Quantcast
[ 3 / biz / cgl / ck / diy / fa / g / ic / jp / lit / sci / tg / vr / vt ] [ index / top / reports / report a bug ] [ 4plebs / archived.moe / rbt ]

Due to resource constraints, /g/ and /tg/ will no longer be archived or available. Other archivers continue to archive these boards.Become a Patron!

/g/ - Technology


View post   

[ Toggle deleted replies ]
File: 44 KB, 650x350, 2700xgtaV.png [View same] [iqdb] [saucenao] [google] [report]
65611781 No.65611781 [Reply] [Original] [archived.moe] [rbt]

lol everyone fucked up the ryzen review as they did not have the spectre meltdown patches on the intel boards

Time2getwrecked intel
The most trusted tech site around shows the 2700x beating the shit out of the intel CPUs

https://www.anandtech.com/show/12625/amd-second-generation-ryzen-7-2700x-2700-ryzen-5-2600x-2600

>> No.65611789
File: 44 KB, 650x350, RoTR2700x.png [View same] [iqdb] [saucenao] [google] [report]
65611789

>> No.65611801
File: 44 KB, 650x350, RL2700x.png [View same] [iqdb] [saucenao] [google] [report]
65611801

>> No.65611816
File: 44 KB, 650x350, Civ62700x.png [View same] [iqdb] [saucenao] [google] [report]
65611816

inb4 hurrdurr gaymes
Gaming is the only thing the intel chips had over the AMD ones until now

>> No.65611853

Ryzen is the only relevant line of X86 CPUs.

>> No.65611922
File: 58 KB, 916x575, Leddit2700x.png [View same] [iqdb] [saucenao] [google] [report]
65611922

so the guys of anandtech posted on the AMD reddit

Hmmm
My guess is they are the only one to apply the firmware updates to the intel boards
Everyone probably "patched" the intel smeltdown problems but no one hardware updated

>> No.65611974
File: 153 KB, 499x771, totes2.jpg [View same] [iqdb] [saucenao] [google] [report]
65611974

>>65611781
Im not really loyal to any company but I think its sad that people get fired up over a 10 fps difference in a videogame. If a chip works for your task or job, cool.

>> No.65612007

>>65611974
thing is the ryzen 2700x is cheaper than the 8700k and comes with a nice cooler and performs better or WAY better in every task possible
Yet people will still buy the 8700k 96% of the time

>> No.65612053

>>65611974
spotted the intel shill.
they have to go with the it just werks macfag shit.

>> No.65612090

The Ryzen 2700x is the best processor money can buy, hands down. There is no reason to buy anything else.
>muh overclocking
>muh gaming
Listen, stop being a kid and grow up. Adults don't fiddle with esoteric bullshit nor do they play video games. Grow up and get Ryzen, then thank me later.

>> No.65612105
File: 1.49 MB, 2000x1367, 1467327921774.png [View same] [iqdb] [saucenao] [google] [report]
65612105

>>65611781
>analtech
>most trusted
>everyone else fuck up
>ryzen besting the shit out of intel in games

>> No.65612110
File: 127 KB, 600x600, 1496686398502.gif [View same] [iqdb] [saucenao] [google] [report]
65612110

>>65611781
>>65611789
>>65611801
>>65611816
I cant take this no more, I was loyal to Intel and bought a new motherboard every time they refreshed.

>> No.65612111

>>65611781
>>65611789
>>65611801
>>65611816

if AMD fanboys are so into productivity, and megatasking and muh workflow why do they spend so much of their supposedly valuable time shitposting discredited benchmarks on /g/? something doesn't add up.

>> No.65612115

>>65612090
Games stopped being fun when devs decided they should be like playable ads or movies.

>> No.65612119

>>65611781
Lol, yes I am sure someone as autistic as the Gamers Nexus guy forgot to apply the mitigations.

Are there some new Ryzen bugs not well know yet? Or is there something else that is requiring this level of shilling from Team Red? Because this post is just fucking sad. Be happy your 8 core can marginally outpace a 6 core in highly parallized tasks and take your "win".

>> No.65612130

>>65612111
discredited by who?
Also this is the only "win" intel has had for the past year~ is gaming performance
Ryzen has them bested in every production benchmark around.

>> No.65612151

>>65612119
Im sure someone who has been around for 20 years doing CPU reviews fucked this up too..even after retesting

>> No.65612152

>>65612130
Excel and stuff has been fasyer on intel. Beleive it or not, most software doesnt use 12+ threads snd the single core perf is important when you can just throw more threads at it

>> No.65612165

>>65612152
lel, fucking LibreOffice can use OpenCL for this shit. Get with the times, gramps.

>> No.65612176

>>65612165
>muh spread sheet calculations!
Lol, fucking anything to support the coarz race eh?

>> No.65612178

>>65612165
>LibreOffice
No one cares, its not an industry standard

>> No.65612192 [DELETED] 
File: 145 KB, 1500x1054, 78334187965987.jpg [View same] [iqdb] [saucenao] [google] [report]
65612192

>>65611781

MODS

>> No.65612195

>>65612178
>point out that open-source software does better than your proprietary shit
>lol no1currs

>> No.65612218
File: 351 KB, 1292x782, excel2700xhmmm.png [View same] [iqdb] [saucenao] [google] [report]
65612218

>>65612152
its nearly twice as fast as a 8400

>> No.65612221
File: 45 KB, 596x628, keksimus_maximus.jpg [View same] [iqdb] [saucenao] [google] [report]
65612221

>>65612152
>Excel and stuff has been fasyer on intel.

>> No.65612224

>>65612218
intlelfags btfo

>> No.65612226

>>65612195
fugging this

>> No.65612264

>>65612195
Libreoffice is slow as fuck on me even with opencl when opening large spreadsheets with lots of formulas

>> No.65612380

>>65612218

oh look, it's just intel getin rekt again

>> No.65612554

>>65612130
nobody has AMD beating intel in GTA but this one retard who fucked up his numbers. Notice how this is the only graph being posted?

>> No.65612582

>>65612554
1 min into this video
Both intel and AMD chips using overspeced ram here

https://www.youtube.com/watch?v=Mr2B0RJd7Nc

>> No.65612587

>>65612554
That's because nobody applied the Smeltdown patches or didn't retest after applying them, despite what they say. Truth isn't democratic, if a thousand people agree that the sky is purple it doesn't make it so, yet this is the argument intelfags are making.

>> No.65612612

>>65612587
whats even weirder is the Intel results are not massively down in performance they are basically the same as 8700k launch performance on anandtechs site
Ryzen performed very well for their results

>> No.65613075

>>65612582
I love how Ryzen is hardly being stressed.

>> No.65613426

>>65613075
well it has a ton of unused threads
games are becoming more threaded each year so the higher thread count should allow the ryzen chip to be relevant longer
this is why no one buys the 7700k now

>> No.65613449

>>65612151
Ian was in his diapers 20y ago. Anand is retired.

>> No.65613457

Holy fuck.

I didn't expect AMD to surpass Intel until maybe Zen 2.

Now they're both the most powerful and the cheaper option.

>> No.65613544

>>65613075
Ryzen is wider than Con Lake IIRC, that's why SMT scales so well compared to Intlel. That also means there are more unused resources.

>> No.65613557

>>65613457
its hard to say fully passing them

ipc, amd was better then intel with ryzen, but clocks, intel had a fairly large advantage that overtook amds side.
but amd now got even better ipc and clocks better, that gap got closed because the gap was barely present to begin with.

however completely passing intel, that will wait till 7nm when amd can have clock parity at the very least.

>> No.65613614

>>65611781
Techpowerup review is using the latest security patches for both systems

>> No.65613647

>tfw bought into the intel meme

looks like i'm building a new rig soon

>> No.65613688

>>65613647
everything is a meme anon

>> No.65615565

>>65612582
>300+ fps
>18% CPU usage
Beast.

>> No.65616032

>>65613075
Very rarely going over 50% usage, man

>> No.65616041

>>65615565

My 1600x hardly ever hits 20% when playing Rise of the Tombraider.
The "Ryzen sucks for gaming" meme is fucking retarded.

>> No.65616135

>>65616041
>The "Ryzen sucks for gaming" meme is fucking retarded
Just intel shitposters

>> No.65616153

>>65615565
>>65613075
>>65616032
>>65616041
check on the benchmark the GPU usage.
there's 5% more gpu usage on the 8700k for less fps.

>> No.65616459

>>65611974
When someone takes the performance crown from Intel for the first time in about a decade, that's a significant event.

>> No.65616809

>>65611974
/g/ - brandwars

>> No.65617009

>>65611922
Why disable turbo on X470? Isn't this the point of buying one of those mobos?

>> No.65617063

>>65612090
>the best processor money can buy
>dual channel memory
>muh adulthood

>> No.65617085
File: 83 KB, 1156x713, PK-IC-DK2.jpg [View same] [iqdb] [saucenao] [google] [report]
65617085

>>65611781
Yes everyone besides this guy

>> No.65617095

>>65616459
Eh, AMD hasn't been competitive with Intel in the recent past but it's not like we didn't know they were capable of it. AMD had it at times and lost it, now has it back.

If someone like VIA came out of nowhere and topped Intel, then there'd be something to be surprised about.

>> No.65617100

Imao intel fags are mad they can't use their graphs anymore

>> No.65617115

>>65611781
>did not have the spectre meltdown patches on the intel boards
(((accidentally)))

>> No.65617149

>>65611922
This is bullshit. Hexus is one example wherw they clearly state it was patched and results were the same as everyone.
Ian cucktress is a long time amd shill

>> No.65617157

Whats with 2700x needing specific type of memory and needing to adjust timings manually? Whats the best purchace money/performance in mind?

>> No.65617212
File: 36 KB, 402x367, ayymd btfo.png [View same] [iqdb] [saucenao] [google] [report]
65617212

>>65611781
>>65611789
>>65611801
>>65611816
i'm a fanboy of neither but amd fanboys are ridiculous

there's no way a 2600 is outperforming an 8700k. the spectre microcode updates did very little to change anything on my 6700k. i maybe lost 3 fps and maybe 9 write speed.

every other test has the 2700x 10-30 fps behind the 8400, 8600k, and 8700k

>> No.65617290

>>65617115
hehe

>> No.65617345

>>65611974
The only thing missing from the 2700X is integrated video, but DESU, anyone who is buying that chip will purchase a discrete video card, or doesn't know that they don't need a chip so powerful.

>> No.65617367

>>65617095
>VIA
Don't you mean Cyrix?

>> No.65617370

>>65617212
>Crystaldisk

Reminds me of the time when I first started in IT as support and I accidentally used the Shizuku version on some old lady's computer.

>> No.65617374
File: 82 KB, 640x640, 1493363142825.jpg [View same] [iqdb] [saucenao] [google] [report]
65617374

>>65612090
>be management fag
>actually no real use of ryzen 2700x as all my job can be done using my x230
>the only excuse will be gaming
>absolutely no desire to play them
Suffering.

>> No.65617386

Ryzen is very sensitive to Bios udates and memory, someone did comparison between motherboards and bios revision and there were significant differences.

Also memory a dual sided 2666 is faster than single sided 3200, and this is only happens for ryzen , Ryzen is new animal and most unless hey get lucky will have it poorly configured.

>> No.65617703

>>65617386
>building a computer requires you to do actual research instead of just buying shit off the shelf
woaoaoaoaaoaoaoa

>> No.65617834

>>65617157
>Whats with 2700x needing specific type of memory and needing to adjust timings manually?
Infinity Fabric (aka InfinityGlue™) and broken BIOSes that didn't accept XMP profiles. It's mostly plug and play by now because the platform matured a lot since launch.
>Whats the best purchace money/performance in mind?
Guru3D has diminishing returns after 3200, there's a bit-tech memory scaling test but I haven't read it yet

>> No.65617850
File: 123 KB, 1643x701, ryzen-2.png [View same] [iqdb] [saucenao] [google] [report]
65617850

>>65611781
The Computerbase test did have all Spectre and Meltdown patches enabled.

>The countermeasures against Spectre Variant 2 now available from Intel, Microsoft and AMD were active in the test on the platforms Intel Coffee Lake (Core i7-8700K, i5-8400), Kaby Lake (Core i7-7700K) and Ryzen 2000 (2700X, 2700, 2600X, 2600, 2400G and 2200G). All other CPUs, including Ryzen 1000, have been tested without active Spectre Variant 2 security measures. Countermeasures against meltdown were active on all Intel CPUs.

>> No.65617860

>>65617386
>dual sided
You mean dual-rank. But dual-rank memory has always been faster on AMD than single-rank due to bank interleaving

>> No.65617936

>>65617009
Then that's a motherboard comparison not a CPU stock comparison.

The reviewers want stock performance first, then overclock as separate thing. If they try to push the overclock settings as "stock", this defeats the purpose of the benchmark.

>> No.65618025 [DELETED] 
File: 125 KB, 960x878, 1487297768743.jpg [View same] [iqdb] [saucenao] [google] [report]
65618025

>>65617149

>> No.65618054

>>65617936
By now we know overclocking Ryzen+ is pretty useless.
You'll actually loose performance on lowly-threaded applications, and win almost nothing at full throttle. All this while shooting power consumption to the moon.
Given the difference in price, I find it much more adequate to spend the few bucks more, and let the ship manage its clock. It will save you days of testing.

>> No.65618073

>>65618054
The reviewers can(and they do) then compare with turbo and manual OC.

That point is to not dilute the initial stock sample with motherboard overclocks.

>> No.65618083
File: 61 KB, 510x681, p0xnyfcH82cVu3FiJOwWFwkV_hHc6ZR9F3vYLP4plxw.jpg [View same] [iqdb] [saucenao] [google] [report]
65618083

>>65611781
Hmmm.... I'm seeing the benchmarks but I'm really finding this hard to believe. And I'm an and fanboy

>> No.65618106

>>65618083
I'm getting a 2700X tomorrow.
Make a list of what you want tested, and I'll tell you how it works on X370 with 2933Mhz memory.

>> No.65618145
File: 203 KB, 720x480, (PNG Image, 720 × 480 pixels).png [View same] [iqdb] [saucenao] [google] [report]
65618145

>>65611781
Oh my god AnandTech has found out the secret to Ryzen performance! Quick! Put the test machine into a sealed room and assemble a team of top experts to find out what even AMD could not figure out!

>> No.65618204
File: 29 KB, 967x553, shit was cash.png [View same] [iqdb] [saucenao] [google] [report]
65618204

Who /Ryzen/ here?
Already got my RAM/SSD for my future Zen 2 (Ryzen 3) build.
I can fucking re-sell the SSD on Ebay for around $260 and walk off with a little over-$100 3600 RAM.

You guys should've checked Newegg lately, shit was cash yo.

>> No.65618258

>>65618106
Can't think of anything. But I think you should try benching OPs benchmarks. Even if you set your rig up shit. You should be able to beat those Intel fps

>> No.65618268

>>65618204
Don't mind me, but Ryzen+ is rated for 2933Mhz memory.
Overclocking your memory to 3600Mhz might damage the cpu's memory controller. So it obviously voids the warranty. Even if you don't oc the cpu itself.
Overclocking is fun, and most likely won't hurt your cpu, but you should save it for when the warranty's over, and you're considering an upgrade.

>> No.65618386
File: 39 KB, 680x458, 55a.jpg [View same] [iqdb] [saucenao] [google] [report]
65618386

>>65612105
>that pic
>Asian woman as CEO = terrible drivers

>> No.65618780

>>65617149
>he believes intel can compete with amd by throwing moar coarz and niggahurtz forever
Your monolithic garbage is on its last legs, schlomo.

>> No.65618925

>>65617345
having integrated video is very useful for troubleshooting. I wouldn't have been able to use my video card on my intel board without it because I had to update the bios in order for it to recognize it first.

>> No.65618952

I'm upgrading from i5 3470 to Ryzen r5 3600, that's it!

>> No.65618983

>>65616041
https://www.youtube.com/watch?v=6mHUWvMDcMI

asks a 7700k user to do this

>> No.65619136

>>65612582
>Ryzen faster in Fallout4
Nani the fuck?

Only thing it's significantly behind in is Arma.
Framegraph is smoother in everything, too, except Arma and Project Cars which looks the same.
Speaking of, holy fuck that flat consistent FPS in Project Cars is so good.

>>65611781
Tom's Hardware used all but Spectre2 patches.
And Anandtech's 8700k results are hardly lower. It's their Ryzen results which are unusually high.

>> No.65619471

>>65617345
Ryzen isn't really meant for gaming though (at least not in 1080p), nothing wrong with buying Ryzen and a mid-/low-range GPU to do coding/compiling, maybe even an RX 570 or 560 for maximum freedumbs

>> No.65620576

>>65618268
This RAM is for Zen 2 though (next year's Ryzen 3).
They might have 3200MHz support by then.
I currently have a 1600 at 2666HMz so I'll be waiting for the next Ryzen.

>> No.65620734

Does anyone know if X470 boards with a PLX PCIe switch are a thing or will come out? I want 2 PCIe x16 slots for SLI - so only bandwidth from card to card (not to CPU) really matters. A 2700X on X470 would be sweet, otherwise I'd have to go with a more expensive Threadripper build or Skylel X.

>> No.65620905

>>65619471
fuck off retard, you're talking about ThreadRipper. Ryzen is very much for gaming, but it excels far more at multi-threaded workloads.

only reason it's slightly worse at gaming is because games tend to be less threaded and thus higher IPC usually gives better result

>> No.65620973

>>65618106
don't bother, intel pajeets will discredit you, it doesn't matter how hard you try. AMD might have won the performance war but Intel will always win the PR judaism war.

>> No.65621077
File: 37 KB, 600x450, Jackie-chan-meme.jpg [View same] [iqdb] [saucenao] [google] [report]
65621077

>>65612110
> I was loyal to Intel and bought a new motherboard every time they refreshed.
>every time they refreshed

Why?
They hardly made any progress with each release.

>> No.65621089
File: 1.88 MB, 288x288, 1504121962694.gif [View same] [iqdb] [saucenao] [google] [report]
65621089

>>65611922
They didn't have XFR enabled in the review? And Ryzen 2 still did that well in gaming?

>> No.65621551

>>65618145
Ian really outdone this time on vassalage role.
Computerbase otherwise always nailed. They were the first to outline perfectly bf1 multiplayer core scale factor very well.

>> No.65621572

>>65618268
>Oc ram to 3600 can damage CPU

Meanwhile cofeve lake deals with 4000+ without even caring

>> No.65622466

>>65621572
Ryzen is usually within 97-99.5% of theoretical RAM bandwidth while Intel had been at 80-85%.

>> No.65622707

>>65611781
sir pls delete this. thank you

>> No.65623438

havent seen any updates on this yet
It seems its not intel that went down on performance but rather AMD is massively up
Not sure what anandtech did but we will see

>> No.65623449

>>65612007
that can change quickly. remember athlon64, opteron and athlon64x2, and the FX cpus?

amd were ahead back then, and if intel doesnt get wise soon, they will be again

>> No.65623459
File: 5 KB, 727x432, 1470712905701.png [View same] [iqdb] [saucenao] [google] [report]
65623459

INTEL BTFO

Am i doin it right?

>> No.65623496

>>65623459

I miss the era of woodscrews and fermi fires.

>> No.65623521

>>65623449
Lmao amd was never ahead with the FX series, all they did was take one step back

>> No.65623861

>>65623521
Are you too young to remember that some of the first Athlon 64's were branded as Athlon FX?

>> No.65623955

>>65612130
he's just pointing out reviewer's inaccuracy
something intel ALWAYS use [most recent, remember the meltdown news fuck fest?]

>> No.65623956

>>65623861
A lot of people on this board are too young to know of the market assrape that was the Athlon 64 series.

>> No.65624088

wait, wait
so security patch fucked intel CPUs?

>> No.65624141

>>65623861
FX as in the amd version of the extreme series sucked balls too (for the cost), the only reason they looked good at first was because athlon 64 on 754 did not have dual channel. When 939 came out like a year later it made that FX series shit.

>> No.65624191
File: 431 KB, 500x383, 24c.gif [View same] [iqdb] [saucenao] [google] [report]
65624191

Its weird. I'm so surprised by these results I actually had a dream where I was personally asking all the tech reviewers what the fuck was going on.
How long till confirmation? Cause this is insane. As much of an and fanboy as I am I actually feel sorry regardless of all the shit Intel's done.
Seriously how the fuck will they recover!!?

>> No.65624218
File: 67 KB, 720x540, 1516473351575.jpg [View same] [iqdb] [saucenao] [google] [report]
65624218

>>65623955
>Every review says zen lags behind on gaming
>Even compared zen to zen+ all reviews seen coherent
>A single guy give completely different numbers where zen+ is well ahead Intel in all scenarios

The only logical conclusion is that all reviews are wrong just anantech is right because reasons

>> No.65624352

>>65624218
>it's not like intel has paid for skewed reviews before

>> No.65624606
File: 91 KB, 825x777, 8f7421b8a4.png [View same] [iqdb] [saucenao] [google] [report]
65624606

>> No.65624671

>>65624606
When you consider where AMD was 18 months ago when the best they had to offer was a Construction Core APU, that is a fucking MASSIVE jump in reaching near-parity with Intel.

>> No.65624812

>>65617850
>unbiased, equally footed test benches
>results that have been verified by literally everyother review outlet.

NOPE DOES NOT COMPUTE AMD IS THE BEST THEY WILL SAVE GAMING

>> No.65625175

It doesn't matter. Initial results are the only important ones. Ryzen+ could be twice the speed in follow up reviews/retesting but your average bozo looks at the initial reviews, see's Intel ahead and buys that.

>> No.65625212

>>65624606
>1% minimum framerates
wat

>> No.65625305
File: 605 KB, 1445x1122, 5.3-4000c16.jpg [View same] [iqdb] [saucenao] [google] [report]
65625305

>>65621572
4000c16 is easy on Coffeelake

>> No.65625349

>>65625212
"99th percentile minimum" and 1% minimum mean the same thing
In any average minute, .6 seconds are spent at that FPS
or,
During an hour of gaming, the FPS for 36 of the 3600 seconds is at this minimum level.

Fun facts:
>avg performance gain from 1800x across all tests is 9.6%
AMD claimed ~10% total improvement
>avg performance relative to stock 8700k is 97.3%,
>removing highest (Anand) and lowest (PCGH) results
>avg perf relative to 8700k is 95%

>> No.65626041
File: 1.08 MB, 1550x1154, 5.3-4200C16.png [View same] [iqdb] [saucenao] [google] [report]
65626041

>>65625305

>> No.65626106

>>65625305
>>65626041
Nice review sample

>> No.65626183

>>65624671
Yet it don't explain Ian cuck strange numbers. I mean kudos for amd, it's better, hope it goes beyond Intel so I can get zen2. Meanwhile Ian have some explain to give, instead of insinuating all others are wrong as he said

>> No.65626193

>>65626041
>Z170
>8700k
>Strange numbers

I think all fit in

>> No.65626282
File: 39 KB, 402x367, ssd numbers.png [View same] [iqdb] [saucenao] [google] [report]
65626282

>>65617212
what ssd are you using?

>> No.65626295

>>65612105
kek, is there one for Intel or Nvidia?
>still putting adored there when he explicitly trashed amdrones

>> No.65626319
File: 408 KB, 628x621, 1517135609852.png [View same] [iqdb] [saucenao] [google] [report]
65626319

hang on lads let me get my oscilloscope and debunk this

>> No.65626326

>>65626319
Better post a few sponsored papers just in case.

>> No.65626348
File: 445 KB, 1121x589, 1524280444602.png [View same] [iqdb] [saucenao] [google] [report]
65626348

I don't believe these numbers at all, but even if they're not true, the 2700x has showed how incredible it is out of the box.
It wasn't enough to make me upgrade from a 1700, but if they're challenging the top of the line gaymen chip with a process not catered to it now, they will fuck Intel up with 5GHz 6-12 cores next year.
Brian is sweating, I bet their 8core mainstream from this year won't even hit 5GHz.

>> No.65626381

>>65611781
>fucked up
>implying it wasn't intentional

>> No.65626416

>>65626348
>I bet their 8core mainstream from this year won't even hit 5GHz.

It will but with a TDP of like 120w

>> No.65626426

>>65624088
it's more like, they cheated to get more performance.

>> No.65626503

>>65626416
5ghz 1.2v 8C = 250W not 120W.

>> No.65626513

>>65626503
125W TDP at 3.7ghz.

>> No.65626943

>>65626513
Half correct.
7820 is 140w if I remember correctly.
My 8700k already peaks 125-130w 5ghz.
8 cores 5ghz might not be possible due to silicon limitations and would be ahead of 250 considering how much you need to drive the current for more cores and frequency

>> No.65626960
File: 2.39 MB, 640x360, 1511400732077.webm [View same] [iqdb] [saucenao] [google] [report]
65626960

Too many words. Is ryzen good yet? Should I buy it or the i5 8400 for gaming?

>> No.65626982

>>65626960
120hz btw

>> No.65627015
File: 35 KB, 402x367, ssd numbahs.png [View same] [iqdb] [saucenao] [google] [report]
65627015

>>65626282

nvme so gud

>> No.65627064

>>65626960
leave /v/eddit tourist

>> No.65627066

>>65627064
nah

>> No.65627092

>>65624606
r5 2600 is value king

>> No.65627118
File: 1.25 MB, 1845x1923, 1492138786706.jpg [View same] [iqdb] [saucenao] [google] [report]
65627118

>>65626295
Intel is still the supreme jew king

>> No.65627136

>>65627118
>nose vacuum
Every fucking time

>> No.65627146

>>65626295
>when he explicitly trashed amdrones
???

>> No.65627155

>>65626982
r5 2600

>> No.65627232

>>65627146
Maybe you should watch his videos.

>> No.65627321

>>65627015
i'm more or less meh with my nvme, if a ssd was cheaper and saturated sata I would have gotten that, but with an nvme and sata saturateing ssd costing the same, I went nvme

>> No.65627508

>>65623521
underageb&

>> No.65627521

>>65626041
good god man, that theme

>> No.65627899

>>65612152
Say it with me now, MAT LAB

>> No.65628769
File: 3.91 MB, 7000x7000, 1515193344699.jpg [View same] [iqdb] [saucenao] [google] [report]
65628769

>>65627118
Even better

>> No.65629175

>>65626943
7820 is 140W at 3.6ghz not 5ghz.

>> No.65629259

>>65626943
https://www.youtube.com/watch?v=GRviKkVUAa0#t=6m07s
Chart with 2 QS and 10 retail chip power draw ( 230 at ~5ghz to ~280 at 5.1/5.2 )

Assuming 5ghz at 1.20v and custom loop because lottery otherwise 5.2 at 1.416v sans delidding. An 8 core chip brings it to 4/3x that chart roughly. And he didn't -2 AVX I believe.

>> No.65629328

>>65611781
Fucking this. Also, Gamer's Nexus was the only youtube tech reviewer who said anything about lowered voltage (from like 1.4v at 4ghz on 1000 series to 1.1v at 4ghz on 2000 series)

Intel is the most disgusting company I have ever seen in my entire life.

Press S to spit on Brian

SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS

>> No.65629353
File: 83 KB, 681x592, fff.png [View same] [iqdb] [saucenao] [google] [report]
65629353

Nothing unusual here

>> No.65629381
File: 37 KB, 780x438, 1516778716769.jpg [View same] [iqdb] [saucenao] [google] [report]
65629381

>>65629353
wtf

Well, we can be sure that this isn't DX12/Vulkan magic since Rocket League use DirectX9

>FUCKING DIRECTX 9

>> No.65629582

But the real question here is which CPU plays Dwarf Fortress at the highest fps?

>> No.65629743
File: 82 KB, 1000x800, 1523029116592.jpg [View same] [iqdb] [saucenao] [google] [report]
65629743

>>65628769
>my OC in this pic
feels good man

>> No.65629767

>>65629381
How playable is that game on actual DirectX 9 cards? The GeForce 7000 series 12-13 years old now.

>> No.65629774

>>65629353
How do you only get 122 FPS average in Rocket League with a 1080?
My 1600X and 7970 get well over than in Rocket League. I'm GPU bottlenecked on it.

Their 1800X results were way low and they didn't retest them.

>>65629582
8350k probably?

>> No.65630081

>>65612587
The sky is purple..
We just see it blue because of physics magic btw

>> No.65630098
File: 135 KB, 383x364, Screenshot_2018-04-21_17-03-30.png [View same] [iqdb] [saucenao] [google] [report]
65630098

>>65627015
Oh, that's nothing
https://www.youtube.com/watch?v=QMfb_kmGLh0

>> No.65630142
File: 111 KB, 1280x720, 1475809244524.jpg [View same] [iqdb] [saucenao] [google] [report]
65630142

>>65630098
Not even my ramdisk is that fast

>> No.65630143

>>65629328
I don't know, man, my 4 hours old 2700X uses like 1.3-1.4V on auto settings.
I don't really care, because no app I use regularly will ever push this thing to 100% for a long amount of time.
Not sure why I upgraded. At least it doesn't get back to 3.7Ghz whenever it's actually used.

>> No.65630224

>>65630098
where do I buy my raid keys?

>> No.65630441

>>65630143

>Not sure why I upgraded.

'Cuz Intel is for incels and Ryzen rocks.

>> No.65630453

>>65611781
>105W

yeah but its a house fire tho.

>> No.65630529
File: 165 KB, 1311x1008, 2018-04-21 16_41_29.png [View same] [iqdb] [saucenao] [google] [report]
65630529

>>65630453

Actually no, it's 10 more watts than Intel i7 8700K while having 2 more cores and 4 more threads

>> No.65630585

>>65617850
Yeah, i'm Not getting out of the 7700K until 2020.

>> No.65630616

>>65629328
>1.1v at 4ghz
so, considering you need to raise voltage significantly to get 4.2, locking it at 4hz OC would be the most efficient setup?

>> No.65630646

>>65630143
My 1600X goes up to 1.425v on auto without OC, too. That's when it's heavily turboing itself.

Still stays under 48C on a torture test with fan quiet, though.

>>65630616
Most efficient set up seems to be to just leave it stock.
XFR2 works very well.
If you can get 4GHz all core at 1.1v, that might be better for like a home server that's constantly working, but for general desktop use it seems best to just leave the 2700X stock unless you have a golden sample that you can get to 4.4.

>> No.65630739
File: 61 KB, 1000x800, 902.jpg [View same] [iqdb] [saucenao] [google] [report]
65630739

>>65630646
>Most efficient set up seems to be to just leave it stock.
I was just planning to build a relatively small system with EK 240g loop (single 240 rad for both CPU and 1080ti) and I was thinking about the best temp/performance setup. So if it will be auto boosting to 1.4 to get ~200mhz more seems like a waste of temps for me.
or am I overthinking this?

>> No.65630825

>>65624606
>>65627092
the chart is missing 8400, which performs ~1% worse in gaming than stock 8600k but costs 50€ more (and would top 2600 in value)

>> No.65630852

>>65611974
Subtle Jewish shill

>> No.65630912
File: 401 KB, 1920x1249, cant even beat a 8400.jpg [View same] [iqdb] [saucenao] [google] [report]
65630912

>>65611781
hmmmm

>> No.65630937

>>65629328
Shouldn't this be obvious? If you increased the clocks but the voltage didn't change. You can therefore use less voltage at the previous threshold?

>> No.65630940

>>65630739
The thing is it won't be boosting that much... It'll be at like 1.125-1.2625v most of the time.

You could potentially clock it lower. I can manually clock my 1600X to 3.6GHz all core at 1.0875v, but then I lose out on the turbo. Not worth it really when it costs some performance.
And that's with XFR.. I think XFR2 is a lot less wasteful, isn't it?

You're over thinking it. The difference of like 65watts at 1.125v vs 140watts at 1.4v might sound huge, but your GPU is going to use much more power.
And that's not much power. A vacuum is generally like 1500watts. Your total system power of 60-350watts is only going to be a few incandescent light bulbs.

>>65630825
It's missing the 8400 because it would be embarrassed compared to the 2600 in non-gaming.

>> No.65630999

>>65630940
>talk about comparing cpus and their price and performance ratio in games
>It's missing the 8400 because it would be embarrassed compared to the 2600 in non-gaming.
Please tell me you amd fanboys aren't this dumb.

>> No.65631041

>>65624606
i hate these performance per dollar measures when they don't take in the motherboard, intel motherboards are more expensive

>> No.65631056

>>65626960
The 8400 is still the best value just for gaming, but the 2600 is more future proof (2x threads) and is far more powerful in non-gaming tasks.
It's also at most 15% behind in raw FPS with a 1080ti, which is a GPU you won't be buying if you're getting a 8400 or 2600

>> No.65631063

>>65611781
Who fucking cares, 10 years old cpu is enough for you guys as you don't play, just shitpost everyday.

>> No.65631171

>>65629175
Yes and no. Intel tdp is hard set as the maximum power allowed for the chip.
I found this with z270 board on default settings, example 8700k is 95w, even if you oc, it will allow the clock to go to your setting, but it drops frequency very quickly, reaching 4.2-4.5 or whatever among cores, until it can sustain 95w. Once you remove current limit it will go beyond.

Same as skx, it will fluctuate frequencies among cores as long as it is within tdp.
Intel by specs allow a short current spike for a limited period of time, but it's defaulted to soon recover into tdp, so Intel on normal settings is hard limited by tdp.

>> No.65631210

>>65631056
what's wrong with getting a 1080ti with 8400/2600?

>> No.65631339

>>65611922
Any news on this?
Is he really the one with the correct benchmarks?

>> No.65631378

>>65612090
>There is no reason to buy anything else.
exept there is waiting for zen 2 which everybody with an old ass intel can still do while ampoors shitty fx systems root in hell

>> No.65631410

>>65631041
except they arent you jew shill. intels boards dropped last week and intel users can run a i5 8400 on h boards while ocing a 2600 on a320 is suicide

>> No.65631469

>>65631210
Loads of CPU bottlenecks, I mean even the 8700k can't get the highest performance out of the 1080ti.
And its not that I found it wrong ,but unlikely, since most people want the best performance out of their 1080ti.

>> No.65631490

>>65631410
B350s are very cheap. 8400s don't OC much at all.

>> No.65631610

>>65631410
Because you might have missed the fact that all of the processors int here are unlocked

>>65631490
>don't OC much
the i5 8400 is LOCKED. It doesn't OC at ll. It's also non hyperthreaded, it's a 6c/6t part.
If that value does not lose its appeal after you consider the fact it has half the threads than another cpu that's only 20 bucks more expensive, then you're a fucking shill.

>> No.65631655

>>65631469
Wendell said his Ryzen 2700x was the only proc he had that could drive his MSI Triforce 1080ti, in regards to cpu utilization.

>> No.65631672

>>65631610
Which is unlocked and comes with a quite decent cooler out of the box.

>> No.65631679

>>65631469
That is very heavily dependent on game and resolution. I seriously doubt you can't choke a 1080ti on 8700k.
The only setup where I found CPU bottleneck is sli of 1080 or 1080ti.

I only managed to get 1440p 144hz stable on bf1 mp with 8700k 5ghz.

Most scenarios you end up bottlenecked by gpu even on 1080p

>> No.65631744

>>65631679

CPU-bottleneck is only an issue with low resolution or certain niches (Grand strategy, massive MMORPGs, dwarf fortress, RTS)

>> No.65631795

>>65631744

There are many, many waus for a cpu to bottleneck a gpu almost regardless of hardware configuration. Its a major issue in vidya specifically due to limitations of APIs and gpu drivers. There is far, far more to performance than the silicon itself.

>> No.65631796
File: 135 KB, 1280x720, maxresdefault.jpg [View same] [iqdb] [saucenao] [google] [report]
65631796

>>65631655

>> No.65631822

>>65631672
and ultimately performs worse, you can list all kinds of "features" that make it go faster, but if even after all that it's slower, whats the fucking point.

>> No.65631837

>>65631679
>>65631744
think of it as a metric of hitting 100% utilization for extended periods of time. Not what the FPS counter shows.

>> No.65631927
File: 45 KB, 500x970, cinebench_multi.png [View same] [iqdb] [saucenao] [google] [report]
65631927

>>65631822
>performs worst
>for you
face it, you'll never buy anything AMD branded. You're here to shitpost exclusively.
that's a 974 MT score for the 8400
vs
1316 for the 2600x, no non x there.

But I already know you'll say who the fuck for Cinebench anyways?
Well, every single intel fanboy who were already capable of reading before 2010 did. They loved it so fucking much you wouldn't believe. It's almost at the same level of your infatuation for gaming benchmarks, I mean, Far Cry, GTA. YOu're thinking of ROTR as betrayer now, but you still loved it 2 weeks ago. Now you don't anymore. That's quite telling.

>> No.65631948
File: 48 KB, 500x970, rise-of-the-tomb-raider_1920_1080.png [View same] [iqdb] [saucenao] [google] [report]
65631948

>>65631822
for you, with love.

>> No.65631978

>>65612130
>discredited by who?

anandtech is MSM, so they've discredited themselves by default - most likely AMD just paid them off to fudge the numbers

>> No.65631986

>>65631655
and that's why it gets less fps in games than 8400?

>> No.65631997
File: 15 KB, 683x301, 3b8d61023a.png [View same] [iqdb] [saucenao] [google] [report]
65631997

>>65631822
That's a 300% + win for all Ryzen procs over any intel.
You'll say it's invalid, but that just proves your confirmation bias even further.
Those single digits gonna hurt you real bad.

>> No.65632011

https://www.extremetech.com/computing/267915-psa-the-amd-ryzen-7-1800x-is-faster-than-it-used-to-be

>> No.65632047

>>65631986
shoo shill
Your script is getting less authorized topics for discussion everyday. One day you'll go back to athlon levels where only lies and misrepresentations could create good press and positive reactions. It saddens me that you fuckers actually were indredibly successful in that endeavor. You'll get the good old soon, don't worry. I hope Millenials can be as bad (good) as Gen X in shilling tactics. You gonna need it. Well, your boss will.

>> No.65632161

I found pretty funny that even though PUBG was one unarguable successful game, no one dared to benchmark it, simply because everybody was keenly aware that it's a piece of shit code wise and always produced laughably biased results.
Guess what intel told them change in their new reviewing guidelines, I mean, counter-reviewing. That they should dust off those anachronistic ways and that OC results for HEDT and K series should get (((shutdown'ed))), especially power concumption graphs. Did anyone else notice all of those graphs for Ryzen 2 vs Ryzen 1, or just OC vs stock on most sites?
Yeah, tha was the power of his boss.
>>65631986
>>65631822

>> No.65632167

>>65632047
I'm sad seeing you do all this for free.

>> No.65632220

>>65632161
>PUBG
Well, gn has tested 2600x vs 8600k in pubg streaming, where 8600k literally was delievering slideshow.

>> No.65632228

>>65632167
nice freudian slip.
we all knew already anyway.

>> No.65632262

>>65632220
Well, I very much doubt intel was expecting that when they told reviwers to bench it. They never thought Steve would find a way to put it under a bad light.
It simply backfired.

>> No.65632327
File: 46 KB, 500x970, excel.png [View same] [iqdb] [saucenao] [google] [report]
65632327

>>65612152
another one bites the dust

>> No.65632346
File: 573 KB, 2220x1080, Screenshot_20180421-191246_YouTube.jpg [View same] [iqdb] [saucenao] [google] [report]
65632346

>>65632047
what? I'm not a shill, I'm trying to pick a processor for my new rig and all I see is people trashing 8400 here. But then I see it outperform 2700x in majority of games. And when I'm trying to understand what's going on I'm called a shill kek.

>> No.65632352
File: 50 KB, 500x970, powerpoint.png [View same] [iqdb] [saucenao] [google] [report]
65632352

>>65612152
>>65632327
and another, another one

>> No.65632370
File: 796 KB, 1290x734, geekbench 3 WR.png [View same] [iqdb] [saucenao] [google] [report]
65632370

>>65612152
>>65632327
>>65632352
TUM, TUM TUM.

>> No.65632381

>>65632346
Bingbus is better at pushing easy drawcalls, more news at 10.

>> No.65632419

>>65632346
AMD boys are really really insecure about their purchases. The bulldozer made them jumpy anon.

>> No.65632456

>>65632346
>And they're not wrong. Our review of the Ryzen 7 2700X revealed that AMD's current flagship processor has surpassed Intel's Core i7-8700K in multi-threaded workloads while also closing the gap in single-threaded workloads. While Intel still has the advantage when it comes to gaming performance, the difference in performance is slim and gets even smaller as you climb the resolution ladder.

Because you're trying VERY HARD to substanciate your buyers remorse over the fact that one processor is btfo'ing the other performing tasks meant for a cpu, number crunching, encryption and productivity by a LARGE margin while it still lacks A BIT on exclusively gaming averages, You can argue from 1% at 4k to 8% at 1080p, with outliers where it actually have some wins too. So you're blinding yourself to the fact that you want your recent purchase validated no matter what, nothing else.

>> No.65632486
File: 1.83 MB, 200x200, 1503634516812.gif [View same] [iqdb] [saucenao] [google] [report]
65632486

>>65632370
geekbench.
intel lost.
remember kids:
if_intel_genuine: optimized path

motherfuckingGEEKfuckingBENCH

>> No.65632498

>>65632419
nice timing, anon.
Gosh, that was karmic.
>>65632456

>> No.65632512

>>65632419
>The bulldozer made them jumpy anon.
Oddly enough, Bulldozer is aging like a fine wine.
https://www.youtube.com/watch?v=16_9mBUCf_c&list=RD16_9mBUCf_c

>> No.65632539
File: 25 KB, 650x365, wwQozWbe5Wkw495mSexa6K-650-80.jpg [View same] [iqdb] [saucenao] [google] [report]
65632539

>>65632486
geekbenck3 WR fell on Ryzen's launch day.
geekbench4 gonna need some fixes, soon.
>notice the ST

>> No.65632806

So the real MAGA was Make AMD Great Again and it worked

>> No.65632888

>>65632456
>buyers remorse
I didn't buy anything yet, retard.
>performing tasks meant for a cpu, number crunching, encryption and productivity
well I'll go buy a lawnmower and ride it to work because it apparently cuts grass better than a toyota

>> No.65633007
File: 46 KB, 640x393, 54caa867cc1f3_-_puzzles_00_0710-650.jpg [View same] [iqdb] [saucenao] [google] [report]
65633007

>>65632888
to the playground you mean.
you are working, right now.
those trips might have even brought a nice bonus shekel.
>also, implying that's a bad thing

>> No.65633026

>>65632888
>>65633007
that's the i5-8400 fac-simile in all of this.

>> No.65633042
File: 26 KB, 220x293, Pushmower1.jpg [View same] [iqdb] [saucenao] [google] [report]
65633042

>>65633026
**that's the i5-8400 fac-simile in all of this.

>> No.65633056

>>65631927
>face it, you'll never buy anything AMD branded. You're here to shitpost exclusively.
Not really I buy the best computer component for my needs and bang for buck is also considered.
At some point it was AMD cpus, and I was shitting on intel users as much as I am shitting on you right now.
Same thing with graphics cards, at some point amd cards were more powerful and cheaper, so naturally I bought one and shat on nvidia idiots.
Right now, amd products are inferior for gaming. Keyword for gaming, because some of you can't accept that.
The fact that amd went for an easily scalable architecture that lets them have high yeilds and corner server market is great.
That does not mean though that I am going to cum in my pants and buy a cpu that has an edge in productivity but is also good at gaming vs a CPU that has an edge in gaming but is also good at productivity.

>But I already know you'll say who the fuck for Cinebench anyways?
Correct. Benchmarks of actual applications you plan to use is far more useful info.

>>65631948
Can you post minimums or 99% percentile or something? Average fps while a nice stat to know, is secondary to the two mentioned above.

>>65631997
How are those relevant to gaming? Because you realize the original post chain is discussing gaming and the 8400 being value for bang.

Or is your actual defence:
>go ahead tell me non gaming benchmarks are irrelevant to accurately measuring gaming performance
Because if that is your case, you are literally borderline retarded.

>> No.65633082

>>65632346
>what? I'm not a shill, I'm trying to pick a processor for my new rig and all I see is people trashing 8400 here. But then I see it outperform 2700x in majority of games. And when I'm trying to understand what's going on I'm called a shill kek.
AMD shills in full force, they literally don't want to admit their favorite company isn't perfect in all regards, so even though AMD is good for productivity in most tasks, it's lacking in gaming performance. This of course wont do, and they will call anybody a jew for even mentioning this and providing proof.

>> No.65633169

>>65633056
Nice, you called the office manager.

>oldest trick in the shilling guide
>find them yourself, you gonna be very disappointed though
>reading comprehension, you lack it. Or it was a bad goal post shift. Either way you both failed. Call the branch office manager or phone your intel contact for more details.

>> No.65633203
File: 540 KB, 1327x1103, 1524149075721.jpg [View same] [iqdb] [saucenao] [google] [report]
65633203

>>65633082

>> No.65633243

>>65633169
>Nice, you called the office manager.
You would know from experience?

>find them yourself, you gonna be very disappointed though
Didn't say that though now did I? I said, average frame times are meh and minimums and 99% percentile are a better benchmark for gaming performance.

But go ahead try to shill your way out.

>>65633203
>average frame times
>time below 60 fps, when there is people with 120/144hz monitors
Is it really so hard to comprehend for you or something?

>> No.65633270

>>65633082
>AMD shills in full force
said the 4th intel shill present ITT

>> No.65633280

>""""breaks"""" something so he can't possibly release an early review of Ryzen 2xxx
>releases a video entirely focused on the worst aspect of Ryzen 2xxx several days later
https://www.youtube.com/watch?v=RZRjoeyz4Z0

is there a worse intel shill than the Jay of many Shekels?
Even the comment section is roughly 99.5% intel drones

>> No.65633304

>>65633280
>trying to manually overclock instead of trying to make the most of PB and XFR2

Literally not even trying to make a good review.

>> No.65633353

>>65633280

I dunno what Jay does but he seems to actively fuckup every AMD specific (be it gpu or cpu) review he does. He is to /g/ what DSP is to /v/.

>> No.65633357
File: 721 KB, 3507x2480, ganbatte.jpg [View same] [iqdb] [saucenao] [google] [report]
65633357

>>65633243
Keep doing Brian's work, anon!

>> No.65633369

>>65631056
>It's also at most 15% behind in raw FPS with a 1080ti

Yikes. 8400 it is then.

>which is a GPU you won't be buying if you're getting a 8400 or 2600

Yeah, nah.

>> No.65633463

>>65633280
shut up goy

>> No.65633515

>>65633203
Basically the only benches that list 2700 ... sad.
I'd be interested to know why the 2600 performs close to equal to the 2700

>> No.65633523
File: 221 KB, 682x599, MultiCoreEnhancement BTFO.png [View same] [iqdb] [saucenao] [google] [report]
65633523

>>65633280
Any honest reviewer noticed that OC'ing Ryzen 2700x brought worst results than out of the box, stock experiences, speaking to the merits of Precision Boost and XFr2, rather than emphasizing on how "badly" it overclocks. It really doesn't OC very much, apart from exotic cooling, on ln2 for benching, most users will have a better stock experience rather than OC, unless they're using heavy MT renderers and the sort.
I'm a bit conflicted, got used to OC just to get that extra bit, not as a hobby. But from the looks of it, not gonna need it now or wait for better techniques and tricks from pro overclockers.
Only youtuber worth watching:
https://www.youtube.com/watch?v=FWzyInXgNXE

He addresses that in the first 15 minutes.

>> No.65633539

>>65633463
schlomo outs himself.

>> No.65633570

>>65633523
>Any honest reviewer noticed that OC'ing Ryzen 2700x brought worst results than out of the box

This is what i've gathered too just seeing what people do on OCN tweaking the XFR and PB, and even adding a bit of baseclock to make it boost slightly higher. Do certain reviewers do this knowingly? To sway the ignorant majority while still being able to claim innocence?

>> No.65633575

>>65633523
it illustrates how much they have done to automatically optimize boosting for any specific number of cores required by an application.
great job on AMDs part, really.

What has me wondering is the 6 GHz under LN2.
That's really close to the maximum clocks achieved with the 8700k.
Maybe we'll be able to squeeze out a couple hundred more under air/water with a Bios update or 5.

>> No.65633598

>>65633270
When your criteria for "intel shill" is
>can amd do no wrong, are amd products superior in every possible categoy
>no?
>you are a shill
Then it's no wonder you think there people like that in this thread.

>>65633357
>give the chance to show ryzen is superior
>instead of posting proof and proving his point decisively you start to shitpost and insult
Just so you know, to anybody from the outside, it just seems like you are a liar/shill, if you behave this way.

>> No.65633655

>>65633570
>>65633575
Perfect illustration, already posted.
Notice how the 4.2ghz OC wins in CB while in games it lags behind
>>65631948
>>65631927

Also, on most everyday task benches, the same applied. Games, and ST, much more pronouncedly due to the XFR2 kicking in -- and how good it is at its intended job.

>> No.65633678

>>65633598
stop. promoting. intel. 8400. you. are. not. fooling. anyone. anymore.

>> No.65633708

>>65633655
Yea. I've seen 2700xs boost beyond 4.5GHz for single threaded shit. Also, do you know anything about this PB and XFR tweaking?

>> No.65633749

>>65633708
nah, let it grow.
Many people gonna rethink real hard about the P state tables overclocking methods from now on.
I had a suspicion about XFR2, gonna try to dig an old post here.

>> No.65633792

>>65633678
Provide proof that 8400 is not the best bang for buck for gaming.
I understand you are upset you cant shill in an echo chamber and claim ryzen is best at everything, but that's reality.

>> No.65633810

>>65633708
>>65633749
xfr2 fell to the background now, as a dependency task of Precision Boost.
> XFR 2 does away with the idea of a fixed frequency increase across single-core and all-core workloads, as seen on all first-generation Ryzen products to some degree. Instead, XFR 2 works more like the Mobile XFR feature we first saw on Raven Ridge mobile chips.

first precision boost kicks in (25mhz increments across all cores entirely dependent on your own cooling solution), then if there's still room xf2 tries to push 1 or 2 cores a bit more (25mhz up to 100mhz, max 4.3ghz and dependant on precision boost increase, also XFR effectivelly raises the TDP rating, power envelope on top of precision boost)
>The second technology is XFR2. This is similar to Precision Boost 2 in that it concerns managing clock speeds. But here it’s about providing a final extra boost if the CPU isn’t too hot. XFR could boost the clock speed of up to two cores in increments of 25MHz, up to a maximum 100MHz above the rated boost clock, if conditions allowed.

Maybe they have a big table of derivative P sates (p' sates) ruled by Precision Boost sensors, or rather MI sense.
4.3+ 1T (xfr can kick in)
4.25 2T (xfr can kick in)
4.2 3T, 4T (no xfr for you)
hardcap
4.1 5T-10T (no xfr for you)
hardcap
4.0 if >13T (no xfr for you)

>> No.65633819 [DELETED] 

I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL I LOVE INTEL

>> No.65633838

>>65633819
>amd fanboy having a literal mental brakedown

>> No.65633846

>>65631171
I also remember lavalake-x reviewers having their samples default to all core 4ghz+ on the 10/8 core chips giving them 90deg temps on 240mm aios. Sure they performed great but it was probably a fire hazard as well. I haven't bothered to read a more up to date review to see if someone has some more sensible bios out, not that any should care about x299 in 90% of cases.

>> No.65633858

>>65633810
Is it correct that you can "add" even more to this via base clock OC? As far as i know upping the base clock will not disable PB and XFR. So let's say 105 base clock.

>> No.65633863

>>65633575
That derbauer OC wasn't even a serious attempt. A quick sponsored trip to Asus offices and he was using/showcasing their new Windows OC aplication -- after a few bios tweaks. Not dissing that, it was actually awesome that he got such a result under all of thsoe constraints. It made me real hopeful for even better results in the future.

>> No.65633865

>>65633515

Because they're the same CPU. The 2600 just has two cores disabled. If games were better optimized to take advantage of moar cores you'd see higher numbers out of the 2700 but as it is they might as well be the exact same CPU.

>> No.65633883

>>65633858
that's P state OC.
it's a bit of an ugly brother for the hardcore OC'ers. What you said is exactly what I meant that the 2700x will make a lot of people rethink that method.

>> No.65633888

>>65611781
so you say the score will be even lower for Intel ?
that's antysemytic goym, delet this !

>> No.65633897

>>65633810

It will be interesting to see how XFR develops as the generations continue. It being much more granular than Intel's turbo opens up a world of possibilities depending on the nuances of the manufacturing node. A hypothetical implementation that would be good to see would be to allow XFR3 on the 7nm to push 2 cores to ludicrous clocks while keeping the rest of the chip effectively idle - this works better the faster the race to zero (and I suppose by extension, the race to 100%) is.

I could imagine if AMD went mental again and made a zen equivalent of the 9590 that relies heavily on XFR being an aboslute monster once you let it loose from TDP values regular plebs need worry about.

>> No.65633909

>>65633865
So i didn't waste my money entirely. Thanks for replying

>> No.65633910

>>65633883
Alright, well thanks for answering my questions man. I thought this was the case so it's nice to get it confirmed. I'm getting my 2700x on Monday so it's going to be interesting to see how much i can push it on my NHD15 and C6H.

>> No.65633925

>>65633792
See the OP

>> No.65633972

>>65633897
What would be a neat addition to it would be the ability to set the desired TDP value directly, and have XFR take care of the rest.

>> No.65633979

>>65633910
another thing. It was aleady confirmed, via the bios updates that all of the xfr2 and precision new tricks are entirely on the cpu's. So, x370 mobos can take full advantage from them. How to check: search for the new options in your bios. Verified already: Asrock Taichi x370 and Crosshair VI

>> No.65634037

>>65611781
Spectre/Meltdown patches have a negligible effect on gaming benchmarks. It mostly affects I/O.

>> No.65634043

Has there been a real comparison on how the 2700x does on a variety of cooling systems? The reviews are all saying it draws up to 140watts while funnily enough the wraith prism is a 140watt cooler.
https://www.youtube.com/watch?v=zYKlMtiO9YY& has a pair and the better cooler is literally letting the chip run .025ghz faster in his test and he never mentions it, only seeing the extra fps in the game test.

>>65633865
>>65633909
If they're both set up normally both will run at a 65w tdp, just means the all core boosts for the 2600 will be slightly higher than that of the 2700.

But I'm waiting to see what the OC results on these two chips are and if they're capable of hitting 4.2/3 depending on the lottery or is it only the X skus.

>> No.65634049

>>65633897
All of those things the 1xxx series already had. I felt they were being severely underutilized. I believe the only mentions I saw from the plethora of temp sensors, complete with a marketeable name sense MI, were in press slides. Not much was said or done about it all, including XFR1, people were even ditching the x series as completely unecessary over the non x's.
This refresh for those 2 (or 3 if you count the sensors) things was very big.

>> No.65634064
File: 31 KB, 707x257, SPEKITER N MELDOWN DUN MATTER.png [View same] [iqdb] [saucenao] [google] [report]
65634064

>>65634037

>> No.65634115
File: 78 KB, 1323x1510, Excel Monte Carlo.png [View same] [iqdb] [saucenao] [google] [report]
65634115

>>65634037
hi intel, you used to be so good at this.
why areyou doing this to me now?

>> No.65634133
File: 55 KB, 500x970, wprime.png [View same] [iqdb] [saucenao] [google] [report]
65634133

>>65634037
almost as good as you were in w prime.
stop please.
we are running out of intel genuine benches.
staaaaaaaaahhhhhhhhhhhp
i-I can't take much more.

>> No.65634149

>>65634043
Personally i have to much shit running and i need to keep my bills low so i bought for the 2700 instead of the X. Will find out how good it is once the package arrives.

>> No.65634202
File: 167 KB, 348x342, confused anime girl.png [View same] [iqdb] [saucenao] [google] [report]
65634202

So with similar stock clock speed between the 2600x and the 8600k, and more or less similar IPC(?), why does intel perform better in most games?

Is it because most game devs are in bed with Intel and only optimize for intel cpus?

Can someone give me a thorough rundown of why?

>> No.65634218
File: 196 KB, 483x487, confused Sachiko on Thinkpad.png [View same] [iqdb] [saucenao] [google] [report]
65634218

>>65630912
The 7820X also loses, what's your point?
Oh no, I get 112 minimums on my 120Hz panel, whatever shall I do?

>> No.65634225

>>65634202
>Is it because most game devs are in bed with Intel and only optimize for intel cpus?

Pretty much. And some intel scheduler shit.

>> No.65634241
File: 380 KB, 747x768, Shrek.jpg [View same] [iqdb] [saucenao] [google] [report]
65634241

I'm going to be honest with you now. I really love onions. And yes, I actually mean onions. Looking forward to buying a new Ryzen 2 computer.

>> No.65634275
File: 99 KB, 401x477, Neat - some Yamatogawa whore.jpg [View same] [iqdb] [saucenao] [google] [report]
65634275

>>65633523
>that XFR graph
My fucking dick, I won't even bother to OC Zen2 TR.

>> No.65634279

>>65634241
Only poozen fans would appreciate such low-brow goyim humour

>> No.65634289

>>65634275
It's actually beneficial to not do a manual OC.

>> No.65634331

>>65633203
It's issues with the nVidia driver on Intel, and about a 5-8% loss from Spectre v2 patches. Anandtech got a normal sample and the framerate differences between different reviewers on the new Ryzen CPUs are identical.

>>65633523
XFR2 is a huge improvement. However, it also costs more power, the process is still pretty good, but people suggest it has a little more leakage current.

>> No.65634449

>>65630739
Set the clocks and volts yourself. XFR is not more efficient than manual tuning.

>> No.65634526

>>65634449
>t. jayz2cents

>> No.65634650

>>65618983
i am now aware clicker game powergamers exist

>> No.65634722

>>65626326
let this bread be buttered please.

>> No.65634917

>>65611781
>9% all-core boost frequency AND single core boost frequency
>most IPC improvements from cache layout improvements, L2$ latency cut down by 35%, L3$ latency cut down by 10%, around 5-8% behind CL in IPC
>no memory controller improvements compared to RR, likely around 50% of 2700Xs can achieve at least 3533MHz, and 9% faster DRAM access speed, similar to RR, X470 still needs maturation
>40% slower DRAM latency though, memory bound scenarios are as expected around 30-40% slower on Ryzen, so some more software optimization will be needed to leverage Pinnacle Ridge better, esp. on GCC since cache latency is up to 40% faster than Coffee Lake
>XFR2 allows much higher granularity to XFR, pulls 1.33V safely all core, 1.48V single core boost, power limit at 142W max (dependent on thermal/VRM parameters). Insanely bad power scaling after boosting past 4.1-4.15GHz, but about 23-25% more power efficient under 4GHz regime, but silicon is already close to limits, will draw max power (142W maximum) if XFR2 is given the opportunity to be stressed to 4.15 all core
>on linux performs almost exactly like 1800X, some software tweaks definitely needed
It will be interesting to see what happens when software catches up. With cache optimization, scientific computing on Ryzen (w/o AVX) will probably close the gap, very excited for that.

>> No.65634970

>>65617212
dip shit, regular sata SSD's are pleb tier in 2018.

It is about NVME's now which are about 6 times as fast. Intel lost almost 30% of performance on those.

>> No.65635034

>>65632512
>Oddly enough, Bulldozer is aging like a fine wine.
how so?

>> No.65635039

>>65634218
>The 7820X also loses, what's your point?
Am I suggesting it for gaming?

>> No.65635054

>>65635039
Not that guy but what's up with the 7820x anyway? Why is it so subpar compared to other intels?

>> No.65635087

>>65635034

Its a claim made by AdoredTV based upon the last few years of Computerbase.de's benchmarks. At face value that statement is correct using computerbase's data.

>> No.65635109

>>65634115
Excel doesn't matter.

>> No.65635126

>>65635087
Nice, I still have a FX next to my Ryzen

>> No.65635160

I got all excited and impulse bought a Ryzen 7 1700x with new case, ram and mboard. It looks really good with the lights but what do i do with it?

>> No.65635168

>>65634449
2700x performs better when it OCs by itself, 4.3 / 4.4 GHz

>> No.65635181

>>65635034
it still holds worlds overclocking record :^)

>> No.65635195

>>65635034
tl;dr modern software can make better use of FX's extra cores than they did 10/5 years ago

>> No.65635201

>>65635181
8722.78 mhz nice

>> No.65635223

>>65635160
>not zen+

>> No.65635224
File: 96 KB, 901x523, order.png [View same] [iqdb] [saucenao] [google] [report]
65635224

You will be missed 2500K.

>> No.65635256

>>65635181

IBM magic science node will do that. Bulldozer was designed to hit well over 5ghz but no mortal cooling solution (or psu) can handle that.

>9590 220w stock

...and that can be overclocked as well!

>> No.65635261
File: 38 KB, 657x527, 1522569054201.jpg [View same] [iqdb] [saucenao] [google] [report]
65635261

>>65635224
>He's gonna do it

>> No.65635272
File: 366 KB, 800x450, 1510664997044.png [View same] [iqdb] [saucenao] [google] [report]
65635272

>>65635224

>not the crosshair hero

>> No.65635294

>>65635272
Stock XFR2 seems to be hitting Zen+'s thermal/stability limits without needing extra tweaks.

>> No.65635298
File: 82 KB, 547x428, Prison School v01.zip-Prison School v01-143.png [View same] [iqdb] [saucenao] [google] [report]
65635298

>>65635224
The madman did it.
I'm a little jelly desu. I have a 1700, but want to upgrade. Zen2 is maybe 14 months away? With Threadripper probably more? AMD pls, I have shekels.

>> No.65635326

>>65635298
sell the cpu you have now and buy a new one? what's stopping you?

>> No.65635435
File: 287 KB, 760x901, distress-8e9828b2a180e0c78b6ee500a4a72a6b.jpg [View same] [iqdb] [saucenao] [google] [report]
65635435

>>65635326
Now that I think about it, the encodes are fine and emulation is mostly great but they could be better.
I don't even need a x470 mobo for XFR2 since I have a NH-D15 and could just manually OC.
It just seems like a waste, but it's tempting. I upgrade GPUs every generation, so it might be catching up to me in CPUs as well.
Then I think about how I want Threadripper on Zen2 and we're looking at maybe 1.5 years so 2.5 years if I don't upgrade now. Shit, I might just upgrade.

>> No.65635464

>>65635435
XFR2 and PB2 works on x370, it is a feature of the CPU, not x470.

>> No.65635483

>>65635464

Did you know that a SoC is anti-semitic?

>> No.65635496

>>65635483
Yes.

>> No.65635501

>>65633539
BAD GOY

>> No.65635519

>>65635464
I thought XFR1 was supported but not the second, maybe I misheard Wendell a few days ago.

>> No.65635578

>>65635519
As long as you have a Pinnacle Ridge CPU XFR2 will work on any b350-x470 mobo. Like i said; it's a CPU specific feature not a mobo one.

https://www.youtube.com/watch?v=YSMXbbw2B8Y

Now, i don't particularly like bitwit but it shows what you need to know. There are also several people in the C6H thread on OCN that has posted benches of their CPUs boosting over 4.5GHz with some base clock and XFR2/PB2 tweaks.

>> No.65635621

>>65635519
>>65635578
I'll add that tweaking XFR2 and PB2 in the bios is also better than manually trying to force 8 cores to run a fixed frequency because you won't have the same single thread performance that the boosts provide, you do lose a tiny bit of multi threading compared to manually fixing all cores to a specific clock but it's negligible. XFR2 and PB2 works REALLY well.

>> No.65635637

>>65635578
Hey looks, it's that newegg guy.
Thanks for the vid, I'll go look into it some more. I might just do this slight upgrade after all and just splurge whenever Threadripper on Zen2 hits.

>>65635621
I'll go look into this as well.

>> No.65635644

>>65635054
It's crap because Intel skx is the same arch as Xeon, it uses mesh instead of ring bus. Mesh is a solution that payoff for a high amount of cores, but Imo, until 8 cores it should remain ring bus, it's much more faster in inter core communication and has much less latency.
As derivative of Xeon purley, skx also has more l2 and less l3, which made l3 cache non inclusive witch done more performance penalties for general desktop apps. There is a lot of performance degradation when you rvict l2 cache and it hasn't a copy in l3 forcing you to go ram.

In this sense, I dont know what good really there is anymore on Intel hedt: core count will go to 8 on main customer platform; massive hot without solder; lack of pcie lanes; lower frequency, more latency and more costs. The only good thing is it's cheaper than Xeon and has 4 ddr4 channel but it's the only good thing a I can list. For a pure performance I would go threadripper as workstation if you don't need avx512

>> No.65635655

>>65635644
Intel is basically killing it's own HEDT in an attempt to stay on top of AMD is what you're saying.

>> No.65635696

>>65635644
>massive hot without solder
Doesn't matter
>lack of pcie lanes
Doesn't matter
>lower frequency
Doesn't matter
>more latency and more costs
Doesn't matter

>> No.65635705

>>65635578
>>65635621
Sometimes I forget how these chips are close to SoCs.

>> No.65635713

>>65635705
What the fuck are you talking about?

>> No.65635719

>>65635696
kek, that's great.

>> No.65635825

>>65635713

For ryzen the chipset provides effectively zero additional functionality.

>> No.65635972

>>65635825
It should be rephrased as: The chipset is not required for the CPU or the system to function as the CPU itself has all of the I/O (save video for the non-APU chips) to run as a fully functional machine. The chipset does provide additional I/O in the form of PCI-E 2.0 lanes, USB ports, and SATA ports.

>> No.65636278

>>65635644
I didn't know even the XEON-W was that much more expensive than just the SK-X. Really intel could have saved a lot of face by just releasing the xeonW renamed as skylake X, performance would have been a tiny bit less, but without the wierd segmentation and socket melting attempts the reception would have been more better even if it did still lose to TR value wise.

Conversely TR could have done with AMD pushing for a more preoffesional branding front as well. When I was looking for a workstation at work every other motherboard had "gaming" or "Xtreme" somewhere stamped on it.

>> No.65636660
File: 69 KB, 1285x399, eh.png [View same] [iqdb] [saucenao] [google] [report]
65636660

what do?

>> No.65636780

>>65631171
>Yes and no. Intel tdp is hard set as the maximum power allowed for the chip.
This is objectively wrong.
Isolated measurements of CPU draw only shows the 8700 and 8700k at 115 watts stock with multi-core-enhancement and so on disabled and a 95w TDP profile.

TDP != power draw, even if Intel didn't straight up lie about their TDPs.

>> No.65637173

>>65619471
>Buying ryzen and a shit GPU
You're next level retarded

>> No.65637261

I have two computers
a 1600 and 580, and
A 7700k and 1060
The amd rig is superior in every way. Games, Linux, compilation.
I just don't understand why people think Intel/ NVIDIA is even with considering

>> No.65637311

>>65637261
I'm not talking max frames either. I'm talking min frames, which is far more important when one setup never drops noticeably and the other leaves some games unplayable. I'm talking trying to use Linux and even the web browser is laggy as fuck

>> No.65637463

>>65637311
>I'm talking trying to use Linux and even the web browser is laggy as fuck

you obviously have other issues then.

>> No.65637657

>>65611781
im planning on upgrading my motherboard and cpu (both are extremely outdated, my i5-2500k still does its job, but overclocks no longer hold) - a guy told me to get good performance out of ryzen you need a lot of ram, and it needs to be fast, i'm sitting on 16gb DDR3 2400mhz, though it only runs at 1333 because >motherboard
is this true?

>> No.65637712

>>65637657
>ryzen you need a lot of ram

no, but you need 3000mhz min, recommend is 3200, you can use 3400+ if you have the money.

>> No.65637722

>>65637712
thanks, i think im going to stick with intel, then
not interested in upgrading ram at all at the moment with how the prices are, same with my gpu, the 1060 will do for a while it seems

>> No.65637919

>>65637722
You can get a Corsair Vengeance LPX 16GB for $189 on amazon. I've been using it with my ryzen vuild for over a year now.

>Ryzen 5 2600x
>any $140 x470 motheboard
>$189 for ram.

>> No.65637942

>>65637311
sound like a gpu problem

>>
Name (leave empty)
Comment (leave empty)
Name
E-mail
Subject
Comment
Password [?]Password used for file deletion.
Captcha
Action