Were waitfags btfo'd once again by zen2 prices?

>patching your cpu
fuck off tinfoil

Waiting implies you needed a new cpu or mobo but put it off despite the NEED. Nobody does that. Zen2 is for people with cpus from 2013-2017

I don't know op, but the fanbois better hurry up and upgrade because I'm in need of a cheap 2600x

If amd's entire business strategy is "more cores" they can't release a mainstream 16 core too quickly or the strategy will run out of steam in like 4 years when we'd get 32 core cpus.

>believing adoredtv asspull chart for even a minute

a retard is born every minute

I paid $275 for my 6c/12t i7 8700 that boosts to 4.3ghz on all cores that has a TDP of 65w, what's the deal with all this hype about Zen2 in comparison?

I don't get it

My real preponderance is should I go for the 8c or the 12c? Rendering is GPU, gaming is one or the other and more rarely CPU, but threads go underutilized. I don't know that I do a whole lot of CPU intensive shit, honestly. As it stands I don't see a whole lot of utility for a 12c segment in the consumer market, or am I wrong?

8700k is typically priced around $350 and more.
This kind of shit doesn't effect you at all with the kind of deal you're getting.

>As it stands I don't see a whole lot of utility for a 12c segment in the consumer market, or am I wrong?
do /p/ stuff. I have a 6c12t and not only does a ton of cores make actually editing large raw photos much more responsive, after you're done editing the process of actually applying the edits and exporting to jpeg or whatever will easily peg as many threads as you can throw at it. Shit eats a bunch of memory, too. This is even more true if you do stuff like panorama stitching, merging to HDR, stacking images for astrophotography, etc. All this stuff is embarrassingly parallel and will make good use of 12 (or 16+) cores. Video people have it a even worse.

vidya gaymen, who knows. I know the next gen consoles are supposed to have eight cores, so we'll probably see games using at least that many. No idea what if any effect "four more cores, but now with inter-chiplet latency" has on that.

I'm talking about the 8700, the non-K, it's "just" 300 right now on Amazon and Newegg

Sure, you can't overclock it, but Ryzen overclocks terrible anyway as AMD usually gets all it can out of their chips and doesn't leave much headroom for manual OC'ing

8700 at 4.3GHz is around the same performance as Zen2 at 4GHz, and is nowhere near 65W with MCE. The 3700X will be faster single- and multithreaded, use less power and be cheaper overall when paired with a B450 board. Otoh, you had your CPU for some time now, while zen2 hasn't hit shelves yet.

No not really considering I bought a 2500k in a bundle deal from microcenter in 2011. That probably put the chip around $200 or slightly below. 3700X is a massive upgrade for not much more in terms of real cost. Will probably go with a 3800x just for the binning and see if it manages to live as long.

Yeah, also add "premium" expensive x570 boards that probably will be faulty and die.

Is 3600 gonna be worth the extra $50 over 2600?

About to upgrade my old pc.

Attached: frodo.jpg (400x399, 93K)

Gonna wait for a month and a half after release and pick up a 3700X for under 300$

>3700X for under 300$
Nice joke

Yes. Buying old CPUs is dumb. Get the newest gear and it will last you longer.

I
I>
I
I3

>An incremental upgrade

>amd gives you 15% ipc between 2 generations
>intel gives 15% between 2 decades

same thing i guess

Where did this Ryzen 5 data come from? I don't recall seeing it in the presentation.

Attached: 1129294958856.jpg (433x380, 37K)

AMD said that the 3800X is THE gaming CPU, so it should perform better than the 3900X on that. If you do a lot of simulations, virtual machines and compilations, 12-cores should come handy.

>that filename
did you save this image in 2005?

or from a newer board ya doofas

>TDP of 65w
>he actually believes that

well lets be honest here.
>2011 - 2017
faildozer. zero improvment
>zen 1 released in 2017
behind intel
>zen 1.5 refresh released in 2018
oh boy. a whole 3% ipc increase. still behind intel
>2019 zen 2 - 15% ipc over zen 1
finally on an equal playing field with intel. amd had a lot of room to grow. as since 2006 all the way till now, 13 years, amd was always behind intel. nice to know after over a decade they finally caught up. and maybe have a 3-4% ipc lead (oh boy, such an amazing advantage after 13 years!)

nah, I probably randomly typed those number, i don't remember the reason tho

>>zen 1 released in 2017
>behind intel
What are you smoking dude? Before zen 1 the highest end i7 was the 7700K, a quad-core. The 1700X destroyed it in every multithreaded workloads, and the single-core/gaming performance is not that far behind.

(((Coincidentally))) the next gen i7 was a six-core.

Maybe if you are that one person who gives no shit about security and is smart enough to not get any security updates it is equal playing field.

Lets be honest here, intel can not hyperthread so 8core intel = 8 threads in places where you need to be completely safe, or zombieload will load in your anus.

Even without turning hyper off, you get 33% slower storage speed and 10%+ slower game/heavy app fps

>zen1 behind intel
yeah you are probably talking about the pre patch era where the 7700k was better than a 1800x
i guess you wont talk about the post patches era eh?

see pic related. that was 99% of all games. not just a little bit behind.
>The 1700X destroyed it in every multithreaded workloads
yeah, in stuff that used more than 8 threads. so for 90% of what 60% of what people used their computer for, that multi-core performance didn't get heavy use.
>Maybe if you are that one person who gives no shit about security and is smart enough to not get any security updates it is equal playing field.
amd gives as much shit about security as intel does. they just got lucky and the last 10 years have bounced around from different architectures. starting with a fresh slate. so an exploit from 8 years ago doesn't affect today. while intel has based all their arch's off of the very first core 9 series which is based off the core series which is based off the p4 mobile / centrino which is based off the p3 and you get the point. why amd hasn't actually taken advantage of the exploits on intel processors by ramming it in their face and promoting their cpu's over intel because of those exploits is because amd knows sooner or later they will get hit by nasty ones too.
>Lets be honest here, intel can not hyperthread so 8core intel = 8 threads in places where you need to be completely safe, or zombieload will load in your anus.
please don't spread these lies. 9th gen was already fixed for zomebieload in the hardware. 9th gen doesn't need to disable smt because of it. intel.com/content/www/us/en/architecture-and-technology/engineering-new-protections-into-hardware.html
>family 158, stepping 12, MFBDS (which is zombieload) "hardware fixed"
So zombieload doesn't affect 9th generation. only fallout and lesser impactful ones effect 9th. which will be fixed in stepping 13. here is a pic of my 9900k tested with the MDS tool to check to see if you are vulnerable: i.imgur.com/o9Y636W.png As you can see, not affected.

Attached: bge9MX6 (1).png (1328x2176, 90K)

If you want to compare multithreaded workloads, shouldn't you include Intel's HEDT chips? The 7800X is a good match, with roughly the same price and release date.

as you can see, 9900k current released stepping, P0, is stepping 12. zdnet.com/article/how-to-test-mds-zombieload-patch-status-on-windows-systems/

Attached: WcA80hX.png (1466x981, 60K)

>with roughly the same price
user...

1700x launch price $399
7800x launch price $389

>so for 90% of what 60% of what people...
nice try pulling those numbers out of your ass, faggot

>pic related
cherrypicking 101 as always
>perfomance in general doesnt matter it only matter when we win

>amd just got lucky it has nothing to do with the fact that they adopted the ibm SMT version that doesnt expose the handshake between the threads on cache pool its all about luck
>9th gen is fixed for zombieload
>meanwhile linux is forcing HT off
>windows is forcing HT off
>literally everyone is forcing HT off but intel says otherwise
>surely we will trust intel
>here let me show you a picture of my trusty 9900k that ISNT vulnerable to zombieload
>posts the screenshot
>doesnt realise that zombie load affect store/load/buffers
>thinks that only the line filled buffers is the problem of zombieload

you are correct user you got the cpu you deserve

not pulling numbers out of your ass when you actually start paying attention to what people do. and if you did you'll notice the vast majority don't sit there running 50 virtual machines while compiling chromium for the 50th time as they sit there masturbating to cartoon drawings of their anime waifu.
>twitch streamers!
most gaymers are not professional streamer whores
>muh gentoo
and you call me a faggot

The 1700 non X destroys the 7800X in price/performance.
> can be used on a reasonably priced mobo
> 2 extra cores
> much cheaper
> comes with a fan

>doesnt realise that zombie load affect store/load/buffers

well you actually expect from intel users to READ what the actual problem is?
he posted a fucking screenshot showcasing that he is ok on the buffer while the handshakes need to be STORED first and foremost
Jow Forums is filled with buyers remorse idiots that need to keep arguing about their cpu despite the rest of the world saying that its not true

Attached: rxiqvtrrbjz21.jpg (512x768, 86K)

well were to start from all these lies
>linux is forcing HT off
this is a lie. some distros may go full paranoid and act like openbsd and disable HT but the linux kernel is not disabling HT. you, have that option to be a freakazoid and do it yourself of course. but what linux has done is implement patches like microsoft did to mitigate it. which phoronix has tested. including disabling smt as well for freakazoids like yourself.
>windows is forcing HT off
windows is not disabling HT. microsoft is leaving that up to those who enjoy fear mongering to do themselves. microsoft has only issued mitigation's for the exploits in the recent windows updates. which, again, those mitigation's are not disabling SMT.
>literally everyone is forcing HT off but intel says otherwise
no. not even apple has. apple only added an option for those, again, who enjoy fear mongering can go ahead and disable HT themselves. by default, its not, and apple has issued mitigation's.
>doesnt realise that zombie load affect store/load/buffers
zombieload is MFBDS. do not confuse it with MDS exploits as a whole that you are. its one of a few that make up the MDS exploits. which in that screenshot you would have seen, such as fallout, which is MSBDS and the other lesser impactful ones. and mind you, again, zomebieload (MFBDS) is the only one they recommend disabling SMT for if you don't have software mitigations or hardware fixed for.
th gen is fixed for zombieload
if a tree is directly in front of you, doesn't matter if you deny its existence, its still there. likewise, regardless that you want to believe in a self imposed delusional, 9th was already fixed for zombieload. only the other, less impactful MDS exploits is what current release stepping of 9th is affected by.

Attached: Blender.png (1327x1222, 60K)

amazes me how /g complain about "idiocracy" of other users when they don't even actually read everything themselves. in my follow up response before any other tard actually replied back to me i talked about the difference between zombieload and fallout.

Zombieload is MFBDS which allows a thread on the core to grab speculatively grab data out of the shared fill buffer which may contain data from the other thread.

Fallout is MSBDS which operates on the statically partitioned store buffer. Since the buffer is statically partitioned, use of HT doesn't have an effect on the side-channel attack.

The other two I haven't seen any researchers name, but the last 3 are prevented by the OS flushing the buffers on context change.

This is like the special olympics. I hope none of you nig/gers actually buy stuff at launch prices. 1700X is $130 at Microcenter without counting bundles or savings and will satisfy 90% of users with the unhappy 10% being autists.

Attached: 30308882513260.jpg (1000x666, 39K)

Is the 3600x a waste once again? I dont get why anyone would buy it over a 3600.

Maybe if you dont overclock.. but unless AMD change anything you can usually OC them to the same level.

>this is a lie. some distros may go full paranoid and act like openbsd and disable HT
>some distros

this is fucking down to the kernel you idiot
phoronix.com/scan.php?page=article&item=mds-zombieload-mit&num=1
>windows is forcing HT off

wait and see the next update all knowning god

>apple didnut nutting
>High Sierra, and Sierra in the latest security updates and may reduce performance by up to 40 percent2, with the most impact on intensive computing tasks that are highly multithreaded

i guess losing 40% comes from a magical place that doesnt involve disabling at hardware level HT right?
>zombieload is MFBDS
zombieload affects all three of the data storage places
store buffers and load you can argue all you want about it but you are still gonna be wrong

get your facts straight a bit first of all smt is very different than ht

overclockers wanting the better bin. of course, if, zen 2 actually overclocks well enough to actually need the better bin of course. i regretted buying the 1800x myself. if i knew the better bin would be worthless in overclocking at the time i would have picked up a 1700 instead. wish amd was more open and honest during its lead up to launch about it.

>this is fucking down to the kernel you idiot
if you read phoronix article would you know he, himself, disabled HT. it wasn'e disabled by default. and he only disabled it because people, like you, are saying it needs to be disabled. so he did a three part test, a test where he disabled the default mitigation's in the kernel to test with no mitigation's, a test with the default mitigation's enabled, and then a third where he tested with the mitigations enabled and HT disabled.
>wait and see the next update all knowning god
biased opinion based off of fear mongering and no actual facts about windows disabling HT on intel processors.
>High Sierra, and Sierra in the latest security updates and may reduce performance by up to 40 percent2, with the most impact on intensive computing tasks that are highly multithreaded
zdnet.com/article/patch-status-for-the-new-mds-attacks-against-intel-cpus/
Mitigations for MDS attacks have been deployed with macOS Mojave 10.14.5, released today.
>"This update prevents exploitation of these vulnerabilities via JavaScript or as a result of navigating to a malicious website in Safari," Apple said.
>The fix has no "measurable performance impact," the company added.
also:
9to5mac.com/2019/05/15/full-mitigation-intel-cpu-mac/
>In its tests, Apple recorded up to a 40 percent drop in performance with full mitigation activated. This is because enabling MDS protection involves disabling hyper-threading entirely, and adds additional barriers when the processor switches contexts.
>Most users do not need to worry about enabling full mitigation. macOS 10.14.5 includes the most important and most relevant patches, like preventing JavaScript exploits through Safari. Apple rolled these critical fixes for all customers as the performance penalty was small/negligible.

>zombieload affects all three of the data storage places
no. zombieload is one of the MDS exploits.
zdnet.com/article/how-to-test-mds-zombieload-patch-status-on-windows-systems/

the link i posted. if you went to it you would have read the following:
>CVE-2018-12126 - Microarchitectural Store Buffer Data Sampling (MSBDS) [codenamed Fallout]
>CVE-2018-12127 - Microarchitectural Load Port Data Sampling (MLPDS) [codenamed RIDL]
>CVE-2018-12130 - Microarchitectural Fill Buffer Data Sampling (MFBDS) [codenamed Zombieload, but also RIDL]
>CVE-2018-11091 - Microarchitectural Data Sampling Uncacheable Memory (MDSUM) [codenamed RIDL]

>CVE-2018-12130 - Microarchitectural Fill Buffer Data Sampling (MFBDS) [codenamed Zombieload, but also RIDL]
zombieload affected fill buffer data sampling.again, as i stated, which i will state again, Zombieload is MFBDS which allows a thread on the core to grab speculatively grab data out of the shared fill buffer which may contain data from the other thread. Fallout is MSBDS which operates on the statically partitioned store buffer. Since the buffer is statically partitioned, use of HT doesn't have an effect on the side-channel attack. The other two I haven't seen any researchers name, but the last 3 are prevented by the OS flushing the buffers on context change.

Thats the thing though even if they do have better binned cpus here its not like its going to be significant. I dont think a potentially extra 100-200 MHz is going to make much difference

Have to see what people get when they start receiving them.

nice bait

is this eunha?

>get your facts straight a bit first of all smt is very different than ht
not op but before you tell someone to get their facts straight you should get your facts straight about HT and SMT. HT is SMT. hyper-threading is just intel's name for their implementation of SMT. AMD calls their implementation of SMT, SMT because they didn't feel the need to give it a fancy name as intel did.

obigatoary wikipedia on hyper-threading. en.wikipedia.org/wiki/Hyper-threading
>Hyper-threading (officially called Hyper-Threading Technology or HT Technology, and abbreviated as HTT or HT) is Intel's proprietary simultaneous multithreading (SMT) implementation used to improve parallelization of computations (doing multiple tasks at once) performed on x86 microprocessors.
en.wikipedia.org/wiki/Simultaneous_multithreading
>Modern commercial implementations
The Intel Pentium 4 was the first modern desktop processor to implement simultaneous multithreading, starting from the 3.06 GHz model released in 2002, and since introduced into a number of their processors. Intel calls the functionality Hyper-threading, and provides a basic two-thread SMT engine. Intel claims up to a 30% speed improvement [3] compared against an otherwise identical, non-SMT Pentium 4.
>AMD Zen microarchitecture has 2-way SMT.
hyper-threading is SMT bro.

>running a day1 2600k
3700x is just fine

more to note how HT is not disabled by default on mac's because apple released a guide for those who want to do it themselves, since apple is not disabling it by default as they advise its not necessary unless you really want to for (((REASONS))) after they released their mitigation's.
>With those qualifiers in mind, to enable full mitigation, follow these steps:
>Restart your Mac and hold Command key and the R key to enter macOS Recovery mode.
>Open the Terminal from the Utilities menu.
>Enter the command ‘nvram boot-args=”cwae=2‘ (without single quotes) and press Return.
>Enter the command ‘nvram SMTDisable=%01’ and press Return.
>Then restart the Mac.

yeah i think zen 2 will over clock "better" than zen 1, but i don't think it will be intel level of overclockability still. from seeing the rated default clocks, it seems like amd caught up to skylake in frequency scaling. with how the upper frequency boosts appear to top off around 4.4 - 4.5ghz. 4.5ghz being reserved for the highest binned chip, the 12 core 3900x. maybe it'll top off around kaby if they can get 4.7ghz overclocks on all cores. if that's the case, then for a lot of hardcore overclockers, again hardcore, it would be worth it. but for most, yeah, just get the cheaper alternative. if i do upgrade ill probably get the 3700x myself. but i think i will skip on zen 2 because my 1800x still good enough for me. i'll wait to see what icelake with intel and zen 3 brings next year.

you know what he means though
intel's implementation of SMT is indeed different

>ITT: Intel cope
I don't get it. I have a fucking dogshit 6600k and you don't see me crying.
3700x seems to be the most based choice.

12 core should have been 449 with the 3800X being 379. To compensate for the x570 chipset being expensive hot garbage.

>intel's implementation of SMT is indeed different
everyone's SMT of SMT is different because SMT is just a basic concept that deals with having a core process more than 1 thread. its up to engineers to figure out a way to implement that concept. jesus christ guys come on. stop being retarded.

it's exactly same prices as for zen1, what do you mean?

> 15% over Zen 1

I think it's over Zen+

Still, 12 cores with potentially the same single thread performance of a 8700k is fucking dope

>stop being retarded
or how about you learn to fucking read

amd doesn't want to be seen as a value company anymore. lisa sue said two years ago she wants to change amd's perception as a value oriented company to a more premium oriented company. an easy way to do that is to charge a more premium price. one of the reasons why zen 1 had such retarded pricing with the 8 cores. like the $500 1800x. got utterly ass raped by the stupid $300 7700k quad core in nearly everything. only losing in highly parallel workloads. so it made the 1800x a great budget workstation processor with one downside. it wasn't on a workstation platform. limited to dual channel memory rather than quad and limited pci-express slot count. 20 tops. then when threadripper came out it took over the budget workstation market for amd. resulting in the 1800x becoming vastly more irrevlent for a stupid $500. only the 1700 (non-x) made sense at $300. slower than a 7700k in everything outside highly parallel workloads but for $300 it made a perfect entry level workstation processor. pick up a cheap b350 board and a cheap dual channel ddr4 kit and you were good to go on a low budget workstation.

>The 7800X is a good match, with roughly the same price and release date.
>literally launched two months after 1st gen zen

I'm 90% certain it was a product released to be more competitive with AMD at that price point.

Nah, AMD actually licensed IBM's implementation I think. Whereas Intel were being cheap basters and made their own implementation which turned out to be buggy and insecure.

I guess I'm the only more excited about navi. NO GCN and it might be cheap when released because of how many they will make for new consoles.

>Nah
>...
>I think.

you could start first by reading instead of being retarded.
you do know ibm didn't make SMT right? ibm's SMT is their own in house implementation of SMT.
en.wikipedia.org/wiki/Simultaneous_multithreading
>The IBM POWER5, announced in May 2004, comes as either a dual core dual-chip module (DCM), or quad-core or oct-core multi-chip module (MCM), with each core including a two-thread SMT engine. IBM's implementation is more sophisticated than the previous ones, because it can assign a different priority to the various threads, is more fine-grained, and the SMT engine can be turned on and off dynamically, to better execute those workloads where an SMT processor would not increase performance. This is IBM's second implementation of generally available hardware multithreading. In 2010, IBM released systems based on the POWER7 processor with eight cores with each having four Simultaneous Intelligent Threads. This switches the threading mode between one thread, two threads or four threads depending on the number of process threads being scheduled at the time. This optimizes the use of the core for minimum response time or maximum throughput. IBM POWER8 has 8 intelligent simultaneous threads per core (SMT8).

also zen's SMT is actually amd's own in house implementation of SMT that they designed themselves. they didn't "license" someone else's SMT implementation. AMD does have a better performing implementation of SMT than intel does though.
anandtech.com/show/10591/amd-zen-microarchiture-part-2-extracting-instructionlevel-parallelism/6
for example Intel uses a 'port' design where floating point and integer instructions can only be issued in different clock cycles per port, whereas Ryzen has ten dedicated execution pipelines. so AMD's SMT scales better. there are a lot of different ways to implement SMT. like IBM is going crazy with 8 thread per core SMT.

holy fuck that intel cope on this thread trying to project that its all good and safe

I can't in good consciousness recommend AMD to my friends. too many security flaws and the speed is subpar

agner.org/optimize/blog/read.php?i=838
>AMD can execute six micro-ops per clock while Intel can do only four. But there is a problem with doing so many operations per clock cycle. It is not possible to do two instructions simultaneously if the second instruction depends on the result of the first instruction, of course. The high throughput of the processor puts an increased burden on the programmer and the compiler to avoid long dependency chains. The maximum throughput can only be obtained if there are many independent instructions that can be executed simultaneously.
>This is where simultaneous multithreading comes in. You can run two threads in the same CPU core (this is what Intel calls hyperthreading). Each thread will then get half of the resources. If the CPU core has a higher capacity than a single thread can utilize then it makes sense to run two threads in the same core. The gain in total performance that you get from running two threads per core is much higher in the Ryzen than in Intel processors because of the higher throughput of the AMD core (except for 256-bit vector code).
since ryzen can do 6 micro-ops per clock cycle it can take advantage of SMT better because it has more holes left open that SMT can fill in those gaps to take advantage of the extra throughput. while intel scales less with SMT because intel can only do 4 micro-ops there are less holes to fill with SMT.

It may have been a case of Intel avoiding some patents and AMD paying up.

what the hell are you talking about? SMT is just a concept. how can you not get that into your head? and again, amd developed their OWN implementation of SMT. they didn't "license" their implementation from anyone else. if anything, you could save intel did in a way, kinda "licensed" their SMT implementation from Alpha as they partnerd with Dean Tullsen and used his work to create their SMT they named HT for marketing reasons.
en.wikipedia.org/wiki/Simultaneous_multithreading
>Historical implementations
>While multithreading CPUs have been around since the 1950s, simultaneous multithreading was first researched by IBM in 1968 as part of the ACS-360 project.[2] The first major commercial microprocessor developed with SMT was the Alpha 21464 (EV8). This microprocessor was developed by DEC in coordination with Dean Tullsen of the University of California, San Diego, and Susan Eggers and Henry Levy of the University of Washington. The microprocessor was never released, since the Alpha line of microprocessors was discontinued shortly before HP acquired Compaq which had in turn acquired DEC. Dean Tullsen's work was also used to develop the Hyper-threading (Hyper-threading technology or HTT) versions of the Intel Pentium 4 microprocessors, such as the "Northwood" and "Prescott".
and

Yes, and certain implementation details of that concept can be patented. You can still borrow most of the original work, so long as avoid the patented parts.

If they don't want to be seens as a value company, then they should have made a non-retarded chipset. Now it's very unattractive to buy the Zen 2 CPU's, while these prices would have been acceptable if you didn't have to by a motherboard that will kill itself in a year or two.

Is it just me or does Intel reek of low IQ? Every time I see an Intel build, my first thought is 'you only went with Intel after being fooled by other pozzed fags."

>14+++++
just noticed that, kek

>if you didn't have to by a motherboard that will kill itself in a year or two.
is this the new meme from HR?

Attached: eyes.jpg (320x371, 62K)

>Buying locked CPUs
Oh you poor retards.

>lisa sue said two years ago she wants to change amd's perception as a value oriented company to a more premium oriented company
Do you enjoy just making things up

Zoomers might not remember how dogshit chipset fans are, but I do. Those things die all the time and make even more noise than a AMD blowerfan gpu

Not him but nForce motherboards were like that.
Still as long nobody cheaps out and uses ball bearing fans it will be ok.
But im sure there will always be that retard that will use sleeve bearings or gimmick sleeve bearing fans(Rifle bearing, Fluid bearing, etc).

If someone used standard 40-80mm fans i would have bought that motherboard.