|
I don't think there's a dedicated RAM thread, and asking in the PC Builder thread will invite a ton of answers and infighting, so I'm asking here: What's the single best 2x16 kit (I know that locks me to ~3600Mhz max) I could get to pair with a 9900K, preferably for under $250? I ask because there's an opportunity for me to switch out my LPX 3200Mhz set. I'm eying the 3600 LPX set, but it's CAS 18, almost assuredly Hynix, and I don't know what else is out there. I've already checked Gigabyte's QVL and focused on the sets that are gettable with Samsung chips over Hynix, but I know that's not everything that's out there and I'd go off-QVL for a good set. EDIT: My general understanding is that 3200@C16 is practically identical speed-wise to 3600@C18? C/D? BIG HEADLINE fucked around with this message at 07:27 on Dec 12, 2019 |
# ? Dec 12, 2019 00:14 |
|
|
# ? May 28, 2024 08:53 |
|
If you already have 3200C16 that works at the advertised speed it would be a poor investment to push for something higher, you are talking single digits improvements outside of specifically targeted benchmarks. Also the higher the frequency or the lower the latency the more of a pain in the rear end it is to get it working even if it is in the QVL, I have 3200C14 and it took quite a while and a ridiculous 1.3v to both VCCSA and VCCIO to get it to kick over.
|
# ? Dec 12, 2019 01:28 |
|
If you have the 3200 c16 kit keep it. A 2x16 bdie kit at 3200 C14 is about as good as 2x16 gets until you get into the extreme binned kits, I think. I run 4x16 at 3533 c14 of 3200 c14 bdie. I'll probably play with timings and try to get up to 3600-3800, but it won't do it at 14 that's for sure.
|
# ? Dec 12, 2019 15:59 |
|
I have 2x16GB LPX 3333C16 and its samsung b-die and does 3600c16 at stock volts, that should be cheaper.
|
# ? Dec 12, 2019 16:05 |
|
Perplx posted:I have 2x16GB LPX 3333C16 and its samsung b-die and does 3600c16 at stock volts, that should be cheaper. Yeah, the problem with Corsair's kits is that even though their support is great, they 'version' the gently caress out of their SKUs, so one version might use B-Die while the other uses binned SKHynix. =/ EDIT: Ended up with this: https://www.newegg.com/g-skill-32gb-288-pin-ddr4-sdram/p/N82E16820232218 3200@C14, guaranteed B-Die. I'll make a game of it to see what more I can get. This same set was over $400 a year ago and they're not making B-Die anymore so . BIG HEADLINE fucked around with this message at 23:44 on Dec 12, 2019 |
# ? Dec 12, 2019 17:04 |
|
BIG HEADLINE posted:Yeah, the problem with Corsair's kits is that even though their support is great, they 'version' the gently caress out of their SKUs, so one version might use B-Die while the other uses binned SKHynix. =/ Apparently B-Die memory is still produced. I think what got EOL'd was Samsung making the entire stick on their own. More B-Die is in the channel than ever at reasonable prices.
|
# ? Dec 13, 2019 07:25 |
|
https://twitter.com/atomicthumbs/status/1203771427679166464?s=20
|
# ? Dec 13, 2019 08:15 |
|
https://twitter.com/TUM_APISAK/status/1206925602168590336 Tiger Lake / Willow Cove cores in the wild. 4/8 mobile CPU on 10nm++, Willow Cove is the replacement for Sunny Cove thats used in Ice Lake, so basically 2 full generations ahead of the desktop parts. Still obviously an early ES, but even at 1.2ghz it beats the retail 6/12 Comet Lake 10710U in both single and multi thread scores. Rumor is this is the core deign that will be backported to 14+++ and become Rocket Lake for desktop.
|
# ? Dec 17, 2019 20:04 |
|
When putting Haswell-series Xeons on Z97 boards, can you adjust the power limit?
|
# ? Dec 18, 2019 00:00 |
|
Beautiful Ninja posted:Apparently B-Die memory is still produced. I think what got EOL'd was Samsung making the entire stick on their own. More B-Die is in the channel than ever at reasonable prices. I'd say it is, seeing as the build date of my CAS 14 3200Mhz G.Skill sticks was "Dec 2019." I'd *really* love to take the spreaders off and check the memory itself - and also replace them with something better since the biggest gripe I've heard about them is that the Ripjaws' spreaders don't make great contact with the modules.
|
# ? Dec 18, 2019 01:12 |
At the risk of veering dangerously into GPU territory, AnandTech just published an analysis of the Xee disclosure.
|
|
# ? Dec 24, 2019 22:46 |
|
D. Ebdrup posted:At the risk of veering dangerously into GPU territory, AnandTech just published an analysis of the Xee disclosure. No worries, definitely germane. Still desperately hoping to see some kind of fusion / actual utilization of the Altera acquisition but the effort they’ve put into the software ecosystem here for Xee is commendable. If there’s anything the past two decades have shown, the “build hardware and wait for software” paradigm doesn’t end well.
|
# ? Dec 25, 2019 03:53 |
|
movax posted:No worries, definitely germane. Cautiously optimistic for a decent Cuda competitor now that someone with a ton of money can buy ports of software
|
# ? Dec 26, 2019 21:44 |
|
Eletriarnation posted:I think the original Moore's law was just about transistor density in a given area of integrated circuit, as the article was titled "Cramming more components onto integrated circuits". All the other conclusions followed from that given some assumptions which were valid at the time. Those assumptions, e.g. "the cost per wafer of size X for process Y is similar to the cost of a size X wafer from process (Y+1)" are not necessarily valid anymore. Obviously, Moore's Law became all about transistor and wire shrinkage later, but I don't think the original article really emphasizes transistor & wire miniaturization. When I read the original article, I got the impression that Moore's thinking at the time (the 60's) was that the way that the cost/function was going to go down and (# of devices)/circuit was going to go up was through improved yields due to better manufacturing know-how, and through larger die sizes (which would not be necessarily low-yielding, due to proportionally greater improvements in manufacturing know-how). See below quote from the article, with the last part in bold which reveals this thinking, in my opinion: quote:The complexity for minimum component costs has increased at a rate of roughly a factor of two per year (see graph on next page). Certainly over the short term this rate can be expected to continue, if not to increase. Over the At one point in the original article he mentions transistor & wire miniaturization improving the power-delay figure of merit, but I think this is the only reference to miniaturization in the article: quote:In addition, power is needed primarily to drive the various lines and capacitances associated with the system. As long as a function is confined to a small area on a wafer, the amount of capacitance which must be driven is distinctly limited. In fact, shrinking dimensions on an integrated structure makes it possible to operate the structure at higher speed for the same power per unit area. silence_kit fucked around with this message at 10:37 on Dec 27, 2019 |
# ? Dec 26, 2019 21:59 |
|
https://videocardz.com/newz/leaked-slide-confirms-intel-10th-gen-core-k-series-are-125w-tdp
|
# ? Dec 27, 2019 18:05 |
|
Hm, no word on the socket. Lately there have been rumors that CML will be on LGA1159 after all and the new socket is coming later (with Tiger Lake?). e: pincount typo eames fucked around with this message at 11:51 on Dec 30, 2019 |
# ? Dec 28, 2019 00:31 |
|
With the amount of leaks of the new 1200 socket, including complete designs and part numbers, I think its coming sooner rather than later. On desktop, after Comet Lake is Rocket Lake, another 14nm part with DDR4 in 2021. Seems pretty likely to me that Intel will keep the 2 year socket cadence going with Comet/Rocket on Socket 1200 (H5), before moving to a new socket in 2022 for 7nm Meteor Lake/DDR5... if you believe the rumors and that Intel can actually ship 7nm on schedule. And when you think about that, it means that not just 10nm is getting skipped on desktop, but also both Sunny Cove (Ice Lake) and Willow Cove (Tiger Lake) core designs, skipping straight to Golden Cove. Desktop would go from 2015 to 2022 on essentially the same core design and lithography. Thats like GCN levels of respinning. Intel still has the IPC and clock advantage at the moment, but Rocket Lake is gonna be competing with Zen 4 on potentially TSMCs 5nm by that point. Lord. Maybe Intel will finally have some motivation to compete on price.
|
# ? Dec 28, 2019 01:18 |
|
Cygni posted:With the amount of leaks of the new 1200 socket, including complete designs and part numbers, I think its coming sooner rather than later. On desktop, after Comet Lake is Rocket Lake, another 14nm part with DDR4 in 2021. Seems pretty likely to me that Intel will keep the 2 year socket cadence going with Comet/Rocket on Socket 1200 (H5), before moving to a new socket in 2022 for 7nm Meteor Lake/DDR5... if you believe the rumors and that Intel can actually ship 7nm on schedule. They lost the IPC advantage with Zen 2 being broadly on par except for avx-512.
|
# ? Dec 28, 2019 03:19 |
|
Malcolm XML posted:They lost the IPC advantage with Zen 2 being broadly on par except for avx-512. In my experience with games, that's not really the case. Still something around 10% in favor of intel with the same ram and videocard in the 10 or so things I tested with an 8700 vs a 3600 box i built for a family member with both stuck to 3ghz. Take that for whatever its worth. I also don't personally care about any other workloads but gaming, and Zen definitely excels more in productivity stuff. e: also that was with 3200mhz ram. I'm probably going to build one of the 10 core Comet Lake machines when they come out and I'm building another 3600 box for a different family member, and gonna have at least some 3600mhz stuff around this time. Will be interesting to see if the comparison is more AMD favorable with that stuff. Cygni fucked around with this message at 07:10 on Dec 28, 2019 |
# ? Dec 28, 2019 07:03 |
|
Intel has developed new vapor chambers utilizing graphene and new ultra flat devices are expected at CES: https://hexus.net/tech/news/cooling/138194-intel-intro-graphene-vapour-chamber-laptop-thermal-solution/ There’s also talk about Comet Lake boosting up to 5.3 GHz using Turbo 3.0 and TVB, so on a limited amount of cores at a controlled temperature. The slides also show Socket 1200 and DDR4-2933. https://informaticacero.com/exclusivo-asi-formara-la-decima-generacion-core-de-intel-i9-10900k-buque-insignia/ 5.3 GHz all-core OC on the new 10C flagship doesn’t sound too bad if you can somehow manage Threadripper-esque power output on a much smaller surface area.
|
# ? Dec 28, 2019 09:14 |
|
eames posted:Intel has developed new vapor chambers utilizing graphene and new ultra flat devices are expected at CES: Prescott 2.0 They gonna try to push for btx again??
|
# ? Dec 28, 2019 19:21 |
|
Malcolm XML posted:Prescott 2.0 I can't wait for 3500rpm turbo fans like on my old 3.8 prescott.
|
# ? Dec 28, 2019 20:30 |
|
Tom's Hardware thinks Intel might release Comet Lake-S for both sockets, with the 125W -K "Enthusiast" SKUs only on LGA1200.
|
# ? Dec 30, 2019 11:58 |
|
nerdrum posted:I can't wait for 3500rpm turbo fans like on my old 3.8 prescott. Delta fans or gtfo
|
# ? Dec 30, 2019 15:03 |
|
That modular "NUC" with room for a PCIe GPU seems to be happening after all https://koolshare.cn/thread-168913-1-1.html
|
# ? Jan 6, 2020 09:10 |
|
eames posted:That modular "NUC" with room for a PCIe GPU seems to be happening after all Next Unit of Computing indeed...wonder what (if) they will rebrand it too.
|
# ? Jan 6, 2020 09:19 |
|
movax posted:Next Unit of Computing indeed...wonder what (if) they will rebrand it too. Former Unit of Computing.
|
# ? Jan 6, 2020 12:29 |
|
movax posted:Next Unit of Computing indeed...wonder what (if) they will rebrand it too. PErsonal NUc System
|
# ? Jan 6, 2020 18:08 |
|
eames posted:That modular "NUC" with room for a PCIe GPU seems to be happening after all Questions, 1. How much is that laptop processor going to hold back gaming? 2. Can I really jam a RTX 2060 in here? Still looking for a 2-Slot height card but power wise it looks like it may be done.
|
# ? Jan 6, 2020 18:27 |
|
It's obviously the Next Unit of Gaming.
|
# ? Jan 6, 2020 18:31 |
|
Current Unit of Computing
|
# ? Jan 6, 2020 18:57 |
|
Wild EEPROM posted:Current Unit of Computing Only interested in the unlocked variants
|
# ? Jan 6, 2020 18:59 |
|
movax posted:Only interested in the unlocked variants Either way people would call them cucks so you'd be getting a kay cuck or cuck kay.
|
# ? Jan 6, 2020 19:02 |
|
Computing Unit of Mobility
|
# ? Jan 6, 2020 19:10 |
|
Tab8715 posted:Questions, Looks like it, but it's hitting 93c on the CPU and throttling, at 49db. That's kinda not ideal?
|
# ? Jan 6, 2020 19:37 |
|
What determines if an intel platform supports the lower 1.35V for DDR3L memory? Specifically a Dell 9010 optiplex that uses intel 3rd gen core, specifically a 3770.
|
# ? Jan 6, 2020 20:13 |
|
The processor is going to be what dictates being able to physically support ddrl since that's where the memory controllers are. But the motherboard/firmware from the OEM can override the compatibility matrix and send any voltage they see fit to the dimm, so if Dell is locking you out of that then tough luck. With that said, a 1.35v dimm will probably run at 1.5v well enough.
|
# ? Jan 6, 2020 20:42 |
|
Yeah I think it will run fine. Just curious if it will take advantage of the lower voltage or not. Going to be pushing this system to the limit of its tiny PSU...maybe.
Shaocaholica fucked around with this message at 20:57 on Jan 6, 2020 |
# ? Jan 6, 2020 20:52 |
|
Those run ~250W PSUs right? 1.5 vs 1.35v on the dimms isn't going to be an appreciable difference on a desktop system, maybe a watt or two. It's really more for battery savings where you're trying to shave every last watt. The GPU and CPU are going to be your big draws, if you're seeing voltage dips with the config you're so far past the capacity of your PSU that ddr3l isn't going to help anything.
|
# ? Jan 6, 2020 21:14 |
|
|
# ? May 28, 2024 08:53 |
|
Good point, thanks!
|
# ? Jan 6, 2020 21:23 |