|
DDR3L is lower voltage, so you might need new RAM, maybe not. Not sure if the same processors will get better performance with DDR4 or DDR3 of comparable clocks and spec throughput with these.
|
# ? Apr 28, 2015 22:13 |
|
|
# ? May 8, 2024 21:23 |
|
Specifically DDR3L is 1.35v. They've been selling it for a while so you might be good.
|
# ? Apr 28, 2015 22:15 |
|
Combat Pretzel posted:Ah for gently caress's sake. The i7-6700 non-K with the extended virtualization features is clocked pretty much the same as my i7-2600, whereas the i7-6700K, which I'd rather like, doesn't come with VT-d and poo poo. (--edit: Then again, I used VT-d only while experimenting with Xen and Linux' KVM. Keep hoping that PEG passthrough gets more reliable... any day now.) A 6700 has to have a significant IPC improvement over a 2600, no?
|
# ? Apr 28, 2015 22:16 |
|
Combat Pretzel posted:Ah for gently caress's sake. The i7-6700 non-K with the extended virtualization features is clocked pretty much the same as my i7-2600, whereas the i7-6700K, which I'd rather like, doesn't come with VT-d and poo poo. (--edit: Then again, I used VT-d only while experimenting with Xen and Linux' KVM. Keep hoping that PEG passthrough gets more reliable... any day now.) That I7-K would be a serious threat to Xeons and selling 300$ X mainboards, can't have that now.
|
# ? Apr 28, 2015 22:25 |
|
repiv posted:Specifically DDR3L is 1.35v. They've been selling it for a while so you might be good. KillHour posted:A 6700 has to have a significant IPC improvement over a 2600, no?
|
# ? Apr 28, 2015 22:26 |
|
Combat Pretzel posted:I just checked, mine are 1.5-1.65v. Meh. New RAM it is, hope DDR4 drops in price a lot. What are you even using the VT-d for on a home machine? sauer kraut posted:That I7-K would be a serious threat to Xeons and selling 300$ X mainboards, can't have that now. E3 Xeons tend to be better deals than non-K i7's anyways, assuming your motherboard supports them (many do). KillHour fucked around with this message at 22:32 on Apr 28, 2015 |
# ? Apr 28, 2015 22:27 |
|
KillHour posted:I thought the current Broadwell chips already have much better overclocking headroom than the Haswells did? The base and turbo frequencies are so high these days that there's not really as big a percentage improvement through OCing compared to Nehalem (2.6->3.8 on the 920) or Sandy Bridge (4.7-4.8ghz). I have a hard time thinking that you'd be able to OC these chips much beyond 4.8 if at all, and when the top turbo bin is already 4.2, that's just not much difference.
|
# ? Apr 28, 2015 22:33 |
|
KillHour posted:What are you even using the VT-d for on a home machine?
|
# ? Apr 28, 2015 22:36 |
|
KillHour posted:What are you even using the VT-d for on a home machine? Combat Pretzel posted:Mostly just experimenting with PCIe passthrough. Still trying to get a VM to boot with hardware accelerated graphics. Worked somewhat when I had an ATI card, but uptimes were rather short.
|
# ? Apr 28, 2015 22:52 |
|
When are they going to stop making new Devils Canyon processors? I might want to pick up an i7-4790K while I can.
|
# ? Apr 29, 2015 00:08 |
|
God drat it, I just bought some 1.5V DDR3 RAM thinking I would be able to put it into a new Skylake computer when the time came.
|
# ? Apr 29, 2015 01:01 |
|
Best case scenario: overclocking enthusiast Z170 boards will have UniDIMM slots and be able to overvolt RAM and allow normal DDR3 sticks to run anyways. Worst case scenario: none of the enthusiast boards will support DDR3L. It's still too early to say, folks. No use in trying to futureproof now if you weren't lucky enough already.
|
# ? Apr 29, 2015 02:10 |
|
Finally buying that low voltage DDR3 ram 4 years ago is paying off in spades. Suck it big ram.
|
# ? Apr 29, 2015 02:17 |
|
As far as I understand it, UniDIMM is supposed to be for SODIMMs only and incompatible with both standard DDR3 and DDR4 so you won't be using your old memory like that.
|
# ? Apr 29, 2015 02:17 |
|
This is the part where I get down and pray that mobo makers hack up something nice. Goofy board with eight RAM slots, anyone?
|
# ? Apr 29, 2015 09:29 |
|
Sidesaddle Cavalry posted:This is the part where I get down and pray that mobo makers hack up something nice. Goofy board with eight RAM slots, anyone? Costing enough over the standard model that you save like over just buying DDR3L or DDR4 RAM
|
# ? Apr 29, 2015 10:13 |
|
I remember buying such a board a decade ago when making the switch from DDR to DDR2 at the same time when the AGP to PCIe transition was ongoing. I don't think it was especially expensive.
|
# ? Apr 29, 2015 10:30 |
|
Grapeshot posted:As far as I understand it, UniDIMM is supposed to be for SODIMMs only and incompatible with both standard DDR3 and DDR4 so you won't be using your old memory like that. My quiet hope was that we'd switch to the so-dimm format for desktop boards as well. It would simplify a lot of stuff and DDR4 is as good a time as ever.
|
# ? Apr 29, 2015 14:06 |
|
Why do we still have the big DIMMs for desktops, anyway? Are there any technical advantages, or is it just inertia?
|
# ? Apr 29, 2015 14:13 |
|
Grim Up North posted:Why do we still have the big DIMMs for desktops, anyway? Are there any technical advantages, or is it just inertia? Uh for blinking LEDs, duh
|
# ? Apr 29, 2015 14:21 |
|
Hace posted:Uh for blinking LEDs, duh Don't forget the ridiculous heat spreaders!
|
# ? Apr 29, 2015 14:22 |
|
Grim Up North posted:I remember buying such a board a decade ago when making the switch from DDR to DDR2 at the same time when the AGP to PCIe transition was ongoing. I don't think it was especially expensive. I had one of those early Asrock socket 939 boards that had agp and PCI-e during that transition. I can't remember it being that great, but it was out at least.
|
# ? Apr 29, 2015 14:25 |
|
KillHour posted:Don't forget the ridiculous heat spreaders! I will never understand how people can think random metal spikes on their RAM are cool.
|
# ? Apr 29, 2015 16:20 |
|
Grim Up North posted:Why do we still have the big DIMMs for desktops, anyway? Are there any technical advantages, or is it just inertia? I always assumed it was cheaper/easier for the RAM manufacturers.
|
# ? Apr 29, 2015 16:22 |
|
Nintendo Kid posted:I always assumed it was cheaper/easier for the RAM manufacturers. On the other hand, two production lines vs. one. Laptop RAM isn't outrageously expensive either.
|
# ? Apr 29, 2015 16:24 |
|
Grim Up North posted:Why do we still have the big DIMMs for desktops, anyway? Are there any technical advantages, or is it just inertia? The supposed trend to go towards "micro" servers that might use SO-DIMMs that some industry pundits tried to push back in 2012 as "the next thing for servers" didn't happen. Instead, Intel just made their Xeon lines a lot more power efficient and the raw performance loss going down to those Moonshot servers or whatever just wasn't worth it for pretty much anyone. If companies like Google and Amazon aren't using it for their massive farms, it is likely not as economically efficient as advertised.
|
# ? Apr 29, 2015 16:37 |
|
necrobobsledder posted:The sheer density of chips from high density compute server lines on DIMM boards cannot be achieved in SO-DIMM form factors, and servers are not going away - they'll go away after SO-DIMMs I would argue (no more laptops or small nettops being made, that is). You try putting 32 chips of the same size as what's on a typical DIMM now into a SO-DIMM and see how well that turns out. I was referring just to desktop. Server DIMM requirements are so far removed from desktop that there is effectively no overlap anyway. Besides, the cap intel puts on the desktop cpu memory controller (32GB total, 8GB per DIMM) makes it worthless for any serious server usage.
|
# ? Apr 29, 2015 17:28 |
|
Grim Up North posted:I remember buying such a board a decade ago when making the switch from DDR to DDR2 at the same time when the AGP to PCIe transition was ongoing. I don't think it was especially expensive.
|
# ? Apr 29, 2015 18:01 |
|
EoRaptor posted:I was referring just to desktop. Server DIMM requirements are so far removed from desktop that there is effectively no overlap anyway. vvv I am enlightened! Sidesaddle Cavalry fucked around with this message at 01:05 on Apr 30, 2015 |
# ? Apr 29, 2015 19:41 |
|
Sidesaddle Cavalry posted:I can't answer that question directly, but I can hypothesize that it wouldn't solve the issue of two production lines like blowfish mentioned. I understand there's a difference between ECC and non-ECC RAM, but memory makers would still need to assemble for two different form factors. We weren't talking only DIMM production lines, more that you'd stop producing a line to desktop DIMMs in favour of producing extra laptop SO-DIMMs. You've just saved a bunch of design, testing, and validation. There would be some savings in the retail channel, as server dimms are pretty rare there already, so the removal of an entire product line (desktop dimms) would clear inventory space and reduce inventory management. You'd also make consumer MB design slightly easier, as the space requirements for memory slots would go down. If you think that the design/testing/validation for server dimms shares anything with desktop dimms, be assured it doesn't. Beyond any ECC requirements, the DIMM itself needs to be much, much stricter on electrical tolerances to keep EM noise down, so larger banks of dimms (slots) can all be populated without errors creeping in. Even though it shares a basic shape with desktop dimms, there really isn't any relationship between them once you begin the design process.
|
# ? Apr 29, 2015 20:07 |
|
The validation process is one thing but when chugging stuff through an assembly line, form factor does matter at least from a PCB form factor and machining perspective. The validation between RDIMMs and UDIMMs alone probably dwarfs that factor far more than between desktop DIMMs and UDIMMs. In any case, simply getting rid of SO-DIMMs is not really feasible given so many desktops are still being produced for OEM channels that you're just kind of stuck with SO-DIMMs, DIMMs + UDIMMs, and RDIMMs as your production lines. Getting rid of things costs money too, and Lenovo, Dell, and HP make too many drat desktops still for dinosaur enterprise companies or low-end small businesses that still buy tower machines for end users out of probably stupid reasons. The day we replace those machines with some stupid Intel NUCs will be a good day for all if you ask me. If any DRAM production line would disappear first, I would think UDIMMs because that's the "prosumer" line and it's already somewhat cheaper in some respects to just buy RDIMMs, enjoy your cheaper RAM (why 8GB UDIMMs cost more than 8GB RDIMMs now when it didn't use to be is beyond me), and suck up that you should be buying E5 CPUs with corresponding expensive motherboards. For me, the day I buy an E5 machine for a desktop is the day I just throw my hands up, buy a Mac Pro, and whine like a bitch. I'm still far more likely to buy an E5 in a server and if I need anything more I'll write some software to farm jobs out to the Cloud. It's really curious how the E3 line still exists given the market segmentation trends by Intel. I thought for a while Google and Amazon may be using them as stupid cheap machines but evidently they all start at the lower end E5 machines for basic workhorse clusters (as I had hoped, honestly), so there goes that line of reasoning. It's not like AMD has enough volume to validate the entire existence of UDIMMs either.
|
# ? Apr 30, 2015 01:17 |
|
necrobobsledder posted:between desktop DIMMs and UDIMMs.
|
# ? Apr 30, 2015 01:33 |
|
Mr Chips posted:what do you mean? A 'desktop' DIMM is an unbuffered DIMM Maybe unbuffered memory which is ECC?
|
# ? Apr 30, 2015 01:48 |
|
manufacturers/vendors are usually explicit if unbuffered RAM is also ECC.
|
# ? Apr 30, 2015 02:15 |
|
The standard nomenclature I've seen is that UDIMM implies unbuffered with ECC. However, there's no such thing as "unregistered" DIMMs either. So there's RDIMMs for Registered DIMMs. Interestingly, I've also never seen anyone say RAM is "buffered" either in a somewhat formal spec. But to me as someone that's written logic circuits in HDLs, I use registers AS a buffer (as opposed to some filter or localized clock trees to avoid glitches from propagation delay), so I dunno why the terms are this way normally. It might help if I actually read about how to design DRAM but I kinda can't care enough to go that far.
|
# ? Apr 30, 2015 02:18 |
|
just to make things lovely and vague, plenty of OEMs will be selling 64-bit DDR3 modules they call UDIMMs.
|
# ? Apr 30, 2015 02:24 |
|
mobby_6kl posted:These are supposedly the Skylake models: Is this the first time that non-K versions have had completely different TDPs and frequencies?
|
# ? Apr 30, 2015 04:46 |
|
.
sincx fucked around with this message at 05:55 on Mar 23, 2021 |
# ? Apr 30, 2015 05:40 |
|
This doesn't seem to relate to intel at all at first, but give it a read. Apparently a machine called the Datapoint 2200 was the embryo of the x86 architecture. http://www.metafilter.com/149540/The-Texas-Instruments-TMX-1795-the-first-forgotten-microprocessor quote:n the late 60's and early 70's, the technology and market were emerging to set the stage for production of monolithic, single-chip CPUs. In 1969, A terminal equipment manufacturer met with Intel to design a processor that was smaller and would generate less heat than the dozens of TTL chips they were using. The resulting design was the 8008, which is well known as the predecessor to the x86 line of processors that are ubiquitous in desktop PC's today. Less well known though, is that Texas Instruments came up with a competing design, and due to development delays at Intel, beat them to production by about nine months. http://www.righto.com/2015/05/the-texas-instruments-tmx-1795-first.html quote:As well as rejecting the TMX 1795, Datapoint also decided not to use the 8008 and gave up their exclusive rights to the chip. Intel, of course, commercialized the 8008, announcing it in April 1972. Two years later, Intel released the 8080, a microprocessor based on the 8008 but with many improvements. (Some people claim that the 8080 incorporates improvements suggested by Datapoint, but a close examination shows that later Datapoint architectures and the 8080 went in totally different directions.) The 8080 was followed by the x86 architecture, which was designed to extend the 8080. Thus, if you're using an x86 computer now, you're using a computer based on the Datapoint 2200 architecture.
|
# ? May 14, 2015 05:21 |
|
|
# ? May 8, 2024 21:23 |
|
Some info on Skylake Xeons/Purley platform: http://wccftech.com/massive-intel-xeon-e5-xeon-e7-skylake-purley-biggest-advancement-nehalem/ Kinda wondering whether they'll be publicly selling the ones with FPGA integration or if that slide is just formalizing the customized chip production they've been doing for big rear end customers.
|
# ? May 26, 2015 01:44 |