOUP. Fabs are clean room environments. I'm sure shattering so many wafers released a bunch of particles into the air. Would not be surprised if an accident like that has serious secondary effects.
|
# ? Sep 5, 2014 01:46 |
|
LeftistMuslimObama posted:I always wondered, why do they make the wafers circles instead of squares anyway? http://en.wikipedia.org/wiki/Spin_coating If you want a layer of something (like photoresist!) deposited uniformly and evenly across a surface, spinning it is a pretty good way to get there. Gets trickier when the wafers are not circular.
|
# ? Sep 5, 2014 04:07 |
|
Two recent stories I'm surprised haven't been discussed here in the Intel thread: ASUS has created a proprietary socket which lets them bypass Intel's voltage regulators for greater overclocking. http://wccftech.com/asus-oc-socket-examined-lga-2011/ http://www.tomshardware.com/news/asus-oc-socket-warranty-x99,27597.html Intel is now admitting that Broadwell is so delayed that it has crashed into Skylake in the schedule so they're now altogether dropping plans for some Broadwell desktop chips (edit: but they will still produce low-power and high-end/high-power Broadwell chips). http://arstechnica.com/gadgets/2014/09/lower-end-desktop-cpus-wont-get-broadwell-will-need-to-wait-for-skylake/ Rastor fucked around with this message at 21:10 on Sep 6, 2014 |
# ? Sep 6, 2014 21:04 |
|
Not really tragic. They could have done the opposite and delayed Skylake some more. --edit: Also, is the IVR really not that good as the socket articles mention?
Combat Pretzel fucked around with this message at 21:42 on Sep 6, 2014 |
# ? Sep 6, 2014 21:32 |
|
First I've heard of dropping the FIVR. Got a link for that? It seemed like a good idea when Haswell came out, but given the hoops that Core M had to jump through to fit it - with a cutout in the motherboard - I can see wanting to simplify. But de-integrating it also seems counter to trying to improve platform power. I also hadn't heard of the Asus OC socket. That's pretty crazy. I want to see that reviewed by [H]. As for the Broadwell desktop delays... Whatever. As much as it's nice to get on the latest and greatest tech-wise, after Sandy Bridge, Ivy was a wet fart, and I don't mind missing out on Haswell's wet fart. I want Skylake details now. And an updated roadmap - we've known of Skylake and Skymont for a while now, but nothing beyond them. I think when Ivy was released, we at least had a very general idea of the big changes with Haswell - FIVR and the GT3e GPU. E: The Skylake Wikipedia page is loaded with a lot more info than a month ago, but much of it unsourced or sourced from rumors taken as authoritative. Factory Factory fucked around with this message at 22:10 on Sep 6, 2014 |
# ? Sep 6, 2014 22:06 |
|
Factory Factory posted:First I've heard of dropping the FIVR. Got a link for that? It seemed like a good idea when Haswell came out, but given the hoops that Core M had to jump through to fit it - with a cutout in the motherboard - I can see wanting to simplify. But de-integrating it also seems counter to trying to improve platform power. Didn't Skymont get renamed Cannonlake?
|
# ? Sep 6, 2014 22:08 |
|
So it did. I missed that. Makes sense, given the patterns of the other names.
|
# ? Sep 6, 2014 22:14 |
|
And here's a third topic for discussion: Phoronix had an X99 motherboard go up in smoke and flames, then Legitreviews experienced a similar event which took out the $1000 processor as well. Phoronix was using an MSI X99S SLI Plus and Legitreviews had an ASUS X99 Deluxe.
|
# ? Sep 7, 2014 03:46 |
|
All this news bodes well for Intel finally being booted off of its high horse.
|
# ? Sep 7, 2014 04:05 |
|
These are all motherboards, I wouldn't be surprised if someone like jetway was building these for them and skating on low tier parts.
|
# ? Sep 7, 2014 04:09 |
|
Factory Factory posted:First I've heard of dropping the FIVR. Got a link for that? It seemed like a good idea when Haswell came out, but given the hoops that Core M had to jump through to fit it - with a cutout in the motherboard - I can see wanting to simplify. But de-integrating it also seems counter to trying to improve platform power. FIVR has some politics behind it. There's a school of thought that enjoys the large benefits in power efficiency that come from having FIVR (and all of those patents / the ability to place that IP in any product they make). There's also another that doesn't want to waste precious room on power conversion vs. being able to place more transistors.
|
# ? Sep 7, 2014 04:10 |
|
Rastor posted:And here's a third topic for discussion: Phoronix had an X99 motherboard go up in smoke and flames, then Legitreviews experienced a similar event which took out the $1000 processor as well.
|
# ? Sep 7, 2014 14:41 |
|
Combat Pretzel posted:If Legitreviews isn't wrong about their analysis of the failure, the FIVR seems to be a rather suboptimal thing. ASUS' OC socket would probably helped then, I wonder if they had it enabled. The legitreviews article doesn't have anything I'd call analysis and I don't see how what is there necessarily points the finger at FIVR. The only thing you know for sure from what was written is that motherboard power conversion components failed. FIVR needs an input voltage of about 1.8V, therefore you still have a power conversion step on the motherboard, and it's really not a lot different from the old ~1.0V core supplies on pre-FIVR boards. Except in that it's considerably easier to design for since current delivered to the CPU drops by almost a factor of 2 and regulation tolerances are relaxed too. The linked wccftech article on the Asus OC socket is nontechnical garbage, the socket simply cannot actually work by completely "bypassing" the FIVR. Maybe they've hacked its control loop to get more control over the final core voltage output so they can boost higher for extreme OC, but no more than that.
|
# ? Sep 7, 2014 22:31 |
|
Rime posted:All this news bodes well for Intel finally being booted off of its high horse. Transmeta will finally have their day in the sun
|
# ? Sep 9, 2014 00:56 |
|
Doesn't look like Intel is going to be holding back Skylake in favor of letting Broadwell have a full product cycle. Instead, they demo'd a laptop running 3DMark on pre-production Skylake silicon. Dev kits go out 1Q15, production ramps up 2H15. Skylake is promised to be a significant improvement over Broadwell in performance, battery life, and efficiency in general. Running 3DMark for the demo kind of hints that GPU is where the focus will be, which is fine I guess. Talking about 4K screens and whatnot. I wonder if it won't be the end of overclocking. I'm fuzzy on timeline in fabrication - It takes months to build up a good selling stock of dies, right? And multiple weeks for one wafer to be processed into dies? So depending on how aggressive Intel is, that could mean anywhere from 4Q15 to 2Q16 launch date for Skylake?
|
# ? Sep 10, 2014 00:25 |
|
I wonder if they are still intending on DDR3 / DDR4 hybrid boards, as the wiki claims.
|
# ? Sep 10, 2014 00:30 |
|
Factory Factory posted:I'm fuzzy on timeline in fabrication - It takes months to build up a good selling stock of dies, right? And multiple weeks for one wafer to be processed into dies? So depending on how aggressive Intel is, that could mean anywhere from 4Q15 to 2Q16 launch date for Skylake? And it's possible to demo Silicon loooooong before any of those processes are ready to kick off, so I'm not sure there's enough points to complete the curve.
|
# ? Sep 10, 2014 00:33 |
|
Skylake brings SATA Express with-it and even this alone is pretty tempting...
|
# ? Sep 10, 2014 01:08 |
|
Researching my future NAS some more, it seems like I can get an ECC capable mini-ITX mainboard and an Xeon E3-1220V3 for just a few more than the four core Avoton. So why exactly would I want that Atom CPU?
|
# ? Sep 10, 2014 02:52 |
|
Combat Pretzel posted:Researching my future NAS some more, it seems like I can get an ECC capable mini-ITX mainboard and an Xeon E3-1220V3 for just a few more than the four core Avoton. So why exactly would I want that Atom CPU? If you'll be deploying a lot of them in a high density environment. For home server use the E3 is probably the better bang for the buck, even with the higher power consumption. It'll be much faster with tasks like transcoding and running VMs if those are things you think you might want to do at some point. You could also hit a lower price point and still get ECC support with an i3.
|
# ? Sep 10, 2014 03:07 |
|
JawnV6 posted:And it's possible to demo Silicon loooooong before any of those processes are ready to kick off, so I'm not sure there's enough points to complete the curve. Demo silicon that's about 90% as awesome as production volume silicon is generally available 6-18 months before production, depending on pretty much everything from node complexity to phase of the moon. Early yields of pre-production silicon are universally dogshit, but you still get a few usable CPUs off the wafer for debug, unit testing and general Foundry Alchemy. I'm still rocking my glorious 5 year old Lynnfield i7 860, and I probably won't upgrade until Skylake is a thing.
|
# ? Sep 10, 2014 03:11 |
|
Combat Pretzel posted:Researching my future NAS some more, it seems like I can get an ECC capable mini-ITX mainboard and an Xeon E3-1220V3 for just a few more than the four core Avoton. So why exactly would I want that Atom CPU? Because a lovely processor will obviously let you leverage lower pow... gently caress I can't even joke about this. You most likely don't ever want an atom cpu if you have a choice. I may as well ask here before I hit the linux thread, does anyone know a way in linux, on the cli to get the current cpu turbo clock / power state? I'm trying to figure out how hard a supposedly cpu bound application is hitting our servers, but I don't know how fast the cpu's are running (eventually I will cacti or snmp this). Ends up intel e5-26xx frequency optimized chips are expensive, so I don't want to switch to those if it won't make a difference. Luckily the Haswell version of the good old e5-2620 chip bumps base and turbo clock a fair amount for no additional cost, so I am hoping to squeeze by on that.
|
# ? Sep 10, 2014 04:49 |
|
Aquila posted:I may as well ask here before I hit the linux thread, does anyone know a way in linux, on the cli to get the current cpu turbo clock / power state? I'm trying to figure out how hard a supposedly cpu bound application is hitting our servers, but I don't know how fast the cpu's are running (eventually I will cacti or snmp this). Ends up intel e5-26xx frequency optimized chips are expensive, so I don't want to switch to those if it won't make a difference. Luckily the Haswell version of the good old e5-2620 chip bumps base and turbo clock a fair amount for no additional cost, so I am hoping to squeeze by on that. Information on each core, including its current clock speed, can be grabbed from /proc/cpuinfo. code:
code:
|
# ? Sep 10, 2014 05:01 |
|
syzygy86 posted:Information on each core, including its current clock speed, can be grabbed from /proc/cpuinfo. Oops, guess I should have actually checked that, I assumed it was informational or set at boot or something: cat /proc/cpuinfo |grep "cpu MHz" cpu MHz : 3401.000 cpu MHz : 1600.000 cpu MHz : 3401.000 Not sure why/how HT factors into this, I'm usually seeing one logical core of eight reporting 1600 or 2100 MHz on a four core chip (Intel(R) Xeon(R) CPU E3-1240 V2 @ 3.40GHz) e: clarified, this is from a single cat /proc/cpuinfo Aquila fucked around with this message at 05:18 on Sep 10, 2014 |
# ? Sep 10, 2014 05:10 |
|
Aquila posted:Because a lovely processor will obviously let you leverage lower pow... gently caress I can't even joke about this. You most likely don't ever want an atom cpu if you have a choice. My initial plan was to run the NAS as VM in my upcoming 5820K build (given that it'd idle at 60% power of my current computer, too), but since it is recommended to have ECC RAM for ZFS, that idea's out the window again.
|
# ? Sep 10, 2014 05:29 |
|
Why is Atom lovely?
|
# ? Sep 10, 2014 05:57 |
|
Eh, according to Intel ARK, the suggested prices of both the E5-1650v3 and the 5930K are the exactly same. I suppose this is a K and non-K style of situation?
|
# ? Sep 10, 2014 06:29 |
|
No Gravitas posted:Why is Atom lovely? Originally, because of intel's product segmentation. The new ones aren't really that bad (Avonton), especially if you were looking at doing something which required either plenty of RAM (64GB) or a lot of network interfaces (4+LOM) in a small form factor. Power use isn't terrible either - idling a Supermicro C2750 serverboard will use about half the watts of an E3-1220v2 on a mATX serverboard, and peak use is way lower with the C2750. The tradeoff is overall CPU performance, which appears to be about 1/2 that of the 1220v2. Mr Chips fucked around with this message at 07:43 on Sep 10, 2014 |
# ? Sep 10, 2014 07:39 |
|
Tab8715 posted:Skylake brings SATA Express with-it and even this alone is pretty tempting...
|
# ? Sep 10, 2014 15:02 |
|
Welmu posted:Not really: most users are well-served by using a single M.2 SSD in a 4x slot which is ~twice faster (with zero cables) than SATA-Express. Yes, you're limited to a single drive, but the SM951 comes in sizes of up to 1TB and supports NMVe to boot. Is there even a proper way to mount the M.2 SSD? Or are newer models still coming out? Not sure if I'm wildly missing the picture but aren't NVMe SSDs geared towards the enterprise and start a a bit over a grand?
|
# ? Sep 10, 2014 15:43 |
|
Combat Pretzel posted:Researching my future NAS some more, it seems like I can get an ECC capable mini-ITX mainboard and an Xeon E3-1220V3 for just a few more than the four core Avoton. So why exactly would I want that Atom CPU? Low power, supports twice as much RAM as the E3, pretty much a perfect choice for stuff like Ceph (and maybe even ZFS?)
|
# ? Sep 10, 2014 15:47 |
|
Tab8715 posted:Is there even a proper way to mount the M.2 SSD? Or are newer models still coming out? There are dedicated M.2 slots, mounting them isn't a problem. If your board doesn't have one, many folks make little adapter boards that are just M.2 slots directly wired to a PCIe slot on an expansion board for PCIe, or on a little 2.5" sled for SATA. God bless Startech and their tons of borderline-useless little adapters. NVMe is currently mostly an enterprise tech, but that's because it's brand new (and therefore expensive, so only enterprise is willing to pay for it) and the benefits of NVMe over AHCI are mostly felt in I/O heavy applications (i.e. in Enterprise). It'll trickle down. Almost every desktop innovation is a trickle-down server innovation from a few years prior.
|
# ? Sep 10, 2014 16:09 |
|
Factory Factory posted:There are dedicated M.2 slots, mounting them isn't a problem. If your board doesn't have one, many folks make little adapter boards that are just M.2 slots directly wired to a PCIe slot on an expansion board for PCIe, or on a little 2.5" sled for SATA. Ha, I had no idea they started putting M.2 on boards although there a little difficult to identify at first. NVMe looks like a great product but once it reaches consumers which looks to be soon what's the point of SATA Express?
|
# ? Sep 10, 2014 16:22 |
|
Tab8715 posted:Ha, I had no idea they started putting M.2 on boards although there a little difficult to identify at first. NVMe looks like a great product but once it reaches consumers which looks to be soon what's the point of SATA Express? SATA Express is the physical spec (the wires) NVMe is the host controller interface (the software API) You can use SATA Express with AHCI like regular SATA, but that would be artificially limiting the controller, so i suspect everyone will jump to NVMe since every OS supports it already.
|
# ? Sep 10, 2014 16:33 |
|
|
# ? May 9, 2024 00:36 |
|
Factory Factory posted:There are dedicated M.2 slots, mounting them isn't a problem. If your board doesn't have one, many folks make little adapter boards that are just M.2 slots directly wired to a PCIe slot on an expansion board for PCIe, or on a little 2.5" sled for SATA.
|
# ? Sep 10, 2014 16:36 |