INTEL IS FINISHED

>>59287949
The 1700 is very good tbqh, but Why did they charge so much for the 1800X? It's literally a 95W Overclocked 1700, same for the 1700X.

R5 will be the best though, with lesses cores, they will manage to get much higher clock speeds.

>>59288044
>Why did they charge so much for the 1800X?
because dumb kids want to spend more of their parents' money for such a thing. the 1700 is the informed customer's choice.

File: order.png (22KB, 732x230px) Image search: [Google]

22KB, 732x230px

Was about to drive to MC today to get the $350 1700X but yeah this will do. Thanks OP

>>59288044
>will manage to get much higher clock speeds.
NO they won't.

Now if only I could find a mobo that isn't shit and in-stock

>>59288209
Stock issues should be solved by the end of the week apparently.

>>59288182
4.5GHz stock cooler OC confirmed

>>59288242
Where? At what voltage? I mean you can clock any 7700k to ~5.3 at the cost of it dying in a few weeks.

>>59288241
any source on that?

>>59288304
his ass, its just a rumor, I'm going to guess tehre won't be sufficient stock for at least ~2 weeks

shipping takes time after all

>>59288335
>>59288335

I saw it on leddit, it's not just my ass. BH and Newegg says that they're going to restock tomorrow as well.

File: intel2.png (495KB, 1070x601px)

495KB, 1070x601px

N-N-NO...
DELET
THIS
RIGHT
FUCKING NOW!!!!

>>59287949
reminder to spam amd to support libreboot

now more than ever we should fight against cia niggers

What's the difference between a 1700 and 1700x?

>>59289833
Clocks. You can OC 1700 to ~3.9 at chepo B350 mobos anyway.

>>59288139
Either that or business don't want to overclock all their cpus and can swallow an extra $100 per unit.

>>59288256
That's not how overclocking works retard.

>>59288044
Don't they come with different heatsinks that are actually bretty good?

>>59287949
thnx op
>59289833
X sku have XFR and a higher baseclock. All 3 chips seem about the same with a manual oc.
XFR only uses +1 multi as of now :/ 1700 w/oc seems like the one. it's the 8320e again.

>>59288976
trannieboot is kill

>>59287949
>buying ayymd
>ever

enjoy your housefire

>>59289833
From what I have seen on benchmarks, the difference between the 1700 and the 1700x is merely the XFR (automatic overclocking) a manual clocked 1700 is the same.

The 1800x, however, does seem to have advantages over the 1700 and 1700x

nice ebayad faggot

report and hide

go

>>59291612
Apparently both of them support XFR. 50 mhz and 100 mhz for the 1700 and 1700x respectively.

1700, with a top tier motherboard and watercooling can be pushed to 3.9(maybe 4.0) where as a 1800x can do 4.3.

I am not sure if I want the 1700 or the 1800x. I think that 0.4 overclock potential is not worth the $150 and will end up going with the 1700

>>59291714
The fuck is even the point of XFR? Why not just say it clocks to 3.5/3.9 instead

File: aavderdaagegamiang.png (72KB, 601x830px) Image search: [Google]

72KB, 601x830px

>>59287949
>buying Ryzen ever

Are you stupid?

>>59290335
Nah they all have xfr. Apparently the x moniker is like with the 290 vs the 290x where the x just means it's more better. Well actually even then it made more sense as the non x version is a cut down imperfect Hawaii die. No idea why they made such a meaningless and inconsistent distinction for the zen chips, but marketing and naming has never really been AMD's strong suit.

>>59292088
nope, but you are

>>59292071
Based on early descriptions and the master oc tool, I think XFR is supposed to be more than it is right now.

Whether you believe it's just early BIOS/chip/firmware issues that just need to be sorted out or just AMD talking up something to seem better than it is is up to you. If you choose the former however, the underwhelming performance of XFR might be currently held back because not everything is perfect yet and it will eventually become a very versatile and useful function. The way it was spun is the chip would sense which cores were being taxed most and clock those the highest while reducing allocation of resources to the less used cores on a task by task basis.

Of course it could all just be a marketing gimmick with no tangible benefit. Considering how many various issues have been seen from all motherboard vendors, however, I'm at least willing to keep an eye on progress to see if all these troubles are just the archetypal result of a brand new product with many brand new features that simply cannnot be perfected on day 1.

lol.
>>59292088
>le gayming 6% difference which clock speed accounts for
kys.

>>59292425
Even if it works as you said, it's still only 100Mhz so big whoop?

>>59287949
So, the thing is, this CPU was meant to compete with the Skylake flagship on a budget. From what I've seen, it's not too bad really, but from what I know, it's more or less aimed at testing new technologies with clock speeds, and frankly seems more based to the computational/video editing load.

>>59292205
>marketing and naming not amds strong suit

Dafuck are you on about. Intels naming scheme is the most retarded and unnecessarily complicated pile of shit I've ever seen.

There is no debating this.

>>59292518
marketing certainly isn't
intel has many commericals on tv which convince goys to buy intels processors

>>59292475
>spending twice as much on a Ryzen CPU that performs worse than an Intel CPU

>>59291666
-t.butthurt Intelfag

>>59292088
thanks to its price, you can get 2 and do crossfire so you get twice the performance of an i7-7700k for less than half the price ;)

>>59292088
7700k can't even get 30 fps over a 8370 what a piece of shit

>>59291304
>buying intel
>ever

So you want to burn down your entire neighbourhood?

>>59292518
>Intels naming scheme is the most retarded and unnecessarily complicated pile of shit I've ever seen.

I liked the naming scheme AMD had going in the Athlon XP and Athlon 64 days.

Their CPUs had names like "Athlon 64 2800+" that actually reflected some kind of objective performance test (I think).

File: uEW7PCL.png (27KB, 533x442px) Image search: [Google]

27KB, 533x442px

>>59292088
>Intards

>>59287949
>8 core 16 thread for $300
Great timeline this.

>>59288151
They 'corrected' it.
Basically, their website doesn't list it as $350 anymore, so they're not going to honor their mistake.

>>59296475
Not sure why they'd sell it at $400. I think the word is out that the 1700X isn't worth it at all. People that want the best will still go for the 1800X.

>>59292088
>8 intel processors have the same performance
nice try shekelstein

>>59296558

i have a 6600k that performs slightly worse than ryzen, i'm talking about 5-10 fps worse, but if i overclock it it will get 10-15-20 fps more than ryzen, keep in mind that its just 4c/4t and 6600k sells only for $240

>>59292088
The fact that it performs better with SMT turned off just tells you have fucked everything is on the software side. Once that gets straightened out, Ryzen is going to be a beast.

In a lot of ways it's a typical AMD launch. Ridiculously powerful hardware for the price but completely lacking on the software front. Still needs time desu.

File: 1488893022307.jpg (69KB, 1440x810px) Image search: [Google]

69KB, 1440x810px

>>59292643

>>59296617
*how fucked

>>59296617
The SMT issue is just a windows thang. From my understanding it works completely fine on Linux in that regard.

File: 1488915545091.jpg (273KB, 1032x774px)

273KB, 1032x774px

>>59291304

When does the r5 series come out? I'll wait for that so my dreams can be shattered again

>>59296729
Same time Intel is releasing new cpu's.

>>59296729
Looking like late April to me.

>>59296836
No, that's R3 and Raven Ridge. R5 is coming Q2 and probably not "late Q2".

>>59287949
>buying on ebay

>>59288044

>R5 will be the best though, with lesses cores, they will manage to get much higher clock speeds.

Stop with this hype faggotry.

It will clock the exact same as the R7. It will even perform worse because it will have a smaller cache.

It will perform between an i7 and an i5. That's your best hope.

File: smugamphibian.jpg (90KB, 1024x768px) Image search: [Google]

90KB, 1024x768px

>>59296891
>yfw he opens up an box full of celerons taped together

>>59296674
There's clearly a lot of things at play. It only makes sense all this new stuff won't be perfect right away. It's not like every single tech company (including and especially video game devs) dont just ship unfinished and fix em later.

>But AMD should have made sure it was just right before launch!

Bruh, they did their damnedest. They ain't got shit for spare cash and they have to carefully allocate what they spend it on and how. It's the reason they spent over 6 years on Zen: equal parts limitations of a small company and hoping to do the best they possibly can. Im sure they had some kind of plan for a stop gap launch at some point but ultimately decided to take their time making zen as good as possible. Just look how long ago they started sending out engineering samples, they've been doing their best to refine the chips and had only set vague launch date goals to prevent missing deadlines or rushing.

But they could, and still can, obviously only do so much. They can't just force everyone else to bend over backwards for them, nor can they take on all that work. So only time will tell if these issues can be sorted out, and what it ultimately means. The good news is they're sure to make at least some wins server side and that many companies are clearly interested in the promises of zen.

Ive said it before and ill say it again: autistic man children on /g/ arent the only ones sick of intels shit.

>>59296891
>>59296970
eBay always sides with the buyer. Selling on eBay is what you need to be afraid of.

>>59297038
>Ebay always sides with the buyer.
Kek no, Ebay gets a cut off the sellers share. Without sellers they're literally not making any money.

File: bob hartley.jpg (23KB, 620x399px) Image search: [Google]

23KB, 620x399px

>>59287949
>monthly financing options available
>they always get their interest

>>59297032
Yeah, it's really a pretty awesome chip when you think about it. More power efficient than comparable intel chips at a very good cost. Reasonable IPC based on expectations. It's a winner for sure.

>>59297102
Literally just google it. You haven't been on eBay long if you think otherwise. They take the sellers word every time on a case.

Intel and AMD have been going back and forth for decades now. But I guess if you're an underage memelord bandwagoner, you wouldn't know that.

>>59297145
>sellers

mean't buyer's

>>59292484
The big whoop isn't so much about raw performance (ignoring that essentially being a selling point from the "it'll overclock as high as your cooling allows!" Claim) as efficient allocation of resources. The idea being if you're running a single thread heavy workload it can try to make that one core as fast as possible without wasting a bunch of power and producing a lot of heat (especially important in servers and laptops) overclocking other cores. And especially neat: in a scalable and automated manner across groups of cores in a simple, automated way.

IN CONCEPT, any way. It's a cool idea. It remains to be seen just how well it actually works. Im hoping it just needs time. We already know one mobo manufacturer had an issue with memory or something I think, and in the process of fixing that inadvertently introced a voltage controller issue. Was it MSI? Someone correct me on this. I want to believe some kinks need time to iron out, not just because im an obvious amd fanboy, but also because its such a neat prospect.

>>59288044

>r5 will be the best though
>less cores
>they are just going to be binned r7's with some cores disabled

seriously, why is everyone thinking that r5s aren't going to be binned r7s?

EXAMPLE: https://en.wikipedia.org/wiki/Phenom_II

or.... https://docs.google.com/spreadsheets/d/19Ms49ip5PBB7nYnf5urxsySvH-Sdy6liE2EBDaB8b54/edit#gid=0

"Four AMD K10 cores chip harvested from Thuban with two cores disabled"

"Three AMD K10 cores chip harvested from Deneb, with one core disabled"

>>59297341
They are going to be binned r7s, r5s have more L3 per core.

>>59296902
>It will perform between an i7 and an i5. That's your best hope
At $150-$200 price range

>>59288182
They will.
My cousin's wife's friend's sister's husband works for AMD. He told me in confidence at a family get together.


mums the word

Yes, they will be getting much higher clock speeds.
Samsungs LPP process is made for lower clock speeds.
IBMs, I forgot what it's called, should be good for at least 10% higher.
>>59296902
>It will perform between an i7 and an i5.
How is that not good?
To get more performance than an i5-7600k in games that use SMT, but for as cheap as an i5-7400 and unlocked.

>>59297425
>>59297447

WRONG. Enjoy your binned chips.

>>59297378
how can a binned chip have more l3 cache per core? do explain faggot

>>59297471
Because execution units are disabled while cache size is maintained and shared amongst fewer cores

>>59297535
No, the 4c ones will likely have an entire 4 core cluster disabled, or the iGPU disabled from Raven Ridge chips.

Chips that have 2 cores bad on one cluster will have the entire cluster disabled and will be sold as a 4c, not a 6c.
Ones that have 1 core disabled, or 1 on each cluster, are the only ones that will be 6 cores with 33% more cache per core.

4cores will not have a cache-per-core benefit, I am certain.

>>59297471
>Enjoy your binned chips.
Is this some sort of insult? I don't get it.
The 7100, 7350k, 7400, 7600k, and so on are all binned 7700k chips. The difference is that their binning is unreasonable, with HT arbitrarily disabled for no reason on so many of them, and every single one of them is overpriced.

>no quad channel memory
>reduced PCIe lanes
AMD why

>>59298155
Naples has 8 channel memory

>>59288044
Ah yes, because the FX-6300 clocked so much better than the FX-8320/8350.

>>59298155
The 16 core will have quad channel.

I do think dual channel is worth the cost savings on the 4-8 cores, frankly.

At least they scale up nicely with memory speed.

>>59287949
>309.99
What the fuck? WOW!

>>59298516
Washington area only and you add on like 24% sales tax.

>>59297102
Ebay always sides with the buyer. You only find this out when you get screwed by lying buyers. The seller is usually screwed unless you pay extra to require the receiver to sign that they got the package.

File: whats_in_a_name.png (317KB, 1275x717px) Image search: [Google]

317KB, 1275x717px

>>59292205

>>59298571
Signature doesn't even work for most things. They can just say the thing doesn't work when it does and you're still fucked. They can send it to you fucking broken and you can't do a goddamn thing about it.

File: 1484554951198.jpg (12KB, 320x320px) Image search: [Google]

12KB, 320x320px

>>59298686
Well shit then never mind.

In the earlier days of bitcoin, I sold coins in groups of 10 for like $200 around 2011 or so. First couple of sales went fine. Then I sold to an asshat from California who said that his account and/or credit card was stolen. I kept the message where he told me his wallet number to send to, and it was easy to prove I sent the coin via the open block chain. But Ebay/Paypal was having none of it and said I should not be selling digital goods anyway. Pretty much stopped me after that.

I've also sent physical goods to places and have customers simply say they never got it, demanding refunds and getting it.

File: Selection_20170308_08:10:33.png (16KB, 521x167px) Image search: [Google]

16KB, 521x167px

>>59297471

>>59298155
>>reduced PCIe lanes
24 out of CPU
whatever for sata and crap on board
SLI 8+8 how many left?

quad channel+ is irrelevant outside datancenter when you need to save that 1.5% a day with massive data arrays

>>59298597
Optimized 1820X, 1850X when?

>>59299007
fcuk off tripfag

>raisin
>fryzen
>"Anon, I'm a female CEO and I don't feel like gaming today"

>the lower core chips will clock highest! They have less cores to feed and thus more thermal/voltage headroom!

This makes sense if TDP or high relative resistance induced by multicore function, or alternatively increased current runaway induced by a relative drop in resistance leading to instability, could be marked as the for sure limiting factor. However there is more to consider.

Remember that all zen dies are fabbed with a full fledged 1800x (or better assuming top tier chips are entierly reserved for servers ala the 8970 chips) in mind. Anything and everything below is a "binned" variant. Assuming the limitation stems from the base architecture of Zen itself, cutting off chips likely won't do much. Furthermore, the fact a large sector of the die would have to be found imperfect to produce a lower core count chip might suggest overall silicon quality is lacking respectively. Alternatively it could just be litho related issues but again, it's just a possibility.
Also for consideration is a limitation of the process. If the silicon itself is the limiting factor, it will always be a limiting factor, no matter what AMD does to their design.
Furthermore BIOS issues are plentiful at the moment. Although an unlikely cause, some more fine tuning might be yet required in that area.

Far too much goes into the production of CPUs to make a blanket statement like that. From the base design process to the node used to the refinement process on the part of amd and glofo together, many things could lead to hard limitations. We absolutely will not know what the lower core parts can do until they launch and reviews come out.

anybody concerned about the 1080p gaming deficiency?

>>59299911
no, if it was like bulldozer where you get low frames,stutter, high temps, high power draw
I would avoid 1700 like fire

but you get low power draw,low heat, no stutter, 15% within single IPC depended games and 5-7% multi threaded games

for $60-70(with this price it's more like $100) cheaper package, decent compromise

>>59299911
No.

Many games now days are 99%+ taxing all 8 threads on the 7700k in multiplayer like BF1 at under 144fps.
Whereas the 1700, while it doesn't get as high FPS in the worst case, is only like 60% taxed under the same situation.

It's clearly a problem of a lack of games utilizing the 1700's power, and not that the power isn't there. If they can utilize 99% of a 7700k at under 144fps then obviously it's not very future-proof, now is it? It might be good today for 100fps, or 120fps, but I want the CPU for upcoming games.
It seems clear to me that the 7700k is already on its last legs with just games that came out the past year when it's a "new" CPU. Things are downhill for it. In 2018 I expect high end 4c/4t to barely be good enough for 60fps in most games, and 4c/8t hardly much better.
And for games that don't need a lot of power that don't tax a 7700k like some current games can, the 1700 is still going to be good enough in them.

Plus it's like.. 5-10% worse on average in todays games, but almost twice as good for productivity.

>>59288907
Why did it have to be this way ;_; Intel was supposed to be the best and then AMD just shows up with this amazing chip. I just hope that the R5 doesn't make me regret spending 400$ on a gaming CPU...

>>59287949
Fuck offffff bitch

>>59287949
Honestly though, it's got a solid out-of-the-gate lead on the x700k-series i7's, when it comes to price-new-to-performance. When software comes to use it's full potential (and AMD chips are apparently aren't being 100% used as >>59300323 pulled out of his ass) it will just go uphill from here.

>go on /g/ last week
>AMD IS FINISHED

>go on /g/ today
>INTEL IS FINISHED

what in the fuck is even happening on this board

>>59301880
Shit-slinging on both sides as more and more benches come out and shit gets optimized.