|
BIG HEADLINE posted:Z390 DARK is finally available: https://www.evga.com/products/product.aspx?pn=131-CS-E399-KR Yeah, I don't regret my Aorus Xtreme at ALL after reading through that. Glad I didn't wait.
|
#
?
Jan 10, 2019 18:56
|
|
|
|
| # ? May 23, 2024 14:29 |
|
|
craig588 posted:Doubling the core count can do nothing if your apps are not able to use 8 cores. If they can though, doubling the core count will double the performance. Not really, software doesn't scale like that. You'll get a benefit in (properly) multithreaded applications, but double the core count will never actually double the performance. Also, while things have been slow since Haswell, a new 14nm chip is still going to give you a very significant performance boost over Sandy Bridge at the same core count just from the DDR4, better power handling and OC headroom, infinitely better motherboard OC tools and a decent IPC advantage. All that said, a well overclocked 3570K is still enough power for most modern applications and games, but that's not going to be true in a year or two, and an upgrade would be noticeable even at the same core count.
|
|
#
?
Jan 10, 2019 19:10
|
|
|
Dr. Fishopolis posted:Not really, software doesn't scale like that. You'll get a benefit in (properly) multithreaded applications, but double the core count will never actually double the performance. Also, while things have been slow since Haswell, a new 14nm chip is still going to give you a very significant performance boost over Sandy Bridge at the same core count just from the DDR4, better power handling and OC headroom, infinitely better motherboard OC tools and a decent IPC advantage. So I'd notice have a noticeable speed boost with today's CPUs, or the Ice Lake and its successors?
|
|
#
?
Jan 10, 2019 19:20
|
|
|
WattsvilleBlues posted:So I'd notice have a noticeable speed boost with today's CPUs, or the Ice Lake and its successors? IPC has increased even if clocks look like they have stalled, so yes you will have a performance boost. e: https://www.hardocp.com/article/2017/01/13/kaby_lake_7700k_vs_sandy_bridge_2600k_ipc_review
|
|
#
?
Jan 10, 2019 19:30
|
|
|
Risky Bisquick posted:IPC has increased even if clocks look like they have stalled, so yes you will have a performance boost. Though in many cases you won't actually notice the difference in day-to-day use for office and desktop applications. The difference will be much more apparent if you're doing content creation / encoding and gaming.
|
|
#
?
Jan 10, 2019 19:33
|
|
|
WattsvilleBlues posted:So I'd notice have a noticeable speed boost with today's CPUs, or the Ice Lake and its successors? Yes, but the answer to your original question comes in two parts. It's also reasonably likely that at least one of Zen 2 and Ice Lake will bring a bigger performance bump than we've seen in years, so if you don't need the additional performance right now, it's also not a bad time to wait.
|
|
#
?
Jan 10, 2019 19:45
|
|
|
K8.0 posted:Yes, but the answer to your original question comes in two parts. It's also reasonably likely that at least one of Zen 2 and Ice Lake will bring a bigger performance bump than we've seen in years, so if you don't need the additional performance right now, it's also not a bad time to wait. Yeah, that's the necessary caveat I forgot to put in my post. We're months away from 7nm and there's still a massive shortage of 14nm parts, so if you can wait, then wait.
|
|
#
?
Jan 10, 2019 19:47
|
|
|
Palladium posted:Buy another used DDR3 stick for 16GB and use the old system till it breaks. This. I'm still on a Sandy Bridge i5 2500k@4.4Ghz. Bought a GTX 980Ti in 2015 and a SSD. It runs all games great on at least high settings on my monitor 1080p60hz.
|
|
#
?
Jan 10, 2019 19:58
|
|
|
Dr. Fishopolis posted:Not really, software doesn't scale like that. You'll get a benefit in (properly) multithreaded applications, but double the core count will never actually double the performance. While that is true, there are absolutely real workloads where the scaling is only short of double to a negligible degree. x264 scales very well with core count (up to a point) and I've seen image processing code that scaled 23.8x with 24 threads.
|
|
#
?
Jan 10, 2019 20:24
|
|
|
Yeah what they call "embarrassingly parallel" workloads where there isn't much communication between threads can have nearly linear scaling, most of that stuff is now GPU accelerated but not always.
MaxxBot fucked around with this message at 20:59 on Jan 10, 2019 |
|
#
?
Jan 10, 2019 20:56
|
|
|
So is Plex video streaming one of the things that would benefit from a newer CPU or more RAM?
|
|
#
?
Jan 10, 2019 21:23
|
|
|
WattsvilleBlues posted:So is Plex video streaming one of the things that would benefit from a newer CPU or more RAM? If you're transcoding multiple streams over the internet or to low-powered devices, yes. For in-home use, it barely matters. My plex server is some old 4 core bulldozer thing with 8gb ram and it's vast overkill for the 8 people who use it, since it's always sending to either a Roku, an nvidia Shield or someone's computer or phone, all of which gets locally decoded. edit: for reference, it used to be a 2011 mac mini with a 2-core mobile processor and I never ran out of headroom with that either. Dr. Fishopolis fucked around with this message at 21:31 on Jan 10, 2019 |
|
#
?
Jan 10, 2019 21:29
|
|
|
The weird yield increasing F skus are starting to show up. 9400F on amazon for $208. https://www.amazon.com/gp/product/B07MRCGQQ4 https://www.anandtech.com/show/13836/intels-core-i59400f-hits-amazon
|
|
#
?
Jan 10, 2019 22:38
|
|
|
28 cores at 5GHz is real. Availability is a different matter.
|
|
#
?
Jan 16, 2019 08:44
|
|
|
ConanTheLibrarian posted:28 cores at 5GHz is real. It's 14c, so a binned i9-9940x.
|
|
#
?
Jan 16, 2019 08:54
|
|
|
ConanTheLibrarian posted:28 cores at 5GHz is real. 28? It says right there in the article that it's a 14 core
|
|
#
?
Jan 16, 2019 09:36
|
|
|
HalloKitty posted:28? It says right there in the article that it's a 14 core I assume he just copied the wrong link. Intel is launching its 28c 5ghz-ish (single core boost was leaked at like 4.3 i think) cpu soon, hence the comical motherboards already being up on their sites and all over CES: https://www.asus.com/us/Motherboards/ROG-Dominus-Extreme/ The CPU is the SNAPPILY named Xeon W-3175X, which is apparently just a binned Xeon Platinum 8180. Cygni fucked around with this message at 10:18 on Jan 16, 2019 |
|
#
?
Jan 16, 2019 10:10
|
|
|
drat, no warranty either. Guess that auction will be rather slow.
|
|
#
?
Jan 16, 2019 15:11
|
|
|
Coffee Lake with the -F suffix (no iGPU) are inexplicably priced the same as the regular chips, or as one commenter put it: https://www.anandtech.com/show/13865/intels-graphics-free-chips-are-also-savings-free-same-price-fewer-features anandtech commenter posted:Intel confirmed that their graphics is worthless
|
|
#
?
Jan 16, 2019 16:32
|
|
|
eames posted:Coffee Lake with the -F suffix (no iGPU) are inexplicably priced the same as the regular chips, or as one commenter put it: Uhh... I guess maybe they're for OEMs? An OEM that builds a machine with a graphics card bundled might be able to negotiate a lower price. Maybe. Otherwise I have NO idea what the hell the point was.
|
|
#
?
Jan 16, 2019 16:42
|
|
|
They don't want to segment out the graphics chips and lost both revenue and graphics market share. There are probably some use cases where you need a chip without a GPU (compatibility, power, maybe OCing?) but Intel doesn't want people skipping out of it to save a few bucks. Pricing of CPUs is rarely about materials or features, it's about segmentation.
|
|
#
?
Jan 16, 2019 16:45
|
|
|
If you need a chip without a GPU you can disable it via a BIOS setting, most motherboards disable it by default when the PCIe port has a GPU installed. They're probably able to increase their yields by selling chips with severely damaged and disabled iGPU parts (which are probably quite a few considering the iGPU seems to be ~25% of the die size) but then they should've either dropped the price or at least increased turbo frequencies by 100 MHz. The iGPU isn't very useful but presumably -F consumers will also lose HVEC decoding, QuickSync and the performance advantage in certain video encoders (i.e. Adobe) in exchange for... nothing at all.
|
|
#
?
Jan 16, 2019 16:57
|
|
|
Lockback posted:They don't want to segment out the graphics chips and lost both revenue and graphics market share. There are probably some use cases where you need a chip without a GPU (compatibility, power, maybe OCing?) but Intel doesn't want people skipping out of it to save a few bucks. The number of self-professed technically inclined people that got mad that Intel made available a way to pay to unlock certain features of their low end chip after purchase still blows my mind.
|
|
#
?
Jan 16, 2019 17:38
|
|
|
HalloKitty posted:Uhh... I guess maybe they're for OEMs? An OEM that builds a machine with a graphics card bundled might be able to negotiate a lower price. Maybe. Otherwise I have NO idea what the hell the point was. Maybe this, or possibly they might get slightly better power utilization, since even when disabled via BIOS, the iGPU will use some (small) amount of power. But, yeah, not providing any price discount is a strange marketing decision.
|
|
#
?
Jan 16, 2019 17:44
|
|
|
It doesn't have to be a particularly competitive product when Intel can't supply enough product. It's competing with not buying a CPU because there aren't any to buy.
|
|
#
?
Jan 16, 2019 18:18
|
|
|
WhyteRyce posted:The number of self-professed technically inclined people that got mad that Intel made available a way to pay to unlock certain features of their low end chip after purchase still blows my mind. It was an incredibly stupid marketing move. Normally, capitalism tries to shield consumers from the ugly realities of business - don't rub our faces in it.
|
|
#
?
Jan 16, 2019 18:49
|
|
|
.
sincx fucked around with this message at 05:55 on Mar 23, 2021 |
|
#
?
Jan 16, 2019 19:30
|
|
|
Lambert posted:It was an incredibly stupid marketing move. Normally, capitalism tries to shield consumers from the ugly realities of business - don't rub our faces in it. Trying to to find a new way to make revenue that was previously frozen out is not in itself a dumb thing. And the only people that raised a stink about it were the ones who weren't buying that stuff in the first place
|
|
#
?
Jan 16, 2019 20:34
|
|
|
I'm interested in seeing some independent KF vs. K tests to see how much of an objective difference we're really seeing here without that extra power leakage.
|
|
#
?
Jan 17, 2019 01:22
|
|
|
WhyteRyce posted:Trying to to find a new way to make revenue that was previously frozen out is not in itself a dumb thing. And the only people that raised a stink about it were the ones who weren't buying that stuff in the first place On a $100k solution that's 40% hardware and 60% software and support contracts, being able to toss some cash at the problem to unlock cores 5 and 6 on the 8 core part you were sold when the software uses per core licensing is pretty magical. Segmenting features the board should support by default just reeks of 'gently caress you, pay me'.
|
|
#
?
Jan 17, 2019 01:39
|
|
|
Methylethylaldehyde posted:Segmenting features the board should support by default just reeks of 'gently caress you, pay me'. Forced segmentation is unjokingly the only reason the industry turns a profit, and everyone inherently knows it. But psychologically, there is a point when this fact becomes too obvious and it turns customers off (especially western customers)... even though its true regardless.
|
|
#
?
Jan 17, 2019 01:57
|
|
|
Methylethylaldehyde posted:On a $100k solution that's 40% hardware and 60% software and support contracts, being able to toss some cash at the problem to unlock cores 5 and 6 on the 8 core part you were sold when the software uses per core licensing is pretty magical. Segmenting features the board should support by default just reeks of 'gently caress you, pay me'. In this case it was hyper threading on a low end CPU on some Gateway system. And there is a good chance that this particular SKU only existed because the vender asked for it to hit some price point and Intel thought of a way to get some additional money back down the road But hey, please demand low end processors get the same full features that high end ones have just because it's technically already there and that R&D and design and validation of it costs nothing.
|
|
#
?
Jan 17, 2019 02:25
|
|
|
WhyteRyce posted:But hey, please demand low end processors get the same full features that high end ones have just because it's technically already there and that R&D and design and validation of it costs nothing. On the other hand Intel's net profit last quarter was 6.3 billion dollars.
|
|
#
?
Jan 17, 2019 02:44
|
|
|
Inept posted:On the other hand Intel's net profit last quarter was 6.3 billion dollars. And?????
|
|
#
?
Jan 17, 2019 02:46
|
|
|
They're obviously not hurting for money. I'm not sure why you'd defend a company making their products artificially crappier other than ~capitalism~
|
|
#
?
Jan 17, 2019 03:03
|
|
|
eames posted:Coffee Lake with the -F suffix (no iGPU) are inexplicably priced the same as the regular chips, or as one commenter put it: It is handy the one time when your GPU fails unexpectedly. That's got to be worth a few dollars at least.
|
|
#
?
Jan 17, 2019 03:06
|
|
|
Inept posted:They're obviously not hurting for money. I'm not sure why you'd defend a company making their products artificially crappier other than ~capitalism~ You really don't understand how different people are willing to pay different amounts for different things, and how this forces prices to cater to those demands?
|
|
#
?
Jan 17, 2019 03:14
|
|
|
I get why it happens. I know that market segmentation gets companies like Intel more money. But artificially crippling products to make more money is dumb, and defending it is dumb.
|
|
#
?
Jan 17, 2019 03:29
|
|
|
Inept posted:They're obviously not hurting for money. I'm not sure why you'd defend a company making their products artificially crappier other than ~capitalism~ What you define as artificially crappier is literally segmentation and it's something that pretty much everyone in the industry does, whether they are making a ton of money or under water. And if software unlock feature didn't exist, then it's pretty likely the same SKU would have existed anyway because their customer (Gateway is the customer here) asked for X number of processors at Y price point so they could sell the whole thing for Z. It gets Intel money because it lets them sell something at a variety of different price points that they wouldn't be able to hit if they just tried selling i7s or i9s to everyone
|
|
#
?
Jan 17, 2019 03:31
|
|
|
|
| # ? May 23, 2024 14:29 |
|
|
So if Intel did no market segmentation, and everything was (and thus was priced as) a Xeon, you'd be happier, because a desktop CPU costs $2k but at least there's no market segmentation?
|
|
#
?
Jan 17, 2019 03:32
|
|