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 ]
>> No.75934464

Wait for zen4

>> No.75934491

>>75934464
>first gen DDR5 and possibly PCIe5

Lol no, new standards are best a year or two after the initial wave.

>> No.75934494

Wait for zen5

>> No.75934536
File: 593 KB, 900x970, 1553572128643.png [View same] [iqdb] [saucenao] [google] [report]
75934536

wow I had a few floating point / integer calculations that I needed to do, and I didn't want to do them on my current CPU because it's 15% slower and has a slightly higher power draw

good thing that these new CPUs are coming out, I'll just wait 3 months and then I'll be able to perform all of these calculations

>> No.75934550

>>75934536
Good call bro

>> No.75934551

>>75934536
Why yes I prefer CPUs to not get faster at all by the year but stay at full price.

Dumb mommyfaggot.

>> No.75934669

>>75934464
the wait is over im jumping ship.
>>75934491
this, better to wait out any initial problems from new shit. rather just get 7nm+ since it's just a more refined version so you'll have a much less chance of having any problems.

>> No.75934699

YES GOY BUY X570 THAT WILL BE ONLY USEFUL FOR ONE GENERATION

X670 USERS WILL GET DDR5, PCIE 5.0 AND 2 GENERATIONS OF 5NM RYZEN

>> No.75934719

>>75934699
>buy $100 motherboard in 2020
>sell it for $50 or repurpose it for a server in 2022 when it's safe to change to DDR5

Wow, I'm broke!

>> No.75934770

>>75934756
This can only be alarming to Intel and their merry band of kikes.

>> No.75934771

>>75934756
>Chinks are supporting AMD
YIKES

>> No.75934799
File: 27 KB, 396x367, 1461755770389.jpg [View same] [iqdb] [saucenao] [google] [report]
75934799

>>75934756
I guess you could say they're ryzen up.

>> No.75934802
File: 1.20 MB, 765x2254, file.png [View same] [iqdb] [saucenao] [google] [report]
75934802

>>75934379
i'm unironically looking forward to Zen2 APU's first. should be very powerful and clock a bit better with 7nm matured af.

>> No.75934849

>>75934819
I figure if im this close to Ryzen 4000, I can try to hold off. But it's showing its age, and might actually be acting up. Noticing lower frames in games that used to be fine a few weeks ago.

>> No.75934879

>>75934861
AVX512 requires a lot of die space, AMD wants to keep their cores small because of development cost and time to market.
5nm will allow them to make a similar sized core while having AVX512

>> No.75934893

>>75934849
Replace thermal paste you dumbass. Or it could be jewish drivers. There's almost no chance your CPU is fried unless you've been running it at 1.4V since 2011

>> No.75934900

Kek, not even Intel's 5.3GHz for a picosecond before it roasts the entire case will save them from this.

>> No.75934965

>>75934917
I'm not too sure how realistic chipset failure is. It never happened and I'm not sure what else could get fucked inside a mobo. I've only ever had ports fail, your PS2 slot probably failed because repeatedly flexing over almost a decade fucked the contacts. You could probably resolder it if you wanted to.
If you end up getting zen 3 get a 8c16t CPU and an x570 board. Should be good for another decade if we're being honest, although zen 4 and zen 5 will both offer reasonable performance uplifts. I'm on zen 2 and waiting for zen 5

>> No.75934983
File: 386 KB, 1000x1000, lisa_su_pepe.png [View same] [iqdb] [saucenao] [google] [report]
75934983

>>75934950
yeah right, thats why you click on this thread and write your cope in here LMAO

>> No.75934999

>>75934965
Are the x570 board better now? I think I remember them coming out and they needed to have their chipset cooled with a little fan like some mid 2000s shit.

>> No.75935000

>>75934979
It takes a lot of years to fix a decade of entrenched "corporate culture" like a 100k+ employee company like Intel has, while Intel is fixing their internal messes and loses talent AMD isn't gonna be sniffing their own ass.

>> No.75935001

>>75934983
dilate

>> No.75935015

>lower memory latencies
doubt

>> No.75935017

>>75935006
no need to be upset friend :)

>> No.75935027

>>75934950
my condolences for intel clowning on you with their hyperthreading

>> No.75935046

>meanwhile Tiger cope
>2.8GHz at 25W TDP for 4 cores on "10nm++"

Rofl

>> No.75935051

>>75934999
I'm not sure how good they are now, but it shouldn't be too bad if you have decent case cooling. I don't think that fan spins all the time unless it gets really hot inside your case. I'd go x570 over b550 because Zen 4 won't be AM4, so your only upgrade path will be higher core count Zen 3 processors and those will need the more robust VRM setup that x570 boards have. You might decide to get a 12c24t or even a 16c32t Zen 3 CPU like 5-6 years down the line and the b550 boards won't be able to handle them.

>> No.75935062
File: 75 KB, 604x453, 1520502958748.jpg [View same] [iqdb] [saucenao] [google] [report]
75935062

>>75935046
>still curbstomped by a 15W 4800U

>> No.75935094

>>75935031
>wasting time on intlel
That's your opinion. I have the best gaming CPU out right now.

>> No.75935095

>>75935087
Bruh nobody's waiting for 2025.

>> No.75935106

>>75935094
but why are you hellbend on telling us that?
buyers remorse is a hell of a drug man

>> No.75935146
File: 29 KB, 753x530, le troll xdd.png [View same] [iqdb] [saucenao] [google] [report]
75935146

>>75935135

>> No.75935147
File: 113 KB, 914x892, 1485079533856.jpg [View same] [iqdb] [saucenao] [google] [report]
75935147

>>75935135
>t.
truly showed us

>> No.75935152

>>75935146
>>75935147
>t. amdrones

>> No.75935155

>>75935152
t. intcel

>> No.75935171

Don't worry bro, Intel is prepared.
With a 14nm backport of a mix of sunny and willow cove with half the cache and features missing, drawing up to 70W per core up to 8 cores!

>> No.75935173

>>75935152
>>75935155
Youre both corporate cocklickers, gratz.
>SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP SHLORP

>> No.75935202

>>75935173
>incel tel aviv headquarters trying to damage control by meeting in the middle
Amd has a product lineup that is superior to intel's at every segment in every way you can think of. Recommending this lineup to people is not at all similar to hard coping by giving legitimately bad advice to people and trying to make intel's housefire look like it's anything but a joke

>> No.75935248

>>75935171
Based Intel, it takes a lot of skill to draw 500W from 200mm^2 of silicon, how does this company do it?

>> No.75935374

>>75935347
He really thinks CPU to memory NANOSECOND latency differences are actually MILLISECOND input and processing latency differences that DPCBitches complain about all the time

>> No.75935671
File: 2.44 MB, 270x396, 10pct body fat?.gif [View same] [iqdb] [saucenao] [google] [report]
75935671

>>75935414
>tfw you buy a new cpu and gpu every generation and don't care about these debates of when its best to upgrade.
>tfw your shit's always ultra snappy

>> No.75936075

>>75934516
How many pluses after 14nm this time?

>> No.75936090
File: 305 KB, 554x400, 1586593838288.png [View same] [iqdb] [saucenao] [google] [report]
75936090

>>75936075
6 MILLION

>> No.75936199
File: 117 KB, 1000x470, t0L4aARlFfrFrsUb.jpg [View same] [iqdb] [saucenao] [google] [report]
75936199

>>75934411
>Hopefully Intel will deliver something that at least can compete with zen2
If intelfags actually think AMD will stand still despite btfoing Intel they're sadly mistaken.

>> No.75936314

>>75934952
>AMD's new iteration with no node has more improvement than Intel's newest uarch on a new node

Top kek

>> No.75936318

>>75936199
>If intelfags actually think AMD will stand still despite btfoing Intel they're sadly mistaken.
the prices will increase. Right now intel CPU's are cheaper than AMD equivalents.

>> No.75936320

>>75935347
>>75935374
Not a meme, skip to 4 min 5 seconds

https://www.youtube.com/watch?v=dsbVSknUK7I#t=4m05s

>> No.75936323

>>75936307
Depends on your budget and what you do with it. 3600 is the best imo

>> No.75936329

>>75935202
and guess what

thats one puzzle piece

a puzzle is made of many pieces.
you dont see the full picture until the puzzle is complete

>> No.75936344
File: 56 KB, 765x513, based.png [View same] [iqdb] [saucenao] [google] [report]
75936344

>>75934756
Wanna bet Zen 3 will shift the desktop market share towards AMD?

At least they still have majority laptop market share...

>> No.75936347

>>75935094
How's that 3950X treating you?

>> No.75936348

>>75936318
The AMD equivalent of Intel's CPUs are also significantly faster, at the same price point they just clobber them.

>> No.75936361

>>75936344
thats probably DIY market only. AMD needs to bring all OEMs in for big bucks

>> No.75936395

>>75936361
Not with Intel MDF lmao.

I love it personally, AMD has a product that's faster, cheaper on both the ecosystem and uses less power but can't get any market penetration because Intel pays off OEMs to not sell those products.

Amazing.

>> No.75936405

>>75936361
No, prebuilts are still included in that chart, because it counts desktop CPUs.

>> No.75936413

>>75934491
>waiting for 4 more years
>when you could simply wait for a full decade
dumb consoomers, never going to make it

>> No.75936426

>>75936323
>what you do with
Gaming essentially and not looking to change parts for as long as possible. I was thinking 3600 as well. Either that or 3300x, but 3600 looks more longer lasting. Thanks

>> No.75936563

>>75935199
>cannot fault AMD for that.

The mindset of die-hard Trump voters who post clever stuff like >>75936090 and think they are smart capitalists because they hold 1 AMD stock certificate.

Lisa is going to squat over that AMD tattoo you got and take a hot steaming price hike on all future products because the Intel model made money by making you an angry gloating AMD fanboy.

tl;dr the MAGA self-own continues

>> No.75936596

>>75936512
I really hope so, Currently I have 5 year old FX-8300 Black Edition, never had issues with it (aside from the awful cooler that you get with it), still works perfectly. I can only wish my next processor is as reliable as this one.

>> No.75936940

>>75934771
The T in TSMC stands for Taiwan

>> No.75936970
File: 100 KB, 1080x1246, lr5kdzqr0ve31.jpg [View same] [iqdb] [saucenao] [google] [report]
75936970

4770K waiting patiently to upgrade to a 4900x.

>> No.75937013
File: 189 KB, 1035x1000, 1563455436921.jpg [View same] [iqdb] [saucenao] [google] [report]
75937013

>>75936563

>> No.75937032

>>75936740
I honestly never dabbled with overclocking, but recently I got quite interested. I bought CM HYPER H411R for my FX-8300 and it works great, even under stress test temperature is not above 40c, we'll see how well it fares with 3600

>> No.75937308

>>75937285
zen4 is slated for a 2022 launch to coincide with consumer ddr5, meaning that zen3 will be quite a long lived generation, lasting 1.25-2 years

>> No.75937321

>>75937285
Buy Zen3 now.
Wait for Zen4 bugs to get ironed out.
Buy Zen5.

>> No.75937366

>>75937342
Um... no? You think that 99% of buyers are the same ones upgrading every generation? There's millions of people still on haslel or older that are upgrading soon.

>> No.75937405

>>75937371
It'll probably only be $200~ and because they can
Quad cores are poverty tier

>> No.75937554

>>75936320
Wait. So he runs CS:GO, records it with 1000fps camera and measures latency that from the recording. The results were in 7ms to 27ms range (averages 11-19). The display he's using apparently is 100Hz so that's at least between 0 and 10ms of latency from the display itself (and probably a couple ms more). Did he do the test at least a couple hundred times?

>> No.75937623

>>75937554
yes, and the result is AMD = lag.

>> No.75937730

>>75936361
OEMs don't give a shit about AMD

>> No.75937738

>>75937711
W8 for 8

>> No.75937762 [DELETED] 

>>75937730
AMD focuses more on servers so it's not like they care about OEMs either. They're a lot more profitable.

>> No.75937765
File: 325 KB, 528x962, 6468915C-6C63-4015-95DB-15CA7CD943D9.jpg [View same] [iqdb] [saucenao] [google] [report]
75937765

Haswell bros it’s finally our time
i5 4690k -> 4800x

>> No.75937775

>>75937371
A lot of people don't really need or want a discrete GPU so long as they can play old games at decent framerates.

>> No.75937784

>>75937762
AMD is trying to focus on servers, most of their revenue is still coming from consumer sales and sales through OEMs for mainstream desktop CPUs and GPUs.

>> No.75937786 [DELETED] 

>>75937765
>4 threads
Obsoleted since 2017

>> No.75937849

>>75937822
AMD is powering TWO exascale super computers now. The first two in the world in fact.
Frontier and El Capitan.

>> No.75937932 [DELETED] 

>>75937881
The world doesn't care about Intel >>75937822 >>75937849

>> No.75937934

>>75937623
Can you point me to where he says how many samples were tested?

>> No.75938042

>>75937730
its not like i already said that and WANT that status to change. back to elementary school anon

>> No.75938149

>>75934952
Intel will never die, they are part of America's national security umbrella.

You really think Spectre and Meltdown are vulnerabilities and not backdoors for alphabet agencies?

>> No.75938176

>>75938158
IBM doesn't manufacture CPUs for consumers.

They make more money than Intel by building custom enterprise stuff.

>> No.75938310

>>75938271
Year of the mature VR market will happen right after the year of the Linux desktop

>> No.75938332
File: 125 KB, 960x878, Intel Wojak 12.jpg [View same] [iqdb] [saucenao] [google] [report]
75938332

>>75934379
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

>> No.75938397 [DELETED] 

>>75938156
Big Navi flagship is already 30% faster than the fastest 2080ti with best binning in existence.

>> No.75938465
File: 174 KB, 2047x1165, IF v2.jpg [View same] [iqdb] [saucenao] [google] [report]
75938465

>>75938377
The IF present in Renoir has some pretty significant tweaks compared to the desktop chips. Its a lot lower power, and has some improved latency characteristics as a result. Zen3 having a unified 8 core CCX will undoubtedly be coupled with major fabric changes as well.
They have the whole "Infinity Architecture" coming to enterprise as well. More than double the bandwidth of PCI-E 4.0, lower latency too. This has pretty big implications for what their data fabric is capable of.

>> No.75938557

>>75938491
>1st gen
Don’t do it

>> No.75938559
File: 834 KB, 609x588, ahegao 23.png [View same] [iqdb] [saucenao] [google] [report]
75938559

>>75938465
>inb4 shared HBM3 system + graphics memory
>inb4 dual socket mobos where one of the sockets is for a GPU die
>inb4 laptop APUs with >20 CU graphics with shared HBM3 memory

>> No.75938607

>>75938520
the right can't meme

>> No.75938645

>>75938397
No, faster than a paper spec 2080ti, which is almost 20% slower than actual 2080tis, which have boost 4.0

>> No.75938717
File: 17 KB, 256x256, 1588238623033.jpg [View same] [iqdb] [saucenao] [google] [report]
75938717

>>75938676
oh, its a ASOOOOOOOOOMer.
thats fine, masks are great tho. i never have to smell you leftists rats ever again. also i'm german, you know, the country with lowest casualites from the chinkflu. but please keep hitting

>> No.75938881
File: 79 KB, 1158x1007, icelel.png [View same] [iqdb] [saucenao] [google] [report]
75938881

>>75938861
heres another one for your collection

>> No.75938911

>>75934379
My R5 3600 will be enough for next 5 years minimum
I would've still used FX8370 but mobo died and getting new one is super expensive

>> No.75938925

>>75938465
>1 performance
>2 performance

>> No.75939005

>>75938708
How can you write so many words and say nothing?

>> No.75939215 [DELETED] 

>>75939018
Seems legit

>> No.75939369

>>75939018
Thats a zen 2 based apu...

>> No.75939524

>>75939224
I'll jump on just because DDR5 has always on ECC.

>> No.75939551

>>75939224
The DIMMs in prouction right now are 5200mt/s. LPDDR5 has been in production for a long time and is shipping in devices right this second.
This fear mongering is nothing but poor NEETs trying to rationalizes not being able to afford new high end hardware. Jet a job, loser.

>> No.75939581

>>75936361
>AMD needs to bring all OEMs in for big bucks
They don't care about OEMs. They got both of the consoles

>> No.75939639

>>75939574
i'm not up to date on that. just caught a thread here last week with anons claiming just that while everyone called them morons. iirc the anons posting HWinfo finally admitted they used some VERY light workloads with low voltage/high clocks and admitted they were wrong.
where do you got your info tho? did you post in one of there threads too? i'm very sceptical about such thing, you know..

>> No.75939907

>>75937765
>4800x
why?

>> No.75939977
File: 21 KB, 666x366, 1589641491501.png [View same] [iqdb] [saucenao] [google] [report]
75939977

>>75938727
Looks like this image triggers the streetshitting inteljeet janitor

>> No.75940130

>>75938096
By the time AM5 comes out AMD will have completely dropped the pretense that they aren't adopting Intel style one chipset one cpu generation.

>> No.75940152

>>75939551
Nobody cares about bandwidth, DIMMs with tight timings won't be here for at least 2-3 years.

>> No.75940174

HOLY FUCKING BASED AMD
POZZTEL DOA HOUSEFIRES BTFO

https://www.notebookcheck.net/AMD-s-Ryzen-4000-desktop-CPUs-rumored-to-have-higher-than-15-17-IPC-gains-over-current-Zen-2-models-could-utterly-destroy-Intel-s-Comet-Lake-S-lineup.465643.0.html

>> No.75940257

>>75937371
It looks like a good idea for casual gaming and workstations that don't need to do render duty but still need to run the latest Autodesk/Adobe shitware
Still, it would be a first, they never made a powerful APU despite the HSA shilling

>> No.75940303

>>75938778
Show me R7 4800U in shop

>> No.75940380

>>75940341
We still won't reach 100GB/s BW with dual channel memory, and a mere IGPu with 20 RDNA CUs is enough to saturate that, which would be piss to make on 7nm, much less on 5nm

>> No.75940404

>>75940380
>We still won't reach 100GB/s BW with dual channel memory,
Dual channel (pseudo quad) LPDDR5 6400 is 102.4GB/s brainlet.

>> No.75940517

>>75939977
Wait, are mods starting to finally crack down on the AMD marketing? Source?

>> No.75940523

>>75940517
>a picture from an Anandtech article
>AMD marketing
pick one

>> No.75940588

>>75940574
See >>75940517 again since you're too big, fat and retarded to read.

>> No.75940976

>>75940930
>day
year
First on roadmaps for 2015, and finally showing up for desktop in 2022 isn't too bad of a spread!

>> No.75941102

>>75938736
based

>> No.75941155

>>75934491
What do you think is going to change in a year hardware wise?

Here’s a hint:
Nothing. It will all be software and firmware compatibility issues that will get ironed out based on level of seriousness.

>> No.75941180

>>75941081
And that's not even refined 7nm since it's basically first gen mobile 7nm, expect 30% higher score from IPC + clock improvements

>> No.75941245

>>75941210
What are you doing that your 1600X feels dated in any way?

>> No.75941263

>>75934999
>they needed to have their chipset cooled with a little fan like some mid 2000s shit
That's still the case, though it's possible to do what some of us did mid-2000: you can replace the fan with an aftermarket cooler. Except.. back in the mid-2000s the chipset fans were typically all 40x40mm with holes in opposite corners making it easy to offer big standardized chipset heatsinks. I got one from Zahlman when the fan on mine went bad. That's not the case with x570 boards, there's no standard off the shelf heatsink fitting all the boards. The new very delayed 550 boards won't have any chipset fan so those could be an option.

>> No.75941405

>>75941368
kek
https://www.youtube.com/watch?v=_1lqaoZhGHU
https://www.youtube.com/watch?v=VkDEUeqQUXY
https://www.youtube.com/watch?v=_ki5kpwJ8J0

>> No.75941431

>>75941385
The creator of Linux IS the beard now.

>> No.75941462

>>75941328
AVX512 is a housefire and a meme, AMD doesn't need it to BTFO Intel.

>> No.75941469

>>75934379
I don't think it's really worth it to upgrade from my 3700x.

>> No.75941561

>>75941536
well in the last 20 years they've been on top for about 15

>> No.75941562

>>75941551
Next year is rocket lake on 14nm++++++++

>> No.75941602
File: 38 KB, 509x500, 299.jpg [View same] [iqdb] [saucenao] [google] [report]
75941602

>>75941538
More like the Pewdiepie of tech world.

>> No.75941618

>>75941481
How much of an upgrade was the 2000 series to the 3000 series?

>> No.75941629

>>75941589
Let him enjoy his final moments of happiness before reality sets in

>> No.75941743
File: 8 KB, 257x196, download.png [View same] [iqdb] [saucenao] [google] [report]
75941743

https://www.notebookcheck.net/AMD-Ryzen-7-4700G-final-version-spotted-4-4-GHz-boost-clock-8-cores-SMT-and-an-8-CU-GPU-set-to-do-battle-with-the-Intel-Core-i7-10700.465693.0.html

>> No.75941745

>>75941701
4690 is locked (although you can force it to use max boost clock on all cores with some motherboards), 4690K is not.

>> No.75942126

>>75942114
It doesn't seem likely in the foreseeable future.

>> No.75942161

>>75941764
Aw, is the incel upset?

>> No.75942267

>>75942143
Good performance any day now bro

>> No.75942297

Can't wait for Intel 7nm to destroy AMD in 2025

>> No.75942343

>>75940535
can't wait to cheer for the underdog scamps at intel

>> No.75942368

>>75942355
Yeah bro, intel must be so good everyone is buying their products, man intel is selling like hotcakes, how could AMD ever recover

>> No.75942413

>>75942396
Oh wow, a time traveler from 2012, neat

>> No.75942414

>>75942368
AMD is trash bro

>> No.75942428

>>75942414
Whatever helps you cope at night bro

>> No.75942438

Intel 3nm will destroy AMD in 2040

>> No.75942463

>>75941066
>slower than x5650 from 10(!) years ago

>> No.75942485

>>75942471
Enjoy your 3.9 GHz bro

>> No.75942500

>>75942351
That's because Ryzen is infinitely better therefore Intel would need an eternity to surpass it.

>> No.75942507

I just wanted DDR5 with an AMD APU man...

>> No.75942535

>>75942521
That's right, they wouldn't.

>> No.75942584
File: 343 KB, 807x867, 1534595224911.png [View same] [iqdb] [saucenao] [google] [report]
75942584

>>75936344
THIS CAN'T BE HAPPENING

>> No.75942613

>>75942577
Doesn't Ryzen already have a small ARM core?

>> No.75942678

>>75942613
There is the Cortex A5 core which powers the TrustZone aka PSP security coprocessor.
AMD has its own customized high performance ARMv8 core arch called K12 though. I'm sure they could pull out all the stops and implement everything they've developed for the Zen family in an ARM core if they needed.

>> No.75942712

>>75942463
7700k loses to 4700u @ 2GHz brah

>> No.75942723

>>75934379
>>2x available cache per core
Digits decide how will AMD will call it
>1 Game cache stays
>2 Rizen cache
>3 Turbo cache
>4 Full Power System cache
>5 Bullet cache
>6 True Velocity boost cache
>7 RGB cache
>8 Deep cache
>9 Cache tracing
>0 AI cache

>> No.75942731

>>75942723
Gamecache sounds too based so it most likely won't change.

>> No.75942742

>>75942723
We already have AI branch predicting so thats out.

>> No.75942799

>>75940424
The purpose of an integrated graphics chip is to be good enough. It’s never going to be the best, it just has too many things holding it back. But it does the job for a lot of stuff.

Don’t delude yourself that significantly faster memory and improved architecture doesn’t raise the bar for good enough quite a bit. You could realistically play all of the video games on the market with one of these bad boys. Not that you can’t on current gen APUs, but you’ll get to compromise on resolution, FPS and graphical quality much less.

>> No.75942821

>>75942799
I just want APUs to make sub-$200 discrete GPUs obsolete, and they sure as fuck could if AMD tries.

>> No.75942902

>>75936563
Cope

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