[ 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: 142 KB, 1072x650, leaked picture Intlel HQ.jpg [View same] [iqdb] [saucenao] [google] [report]
66240676 No.66240676 [Reply] [Original] [archived.moe] [rbt]

>32 core 64 Thread Threadrippers
>3GHz base 3.4Ghz Boost
>32 Threadrippers are 250W (as compared to Intlels 28 core housefire)

>7nm Epyc CPU's
>new Vega 56 "Nano card"
>New vega GPU at 7nm with 32GB of HBM2
>1.35x Performance increase
>2x more energy efficient

Meanwhile Nvidia's CEO doesn't want to speak about new card releases and Intel is trying to make it's move in the nuclear power industry.

is this it bros?

>> No.66240847
File: 77 KB, 501x462, t-threadripper 2 will be DOA r-right.png [View same] [iqdb] [saucenao] [google] [report]

>> No.66240867

Nvidia still dominates scientific computing/deep learning

>> No.66240924

>32 core 64 Thread Threadrippers
Which will be beat by the lower core higher clocked intel equivalent.
>>1.35x Performance increase
Which will be beat by the new nvidia cards coming out in a few months. The ceo was making a joke, when he said in a long time.

>> No.66240980

IF what we heard about rcom is true and amd managed to create another HIP tool that directly translates cuda to its own without modifications

then its over for nvidia they simply wont be able to keep up with the superscalar desing at any level

>> No.66241439

where's your source

>> No.66241568
File: 66 KB, 490x554, b044444ac6.jpg [View same] [iqdb] [saucenao] [google] [report]

>streetshitter thinks nvidia has competition

i bet you think vega was a success

>> No.66241683

nice filename :)

>> No.66241719
File: 82 KB, 1464x555, market share.png [View same] [iqdb] [saucenao] [google] [report]

>i bet you think vega was a success

>> No.66241752

It's happening
Hopefully my 1080 g1 lasts til 2020 by then mcm 7nm will be mature enough and I can comfortably upgrade to 4k 250hz

>> No.66241767

That's a really dumb take. RTG remains terrible and the Vega nano is still on shitty 14nm by gloflo. Until amd gets real yields on a non-garbage fab don't expect them to remotely compete with Nvidia. CPU is a completely different story, and they're poised to lose big time again when Intel pulls bullshit lock-in agreements in the hpc space to prevent the superior amd offering getting traction. It's all happened before, it'll all happen again.

>> No.66241779

Nvidia is safe, I doubt AMD will be capable to salvage the garbage that is vega. Intel though is in full PANIC mode.

>> No.66241795

What source do you need if only one manufacturer makes CUDA GPUs.

>> No.66241802

Nvidia still dominates overfunded institutions who are stupid enough to invest 100k+ into a cludgy gaming platform rebadged as a research tool.

>> No.66241807

>What source do you need if only one manufacturer supports a niche, proprietary API

What did he mean by this?

>> No.66241819

Please read again what we are talking about.

>> No.66241827
File: 427 KB, 978x478, sorry.gif [View same] [iqdb] [saucenao] [google] [report]

>Intel will calmly execute their master plan to defeat Ryzen
>rushes an overclocked Xeon pulling 1500W on stage to try and grab a headline
>gets BTFO the next day by Kikeripper
>immediately demands the overclocked Xeons back from Asus and Gigabyte
Truly the image of a calm, collected company with a long-term plan for dominance.

>> No.66241837

>It's another AMDrone thread from /pol/ack crossboarders who think their favorite multi billion dollar company isn't jew tier due to mental gymnastics episode.
Protip they are there for profit so you all should stop giving them free marketing, at least ask for something in return you dumb NEETs

>> No.66241850

You wrote precisely why Nvidia still dominates the scientific field. Why act so obtuse when you know yourself what exactly is happening.

>> No.66241968

But we are getting something in return. Awesome CPUs at a fair price. And your tears. That's worth a lot, anon.

>> No.66241981

>3GHz base 3.4Ghz Boost
engineering sample

>> No.66242031

Thats a good one marketeer kun.
Dont forget to buy the advertiser rights on 4chan, send an email to hiroshimoot.

>> No.66242078
File: 37 KB, 702x317, Selection_022.png [View same] [iqdb] [saucenao] [google] [report]

But I know a good upgrade when I see one. Come August, Threadripper mine.

>> No.66242094

Please explain what exactly is happening.

>Nvidia still dominates the scientific field.

oh, I guess that's why Intel put FPGAs in their newest niche product.

I really have no idea what you're trying to say. Not everyone thinks the same way you do.

>> No.66242106

Wait they demanded it back? Source I haven't seen this.
T. Intel to Zen cuck

>> No.66242121

Thanks Anon

>> No.66242136

Scientists use CUDA because they can't be bothered to learn anything else(these are the people that stop use Fortran ffs). Only one company makes CUDA GPUs. CUDA is proprietary. Therefore Nvidia dominates the scientific field.

>> No.66242154

Give us Vega40 7nm 2.2ghz AMD

>> No.66242174

Intel still rules the third world countries in cpus market. A few people in these countries know about amd.
So even amd make the super duper micro circuit, intel still have better marketing for poor countries. So... intel still relevant.
>t. Sopa de Primata

>> No.66242222
File: 294 KB, 1565x913, Darwin4athiests[1].jpg [View same] [iqdb] [saucenao] [google] [report]

Well you got me there. You have sources and everything.

>> No.66242741

>Which will be beat by the lower core higher clocked intel equivalent.
Only on liquid nitrogen LUL

>> No.66242813
File: 74 KB, 1536x869, 43awr5p06d211.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.66242827

There's nothing wrong with FORTRAN, brainlet

>> No.66242851


>> No.66242895

I like this timeline

>> No.66242909
File: 51 KB, 711x457, dRINhkW.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.66242935

>placed on a trolley cart
Truly cascade lake.

>> No.66242952
File: 436 KB, 823x520, 1528257832783.png [View same] [iqdb] [saucenao] [google] [report]


>> No.66243001

Nice stock cooler there, Intel. Nothing like those shitty ones that AMD packages for their cpus.

>> No.66243051
File: 110 KB, 1200x800, threadripper_cooler_vladsavov_computex18.jpg [View same] [iqdb] [saucenao] [google] [report]

This is what AMD demoed with.

>> No.66243064
File: 84 KB, 653x726, Everything's fine.png [View same] [iqdb] [saucenao] [google] [report]


>> No.66243098

>>32 core 64 Thread Threadrippers
>half of dies does not have direct access to RAM

>> No.66243104
File: 468 KB, 882x758, 1526745411872.png [View same] [iqdb] [saucenao] [google] [report]

Someone should add one extra layer to this

>> No.66243111

You want to game on a 32 core machine?

>> No.66243119

So intel doesn't have legit PCIe because half of the lanes are not direct?


>> No.66243124

How big a deal is this for anything that isn't memory intensive?

>> No.66243133

Nvidia has proven they do think about amd when they liked epyc fag

>> No.66243150

Get epyc if your are concerned.

>> No.66243179

I don't know if I should be concerned. Nothing I do is specifically memory intensive, it just likes moar coars.

Worst case scenario you can just limit tasks to the 16 cores with direct memory access.

>> No.66243191

the cards arent the problem the software is anon
amd doesnt have anything to compete with cuda and gameworks

>> No.66243286

I'll put it like this. If you make money with your computer and you're worried TR might not be enough then go for Epyc. I mean, if we're talking money here then what's the difference an extra grand for the CPU make? In my case, I need the cores but the memory bandwidth isn't that important and even though I make about 7 grand a week on my computer, I'm going 32 core TR in August. If I needed the memory bandwidth though, I wouldn't hesitate to just go Epyc. I'd go Epyc now if the single process speed was competitive with TR.

>> No.66243320

Like I said, as far as I know it shouldn't be an issue. I'm just not that educated on the impact this is going to have to latency.

And I'm holding out until 7nm for an upgrade. I'd love to see a 48c or even 64c 3950x.

>> No.66243348

>7% increase over a year
>still 65% nVidia

>> No.66243440

>Which will be beat by the lower core higher clocked intel equivalent.

1) The stock clock is listed in the photos as 2.7GHz which is lower than AMD's
2) They have a phase change cooler needed to overclock it to 5GHz which isn't that impressive. Threadripper can do the same with enough cooling too.

>> No.66243442
File: 240 KB, 765x586, 60cd5ad54b906bc9aa7d6c8ed6d71226b1eaf828c8b683b5194f0f42eac1ec6a.png [View same] [iqdb] [saucenao] [google] [report]

How incredibly mismanaged do you have to be to fuck up as badly as intel did?
It just doesn't end.

>> No.66243469

>all that shit and Poozen still can't game and has shit single core perf

>> No.66243500

IAEA classified it as a potential threat and they had to remove it?

>> No.66243501

>32 cores for MUH GAYMEN

>> No.66243528

>New vega GPU at 7nm with 32GB of HBM2

Are these official claims!??


>> No.66243536
File: 161 KB, 1079x784, 1525554090644.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.66243546

>Intel is trying to make it's move in the nuclear power industry.


>> No.66243550

Intel was demoed with an industrial cooler that's massive enough to have its output measured in horsepower rather than watts.

>> No.66243584

Yes and it's already being sampled to partners. It's business part and it's still an old Vega shrink.
>only games are latency-sensitive
It's the first CPU to have dies without direct memory access. Remeber what GTX970 was BTW.
I really do not know. One could probably simulate that on current Threadripper by using other die's memory only.
Non-overclockable, slower RAM support, different motherboards.

I wish that AMD still made a two die TR though.

>> No.66243601

What was the point of this?

>> No.66243606

>Gets molested by original Ryzen
>Gets penetrated by Threadripper
>Gets slam dunked by Meltdown and Spectre
>Gets dunked on EVEN HARDER by Spectre
>Gets brutalized by Threadripper 2

>> No.66243609

Yes, but they are for workstation and server use. Consumer 7nm announcements later.

>> No.66243643

Something I've always wanted to know, why doesn't the speed of processors increase with new models? A 20 year old processor is still 3Ghz.

As I understand it, that is how many instructions it handles, right? So why doesn't it improve?

>> No.66243651

That's the 1700w cooler Intel used for the demo.

>> No.66243669
File: 24 KB, 881x188, Screenshot_20180606_163636.png [View same] [iqdb] [saucenao] [google] [report]

They weren't.

>> No.66243671
File: 107 KB, 964x960, Amada.jpg [View same] [iqdb] [saucenao] [google] [report]

35% of the market is a healthy share, especially for a company where resources are stretched much thinner than their competition.

But that wasn't the subject anyway - it was whether Vega was a success, and AMD's market share jumped 6% in the quarter following Vega's launch.

>> No.66243673

Yes but why? Just to show what their CPUs can do at extreme conditions?

>> No.66243698

>another trap mascot

>> No.66243704

I see, so they do increase?

>> No.66243736

More recently we've been coming up against limits that we can't quite break, at least not until quantum computers become viable.
So the second best course of action is to spread out the workload and do as much as we can at the same time.

>> No.66243746

Because otherwise this housefire will throttle to 10% speed.

>> No.66243752

Thanks anon.

>> No.66243755

>jumped 6%
That's what I call a success!

>> No.66243761

>Need a dedicated 15A circuit for the CPU alone and another 15A circuit to power the cooler and accessories with little headroom left over

You can't make this shit up

>> No.66243769
File: 996 KB, 1166x925, RipperZ.png [View same] [iqdb] [saucenao] [google] [report]


>> No.66243773



>> No.66243781

Quantum computers are entirely different beast, they can't even do "normal" computations like normal computer do - not an upgrade over transistor based technology.

We can however still go back to vacuum tubes - litographically print them in silicon and copper like normal microtransistors. They can be made smaller and can operate at much greater frequencies.

>> No.66243803
File: 277 KB, 1616x2048, Amada!!!.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.66243806

Also because the gate has nothing in it (literally vacuum) it doesn't heat up, the only heat dissipation would be from current going through copper traces - that would be several orders of magnitude less heat output than semiconductor based CPUs.

>> No.66243807

None of the dies have direct access to ~all~ the ram in threadripper 1. It's not that big of a deal.

>> No.66243813

>As I understand it, that is how many instructions it handles, right? So why doesn't it improve?
Frequency of CPU is frequency of internals and also the maximum possible instruction execution frequency. Modern CPUs can execute more instructions per cycle at same frequency.
Memory latency reduces instruction throughput.

Only RISC processors can execute one instruction per cycle and they do not have that high frequencies.

>> No.66243836

It not that big deal because OS kernel maps virtual memory onto physical memory accordingly among other reasons.

>> No.66243843

>he needs a 1200W PSU for his cooler
>he needs active cooling for his fucking cooler because it itself is a fucking housefire
I wonder if the two PSUs needs their own water cooling


>> No.66243856


>> No.66243857

If Vega 64 and 56 are anything to go by, AMD is dead when it comes to GPUs, especially high end GPUs

>> No.66243858

no one was supposed to have noticed that. we only know about it because anandtech managed to get a short hands-on with the system. it litelarily is just an overclocked $9k server chip

>> No.66243874

Honestly it's not quite as bizarre as people think, the first 1ghz PCs had fridges built into the case, but obviously this set up is not practical at all

>> No.66243877

>Yes but why? Just to show what their CPUs can do at extreme conditions?

Because intel has nothing, this was an attending grab hoping to steal thunder from amd.

>> No.66243887

Because dennard scaling is over. Power dissipation problems goes up the roof over 5 Ghz even with the Intel's 14nm


>> No.66243896
File: 87 KB, 645x773, 1525082832211.png [View same] [iqdb] [saucenao] [google] [report]


>> No.66243913
File: 971 KB, 800x450, 1528208645545.png [View same] [iqdb] [saucenao] [google] [report]

Nice picture, Jordanian Peterson cock sucker

>> No.66243940

amd is moving vega to 7nm, which is cutting power cost significantly, while increasing power by 35%
even if everything stayed exactly the same and its just a 35% increase in power, that puts it at 1080 ti territory.

if this becomes a 300$ part, its very hard to argue with the performance, if it stays a 400$ one, its nvidia's pricing which will determine how good a deal amd is.

>> No.66243943
File: 41 KB, 768x539, Muh Power density.png [View same] [iqdb] [saucenao] [google] [report]

Power=2π * frequency * voltage * gate capacitance
voltage ∝ frequency

Increasing frequency increase voltage too leading to quadratic growth
Increasing cores just linearly increases the number of transistors and their capacitance

>> No.66243951
File: 196 KB, 335x455, homer.png [View same] [iqdb] [saucenao] [google] [report]

The reason they snatched the CPUs back from Asus and Gigabyte is because they let the cat out of the bag about the chiller and the overclock. Intel were trying to present it as running at stock and without exotic cooling. I guess they forgot to tell their partners that they weren't supposed to take the system apart and show it to everybody.

>> No.66243967

they wont on "10nm" compared to 7nm amd

>> No.66243968

It did, but only to the dumb normalfags(and that buttmad French shill ex-engineer on Twitter) who won't even buy this, anyone with a speck of silicon knowledge would notice the red flag as soon as they saw the clocks.

Pure publicity stunt.
I fucking wonder how half these 'tech sites' didn't notice until a day latter..

>> No.66243969

48+ core Threadripper 3 though.

>> No.66243993

More likely to be 64 cores.

>> No.66244003
File: 484 KB, 574x516, buy it.png [View same] [iqdb] [saucenao] [google] [report]

>A 20 year old processor is still 3Ghz.

>> No.66244020

I put 48+ cores to be safe. I'm expecting they won't need to release that with Zen2 but they will do it on Zen3 just like how they didn't go 32 cores with TR1.

>> No.66244060

>770mm2 die
>1000/7.7 = ~130W/cm2
>Literally as power dense a nuclear reactor

lmao OP's image checks out.

It seems more likely to be 48, if the rumors of a 3 CCX die for 7nm are true.

>> No.66244063

>1700W / 484mm^2 die size (according to anandtech) = 350W/cm^2

Holy shit, it literally IS hotter than a nuclear reactor

>> No.66244305


>> No.66244356

Why would you want even more memory-starved cores? I honestly doubt TR will go above 32c64t on TR4. Any more than what they have now and they WILL need direct access to the memory to not have an enormous amount of laggy cores.

>> No.66244408
File: 1.28 MB, 1452x892, 1528257047833.png [View same] [iqdb] [saucenao] [google] [report]


>> No.66245021

Pajeet pls

>> No.66245058

>4 updates
>have to wait for 1 to download completely before you can download the next
Fuck Intel

>> No.66245300

How much watts per mm2 does a nuclear reactor actually generate?

>> No.66245566

Intel side, kike are literally burning themselves to death, Nvidia side, nah those fuckers dominate AI and Gamer markets alright, hell even Mobile segment all i see is Gamer laptops with Nvidia/Intel hardware.

>> No.66245678

nope. HIP won't be able to compile CUDA into OpenCL and AMD GPUs need better frameworks for BLAS/LAPACK - basic linear algebra routines before anything else can go down. the frameworks in DL/ML and OpenCL is rapidly being integrated into HPC stuff like trilinos which are important in HPC projects my research group does but RTG resources are stretched very thin so it's moving much slower than it should. if the OpenCL compiler gets fixed (it likely won't and we'll just transition to using compute on vulkan) and the linalg libraries get upgraded for AMD most of the work is done for getting AMD GPUs to run just as well as nvidia gpus for HPC.

>> No.66246034
File: 149 KB, 800x505, o0800050512308060428.png [View same] [iqdb] [saucenao] [google] [report]

Modern 4300MW (thermal power) EPR reactors have 241 fuel rod assemblies that each contain 289 individual 4,2m long 1cm thick fuel rods. If we assume that all that heat is divided equaly onto the surface area of all those rods and that I didn't fuck up the math then 4300MW / ( 289 x 241 x 4200mm x 31.4mm ) = 0.47W / mm2
it would seems that this new CPU is producing more than 4 times the heat per mm2 than what this particular nuclear reactor type does.

(sorry for formating of the math, 4chan thinks it's spam)

>> No.66246086

its already doing that but it needs some minor coding alongside..

>> No.66246115

Well first of all, congrads on actually doing the search and math.
>Intel beat nuclear reactor
Was Intel motto, when they calculated thermals of CPU, "to Infinity and beyond"?

>> No.66246127

I'm pretty sure the 4300 MW is the electrical output not the heat transferred to the coolant.

>> No.66246198

Nah that would be too much.
There are 12 NPP in the World that make more electricity than 4300MW, and each of them use at least 3-4 reactors

>> No.66246202

I was using the numbers for Olkiluoto 3. 4300MW thermal, 1600MW electric.

>> No.66246267

I'm fucking crying

Where's 31.4mm coming from?

>> No.66246297


>> No.66246322

pretty sure Nvidia is going to move into AI big time. They have some of the best (if not the best) neural net hardware in the business.

>> No.66246376

>Vega with moar gigs
This is absolutely sad. The 1080 Ti is years old now, Volta is just around the corner and AMD is still pushing this shit architectural housefire.

>> No.66246381
File: 207 KB, 485x362, that feel when jury duty.png [View same] [iqdb] [saucenao] [google] [report]

>is going to

>> No.66246410

Everyone knows current """AI""" is a meme, they are just enjoying the ship while it lasts.

>> No.66246491

> Intel
> 32 core 64 thread ayymd cpu
Whoever dominates the sub $400 segment wins. Normies (which are 95-98% of the population) are going to look at benchmarks in youtube for gaymes and going to buy intel because of 10-20 fps more. Also prebuilts which mostly comes with intel.

> nvidia
Top kek. Nvidia doesn't even try because no competition.

>> No.66246558

wanna compare matrix cores of nvidia and google?

nvidia was never good on anything that involves parallerism anon they dropped their superscalar desing for a reason

>> No.66246966

>28 cores
>32 phase vrm
>4 x 8pins
>cooling with insulated tube presumably for sub zero solution
who thought this was a good idea?

>> No.66246973

>vrm fans

>> No.66246985

evem intel known shills are jumping ship

>it's not even a new chip, it's an old xeon platinum

>> No.66247024

>we don't even know if it's gonna become a real product

>> No.66247049

Not for long

>> No.66247075
File: 84 KB, 1590x888, ZH4IpBo.jpg [View same] [iqdb] [saucenao] [google] [report]


AMD isn't even done yet. Their final form is yet to be announced:

>> No.66247076
File: 42 KB, 712x534, aHR0cDovL21lZGlhLmJlc3RvZm1pY3JvLmNvbS82LzgvNzc2NTI4L29yaWdpbmFsL0ltYWdlMS5wbmc=.png [View same] [iqdb] [saucenao] [google] [report]


>> No.66247088

Everyone with half a brain knows that intel doesn't have anything new other than the 10nm catastrophe. As AMD gets more money, the paid shills will leave intel's side and join them.

>> No.66247093

>The impact of leapfrogging Intel and using 7nm is several-fold. First, Rome will have up to 64 cores and 128 threads in a single socket. The new CPUs will be socket compatible with the current SP3 socket motherboards with a small caveat. At STH, we expect Rome to adopt PCIe Gen 4 so motherboards will have to support the higher signaling rates to achieve PCIe Gen 4. We also expect the next generation to have greatly improved Infinity Fabric, an area that the first generation product has room to improve upon.

>> No.66247105

Where is the 7601 EPYC?

>> No.66247128
File: 17 KB, 260x194, 5 stage cascading phase change.jpg [View same] [iqdb] [saucenao] [google] [report]

sure kid.
this will be your "aio" cooler.

>> No.66247144
File: 21 KB, 480x231, DRJBQdfX0AEr7gx.jpg [View same] [iqdb] [saucenao] [google] [report]

>it's not even cascade lake
>they're still skylake chips quite literally extreme overclocked.

>> No.66247162
File: 2.07 MB, 4423x3317, 2018-06-06 08.54.08.jpg [View same] [iqdb] [saucenao] [google] [report]

Using 20% of the power at 70% the performance.

>> No.66247165
File: 91 KB, 700x565, IMG_3668.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.66247172

Laffd and saved

>> No.66247208

>amd still cant reach 5ghz
>2x energy eff with 35% more gpu perf means that amd @ 7nm will be equal to current nvidia @ 12

catching up by process improvements but still not at the top. amd still doesn't have any CUDA meaning its useless for productivy.

the only justification for buying a new amd product is if the only thing you do is compaling linux.

>> No.66247215

just imagine the lengths intel had to go because of their 10nm failure

find a highly binned xeon
order asus to create a monstrosity that will never be seen again
order a 2 hp chiller
run the benchmark on a different setup than its showcased
and all of that just to make a cinebench number that probably no one will see

>> No.66247244

>amd still doesn't have a feature tech that is patented by its creator and made sure by all means that amd can never have it since it would mean the end for their monopoly
No shit, sherlock.

>> No.66247259

For the genuine retards. That will never be a true product for the consumers, it's just a stunt to please shareholders. It could top 10k in Cinebench and would still be disregarded. You would need fire coverage, a way to dissipate 1600W of heat, and a house wired in such a way it won't implode the entire grid when you turn the thing on.

>> No.66247262

All this being said, I imagine the 28c with a 1-2 core 5Ghz turbo is what they anticipate ending up at with 10nm. For now though, it's blatantly obvious it won't be here anytime soon and that we saw a Skylake-X chip experiencing the thermal equivalent of atmospheric reentry.

>> No.66247266

>find a highly binned xeon
>intel had to ask NSA to give back their diamond binned chip for the showcase
>they told asus to give it back with a 20 minute notice because the CIA was after some terrorists and needed the chip

>> No.66247283

But they claimed 2x power efficiency and 35% performance boost over previous Vega

>> No.66247286

But as core counts are finally skyrocketing on both sides of the shilling fence, apps and OSs - especially high-perf - should get more savvy about NUMA.

>> No.66247292

Why would they, they spend a lot of money to develop the best possible GPUGPU programming model only to share it with their competition. Makes no sense. Where is AMDs alterantive? OpenCL? Sorry im not autistic enough for that shit.

>> No.66247294


>> No.66247302

Your point was that amd doesn't have cuda. You are autistic still. The fault lies with you.

>> No.66247305

>Volta is just around the corner
Who the fuck cares? That's just hugeass die with nothing to contribute on the consumer side. They might release ampere or some shit, that is what you gaymen faggots are looking for.

>> No.66247331

7nm won't magically make Vega good

>> No.66247376

its a compute monster anon amd tried to push for games to be more multithreaded but having to fight gameworks and the money of nvidia didnt really paid off

but vulkan is slowly being adopted and this is the only win amd has seen having its own api as a standar means that when they perfect the culling shit is going to hit the floor

>> No.66247397

7nm offers
40% more perfomance on the same power
or 60% less power on the same perf

usually they go for a middle solution..
i will be suprised if amd doesnt get a 1.8ghz with a low enough tdp

>> No.66247416
File: 223 KB, 379x367, 1528134724 .png [View same] [iqdb] [saucenao] [google] [report]

I just logged in, what the fuck is this.

>> No.66247428

the new bundle intel is making for their hedt market

>> No.66247456

It will most definitely beat the 1080Ti before Nvidia responds with Ampere or a feature removed Titan XxX for the consumer.

>> No.66247466
File: 1.97 MB, 615x413, YaU9Hrf.gif [View same] [iqdb] [saucenao] [google] [report]

>atmospheric reentry

>> No.66247472

The new intel stock cooler.

>> No.66247484

>giga joules
it's a brave new world of new units used for cpus

>> No.66247497

Please note that not all wall sockets can handle the combined 2500W+ load of that setup

>> No.66247518

Now I just want to see Intel get sued when some guy starts a rendering job overnight and burns his house down.

>> No.66247528

The absolute egocentrism of fairytale believers condensed into a single two panel image.

>> No.66247542

im more concerned that this setup can cause harmonics on the electrical grid of their neighborhood

>> No.66247571

lol, the last time I turned on the water heater and microwave at the same time, the fuse blew.

>> No.66247740


The only "limit" has been the lack of competition. Jewtel will keep selling you the same shit for decades over and over again with a new name if they can get away with it. Notice how suddenly Intel can pull a 5Ghz chip our of their ass? There is no real limiting factor other than Intel's complacency and lack of motivation or will to serve the customer.

>> No.66248090

literally the netburst of the era

>> No.66248469

For your reference, a dual-7601 system caps out at around 7252:


Note, this is at 2.7Ghz, so a little over half the clock speed the Intel 28-core CPU was running at. So, the 7601 clock-for-clock is actually almost on par with the 28-core Intel chip, which makes me hopeful for the TR2 32-core chip.
If the scaling holds (which, with Zen, it almost always has), the TR2 32-core chip under water should be able to hit in the 6-7k mark and 4.2-4.3GHz. The 2700x draws ~110-130w at those speeds, so if we assume it's about equal to 4 2700x's under a single heat spreader, it'll be coolable with the LiqTech 360 (500w TDP) for day-to-day use at 4.3GHz. Hell, at 4GHz the 2700x consumes ~75w, the TR2 chip could hit as low as 300W peak consumption at 32x the highest-end Ryzen 1000-series performance tier. That's hilariously overpowered.
tl;dr: the Threadripper 2 32-core CPU will be able to hit 90% of the 28-core Intel chip @ 5GHz's performance for day-to-day usage without issue.

>> No.66248528

>intel/nvidia bros
What does that even mean you nigger? Intel is the one getting their shit pushed.
nvidia is in literal monopoly mode doing fuckall and holding their new generation GPU so they can milk everyone more due to AMD taking ages to release something new.
And let's not even start on the CUDA monopoly part in the enterprise and workstation segments.
Vega frontiers edition for example was a huge meme that had to be rebranded as blockchain pioneers since only miners bought it.
OpenCL is literally dead right now and even the homosexuals at apple are dropping it.
It's over dude, CUDA won the monopoly of GPGPU, the best thing AMD can do is get into some sort of deal with nvidia to license CUDA and make their own GPUs who can run that shit.

>> No.66248571

Nvidia doesn't seem to be stupid, but they recently started to go down on the same path that got intel into its current state.

>> No.66248608

Double the TR number for TR2 and you have 6918 but without the $600-$1000 cooler and $10,000 cpu

>> No.66248728


Are you retarded or something?

7% increase in a year over the beheamoth nvidia is fucking insane.

but then again its mainly due to miners. and this might fuck AMD over badly soon

>> No.66248748

Zen+ actually has IPC increases.

>> No.66248867
File: 327 KB, 823x520, 1528270990961.png [View same] [iqdb] [saucenao] [google] [report]

here is a quick rundown

>> No.66248919

>6% jump is nothing guys -.-
have you ever worked a real job in your life?

>> No.66248928
File: 112 KB, 640x800, 18879778_1869105773337809_8791709922183086080_n.jpg [View same] [iqdb] [saucenao] [google] [report]

>sub ambient temps to hit that score

>> No.66248963

the housefire meme went to far. 4chan is a dangerous place to be.

>> No.66248976

well Intel said that their focus is not on PC anymore in 2016

>> No.66248995

Every fucking week, every fucking month, everything fucking year these fucking threads are made and every fucking time Intel and nvidia come out on top with better performance. Just stop fucking crying you amdrones

>> No.66249010


>> No.66249023

thank God for putting me on the greatest timeline, i might still be a virgin at 26 (soon 27) but at least I get to witness a real holocaust.

>> No.66249055
File: 162 KB, 633x900, 1494886290711.png [View same] [iqdb] [saucenao] [google] [report]


>> No.66249101 [DELETED] 
File: 54 KB, 679x758, 1527629778452.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.66249488

Shintelkikes literally used rocket science to cool their fucking cpu

>> No.66249553

>this delusion when 2700x exists and Vega is behind only the 1080Ti
Nvidia is stronger for sure, but Intel is just embarrassing.

>> No.66249581

circumference is π times diameter

>> No.66249607

For reference, a wall outlet can only supply 15A/1800W

>> No.66249640

Oh boy nvidiots amazing telemetry shit software. Such value added

>> No.66249647

OK I'm really late, but... Can you compare this to the heart of the Chernobyl reactor when it catastrophicated.

>> No.66249656

There's 20A outlets all over the place

>> No.66249679

Also it's a buggy shit pile and gives you wrong drivers or misidentifies the drivers you have half the time.

>> No.66249693

It was pulling over 25A

>> No.66249717

intel are facing an uphill battle now but nvidia are smooth sailing. they can bury amd graphics tomorrow if they wanted.

>> No.66249724

>muh gtx 1100
Nvidia is waiting for AMD's move, they did the same thing with the 1060.

>> No.66249731

Oh right I hadn't realized they were round

>> No.66249733
File: 42 KB, 664x538, Capture.jpg [View same] [iqdb] [saucenao] [google] [report]

feelin ok actually

>> No.66249736

They get ad dollars from intel

>> No.66249750

>they can bury amd graphics tomorrow if they wanted.
They absolutely do not want that because then they get split due to monopoly laws.
Nvidia fucking loves AMD to the point if they were really going under in the graphics division they would probably give them the license for making CUDA supported hardware.
As long as their graphics division is alive nvidia can keep on making money from international contracts for AI and other crap like they just did a while ago with lockheed martin, boeing and volkswagen.

>> No.66249767

It depends on the building. Some older ones can only do 10A on a single breaker.

>> No.66249783

>They absolutely do not want that because then they get split due to monopoly laws.
i know i was just saying they have the power to do that because they're so far ahead and basically dominating their market. intel could have been in the same position but they decided to go maximum jew with 5% incremental upgrades. jensen might be on the greedy side but he is always wary of competition e.g. 1070 ti to combat the 56. if that was intel they'd just let it go because they're too retarded and underestimate everything.

>> No.66249800

Intel should burn another 300m on worthless dyke cunts lol god I hope they go bankrupt

>> No.66249869

>stock price driven by paperlaunches and housefires and no actual marketshare growth
its a bubble and its going to pop

>> No.66250050

1cm^2 = 100mm^2
0.77cm^2 friendo.

>> No.66250082

> as compared to Intlels 28 core housefire
But can Threadripper do hardware simulation of shuttle reentrance for NASA? I don't think so...

>> No.66250120
File: 25 KB, 720x718, 1513534450670.jpg [View same] [iqdb] [saucenao] [google] [report]

This. I'm seriously considering shorting it.

>> No.66250267

*tips trilby*

>> No.66250276
File: 71 KB, 1280x720, Meteor.jpg [View same] [iqdb] [saucenao] [google] [report]

> 15A/1800W

Is that the standard in America?

I know wall outlet can pull 16A / ~3.68 kW at max in Germany, but even then, that's already 3.27kW used by CPU and cooling alone. Like holy shit.

Who exactly at Intel thought this was a good idea? Anybody with a bit of grasp of tech knows they're making fools out of themselves for something that's never going to see the consumer market and the normal consumer will eventually catch on with all the memes born from this.

Having surface power density of a reentry vehicle is ridiculous. Somebody should shoop a
> Intel Inside
on a meteor or a space shuttle.

>> No.66250414
File: 2.23 MB, 320x384, 1508125391216.gif [View same] [iqdb] [saucenao] [google] [report]

>15B company can beat 160B and 265B company this easily

>> No.66250657
File: 37 KB, 567x499, 1523170664517.jpg [View same] [iqdb] [saucenao] [google] [report]


It's because the certified shitwrecker worked his magic.
Now he's going to be working at Intel to unfuck them out of this hole that they've dug for themselves.
>Mfw the future is going to be Keller vs Keller.

>> No.66250713

But Keller can't unfuck the fabs

>> No.66250723

No institution worth any grain of salt owns a Threadripper/Epyc other than AMD itself

Threadripper are for system builders and no Fortune 500 company has ever moved to TR/Epyc either and if some of them do Intel still dominates the market

Laptops are all Intel

Nvidia has meme AI and deep learning shit plus unlike Intel at their height they're not giving a millimeter to AMD because they haven't been hit with anti-comp memes yet

>> No.66250763

They have to spin off their fabs sooner or later.

Otherwise, shekels will be lost, and they can't have that.

>> No.66250816 [DELETED] 
File: 365 KB, 2000x1153, 1497658837285.gif [View same] [iqdb] [saucenao] [google] [report]


>> No.66250855


>> No.66250905

Intel really expected to coast on their monopoly and now are scrambling to not lose marketshare and profits. What an absolute hilarious turn of events for them.

>> No.66250960

Not even that.

AMD's 64 core 128 thread is already in the labs and will be sampling to OEM for testing.

>> No.66251033


>> No.66251148
File: 2.51 MB, 1920x1080, delid.png [View same] [iqdb] [saucenao] [google] [report]

How will Intel Aviv ever recover?

>> No.66251191

Anon...I'm sorry, OpenCL is dead.

>> No.66251224

Nearly bankrupting would be the best way for AMD to beat Nvidia, but then they'd have to lose out on all the Ryzen progress. Since they could have CUDA source and and some money injected and then using source pump out a Vega+ compute monster that uses CUDA.

>> No.66251644

>50W/mm^2 for the shuttle
>5W/mm^2 for the cpu
>literally 1/2 of the reentry vehicle
am i being retarded or is he

>> No.66251685
File: 3.69 MB, 418x335, 1515983268443.gif [View same] [iqdb] [saucenao] [google] [report]


>> No.66251791
File: 323 KB, 1024x1024, Intel_New_Core_i9_28_cores_5ghz.jpg [View same] [iqdb] [saucenao] [google] [report]

I got something even better senpai.

>> No.66251951

50w is 5tw he made a mistake
its 5w

>> No.66253107

>quantum computers
whenever anyone mentions this you can pretty much guarantee they have no idea what they're talking about

>> No.66253137

>is this it bros?
It would be if AMD had faster products but they don't lol.

>> No.66253138
File: 23 KB, 500x294, 5985698656.png [View same] [iqdb] [saucenao] [google] [report]

i thought only one company was supposed to make housefires,whens the nuclear reactor powered by them coming?

>> No.66253147

>Threadripper are for system builders
Eh... not really.

>> No.66253204

Holy shit, they're rekt

>> No.66253217

>Brought a 1 horsepower chiller just to demonstrate their CPU on stage at 5GHz


>> No.66253282


>only displaying add-in card metrics


>> No.66253382


we have them.

1920s and 1950's.
dozens of them in a few labs near me.


>> No.66253394



>> No.66253526
File: 354 KB, 921x865, brianjust.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.66253579

Yes, their focus is nuclear reactors nowadays

>> No.66253739


>> No.66253803

>We can however still go back to vacuum tubes - litographically print them in silicon and copper like normal microtransistors. They can be made smaller...

Really? Is there a working process to make circuits out of vacuum tubes lithographically at densities similar to an equivalent 14nm FinFET circuit?

>> No.66253931

I thought the problems with Vega are high power draw and low clocks.
Both might be fixed by the move to 7nm.

But yeah, fucking GCN seems like an architectural dead end. They keep promising interesting shit and then don't deliver - primitive shaders for example

>> No.66253939

I think a group at Caltech is working on those, financed by Boeing.
And actually, the smaller the lithography is, the better they'll work, because they aren't "vacuum" tubes, they exploit the minuscule sizes at play to work at ambient pressure.
Basically, there are so few molecules in between, that most electrons flowing won't clash with them.

>> No.66254006

there is no excuse to that housefire cpu, like we have the tech but intel is just tryna cuck itself now

>> No.66254144
File: 8 KB, 76x29, Skjermbilde 2018-06-07 kl. 09.48.00.png [View same] [iqdb] [saucenao] [google] [report]


>> No.66254170


he has tits

>> No.66254311

primitives worked on the FE and the prosumers cards
but in games they tried to make them automatic which was an insane task to begin with and the usual result was that they let the devs to tap into them..

the problem of amd was and is their software team

>> No.66254337

ROCm aims to do that but the libraries are utter shit in real world applications and much slower than nVidia's much more mature libraries.

opencl will eventually be merged into vulkan with compute kernel modes enabled

>> No.66254480

No. The factor is 100. If you divide 770 by 100 you get 7.7
The anon was right.

>> No.66254508


Well yeah most still use their old Intel stuff from when AMD didn't really offer anything competitive.

But server farms were first hit by an up to 35% performance hit with Intel's shoddy security patches. Then AMD keeps on churning out pretty good server chips at low TDP and with comparable or even better processing power.

And every sysadmin is looking at Intel and is disgusted, and soon even the suits will realize that AMD is cheaper.

>> No.66254534

AMD Vega is trash the only reason why anyone would buy something like an RX 580 over a GTX 1060 is because of Freesynch

>> No.66254552

He's a wily and very smart bastard. It's no coincidence that he's, no shit, Lisa Su's uncle (technically - he's only 6 years older). Don't expect Jensen "Pork Shoulders" Huang to make the same dumbfuck mistakes that nearly a decade of clueless bean counters running the show has put Intel in the shitty position it finds itself.

>> No.66254555

>primitives worked on the FE and the prosumers cards
>the problem of amd was and is their software team

Well yeah, but if your hardware is so complicated to use (i.e. you can't just flip a switch in the driver and it gets used), the normies of software developers (the gaming studios) will simply not use it. Ergo, you might as well mothball the tech and pour your resources into more 'classical' improvements.

And don't get me wrong, I want some company to develop new ways of doing stuff. But if AMD sucks at providing ample support for their super duper hardware advancements, they might as well stop, cause no one uses their shit.

>> No.66254563

>Meanwhile Nvidia's CEO doesn't want to speak about new card releases

I dunno about Intel, but you are a FUCKING IDIOT if you think AMD is anywhere close to catching up to nvidia, let alone surpassing them. The ONLY reason why nvidia's CEO doesn't want to discuss new cards is because Pascal is still selling like hotcakes at above MSRP over two years after launch *because* AMD is so non-competitive.

>> No.66254611

Blame boomers who buy them.

>> No.66254629

>blame people for buying the best product in an industry where even the best is not good enough

>> No.66254644

You can measure power of any magnitude in watts, you heathen: that's what prefixes are for. There is no excuse to debase oneself by using non-SI units not approved for use with the SI in this manner.

>> No.66254657

the hardware isnt complicated to use at all

you cant explain the ability to code for consoles given that the card can achieve 30-40% async levels on them
and then say amd is difficult to code on pc's....
thats a load of bullshit everyone knows that gameworks is the reason why this shit is done
they save a lot of money by letting nvidia devs to implement the extensions

>> No.66254663

NGG and primitives never worked; not sure if postponed indefinitely or just canceled entirely. Bega for gaymurrs is dead in the water it seems. I think it would be more wise for AMD to target prosumer/HPC first since there's a lot of investors after this market and a lot of companies willing to buy cheaper alternatives for GPGPU, AND the GCN layout is really prime for HPC. AMD will need to funnel more money into software in that area; it's also much easier to get into, graphics pipeline and GPU programmers are not cheap nor easy to come by. AMD missed their chance completely when they had a massive ALU density advantage since they had 0 ecosystem for their cards and no one gives a shit about OpenCL. 7nm will at least give them a change to work on the ecosystem I'd hope.

>> No.66254682

Nvidia still has the upper hand. Intel tho, they're gonna tank till 2020. Not totally finished, they still have that mindshare and normies buy "muhhh intel inside" but AMD will make significant gains.
Workstation and server market should start reflecting it soon, but it'll be very slow.

>> No.66254929

>Threadripper are for system builders and no Fortune 500 company has ever moved to TR/Epyc either and if some of them do Intel still dominates the market
I see you get your news directly from Intel.

>> No.66254955

>AMD is so non-competitive.
The only reason AMD isn't competitive is because miners are STILL driving up the prices on their GPUs. If both AMD and Nvidia GPUs were selling at their MSRP, AMD would be perfectly competitive.

>> No.66255082
File: 32 KB, 410x266, needful.jpg [View same] [iqdb] [saucenao] [google] [report]

Sir... Sir please.

>> No.66255147

>1.35x Performance increase
>2x more energy efficient
it's either one or the other dumbfuck

>> No.66255224

>Those bent fins
Low quality trash, stay with bequiet or Noctua.

>> No.66255241
File: 222 KB, 590x332, David Wang 7nm process.png [View same] [iqdb] [saucenao] [google] [report]

OP here, you're making yourself look dumb

>> No.66255267

>AMD scales nearly 1:1 with more cores
>[email protected] 3459
>[email protected] projected 7500
>powerdraw under 500W
>cooled by air, not a fucking fridge

>> No.66255286
File: 143 KB, 1024x702, 4maAVRQTt9Q_fpvyhL9Li_HiHGnLVdi2U31C7hAD-Jo.jpg [View same] [iqdb] [saucenao] [google] [report]

at least i'm not falling for VEGA slides again. where are the prim shaders for gaymes, anon? remember this PR stunt?

>> No.66255293

Not the anon you're talking to but you're believing a powerpoint slide too much my dude.
2x power efficiency there could be anything.
I will wait for actual benchmarks before I start telling everyone their new generation cpus will have a TDP of 45 watts instead of 90.
Why I never see autismos who own intel shit post these slides tho? They just come with their muh gaymes run faster but you guys literally post press event pictures for media / marketing as if they were technical spreadsheets tested multiple times on oscilloscopes, multimeters and you're mums dick.

>> No.66255315

No it's not. It could be twice as efficient *and* use more/less energy to reach the 1.35x performance increase.

It's ironic you would call anyone a dumbass.

>> No.66255342
File: 218 KB, 1000x505, AMD-Vega-architecture-1000x505.png [View same] [iqdb] [saucenao] [google] [report]

>REEEEEEE why won't you fall for marketing slides again!
vega was alright but come on anon, you can't be this delusional.
where are prim shaders for desktop again? do us a favor and check if all this pic related stuff got into Vega and working, okay?

>> No.66255347

at the moment there are no benchmarks whatsoever so this is the only thing we can base our predictions on, remember I'm not saying that everything is confirmed, just asking people on here for their view and opinion on how things are going to end up in the PC market.

it's a server oriented card anon, not aimed at gaymes.

>> No.66255360


Those two answers are mutually exclusive. As I am not a hardware or a software engineer, I don't know who to believe here.

Anyone care to weigh in?

>> No.66255361

I'm not even an AMD fanboy. I was literally pointing out the flaw in your logic and that you should think twice before insulting others' intelligence.

Do you understand that it is possible to have both a 1.35x performance increase and 2x increase in energy efficiency?

>> No.66255376

AMD may have more cores, but the individual cores of Intel will be faster, making intel faster overall.

Amd can't even beat a gtx 1080 well over a year it had already been out.

>> No.66255382

>Do you understand that it is possible to have both a 1.35x performance increase and 2x increase in energy efficiency?
if the marketing slides said so it must be true! AMD didnt even deliver on their original promises for Vega AT ALL. you can't even prove me wrong. stick to their PR team, liddle man.

>> No.66255407

Marketing slides have nothing to do with it.
You should take some ritalin.

>> No.66255560

>Not the anon you're talking to but you're believing a powerpoint slide too much my dude.
Only applefags are as much emotinally invested in a tech company as amd fanboys. They shill for free. AMD created a cult of neet slaves. Pretty ingenious and financially smart imo.

>> No.66255569

well at least mining fags were never disappointed by AMD, do you agree?

>> No.66255582

miners aren't good to company tho, autists on here will probably never switch from AMD whereas a smart miner will just buy whatever is best at the time.

>> No.66255706


>> No.66255726

>the chad air-convection vs the virgin fridge-magnet

>> No.66255727

Primitive shades never happened because Raja tried to crash RTG with no survivors. The RTG software devs little sat around stroking their dicks for two years and didn't even start the Vega drivers until 3 months before hardlaunch. They were so far behind on Vega software development that it would have been nearing Navi's launch window by the time they finished actually developing the features Raja claimed for Vega, and seeing that Raja let RTG stroke their dicks for two years doing nothing while collecting paychecks, AMD decided to cull the cancer by firing that scam artist piece of shit and shit caning development of all of his vapor ware ideas instead of throwing good money and dev time after bad.

They've already made the decision the a substantial ground up revision or replacement to GCN is going to be required and to not waste any more dev time on Raja's bandaids. They need to change GCN to eliminate the enormous front end bottleneck that has always destroyed its ability to compete in the high end market.

>> No.66255848


>> No.66256036
File: 311 KB, 554x720, Confused Megumi.png [View same] [iqdb] [saucenao] [google] [report]

>mentions Polaris
Isn't the 580 stronger than the 1060?

>> No.66256096

make me a sandwich, bitch, and then suck me off

>> No.66256115

you have to be 18+ for this site, little man

>> No.66256117

Krzanich was a mistake

>> No.66256125
File: 1.25 MB, 1002x1480, t.JUSTnich.png [View same] [iqdb] [saucenao] [google] [report]


>> No.66256151


>> No.66256165

keller is doing some stuff with ARM last time i checked. dunno if he wants to help Intelaviv with CPU parts

>> No.66256195


Holy fucking shit

>> No.66256249
File: 190 KB, 853x480, 1454140123431.jpg [View same] [iqdb] [saucenao] [google] [report]

>intel individual cores will be faster
but at what price

>> No.66256257

That's what ur waifu is good only at.

>> No.66256269
File: 742 KB, 1055x808, 1525363070293.png [View same] [iqdb] [saucenao] [google] [report]

>implying my waifu isn't 2D
kek, what a fucking pleb

>> No.66256299

Pretty even and depends on benchmark which one is on the top with marginal lead. However as long as nvidias driver situation on linux is what it is their cards are not a realistic option

>> No.66256301

i can

"Melting/10 would go critical mass again" - IGN

>> No.66256336

>However as long as nvidias driver situation on linux is what it is their cards are not a realistic option
What do you mean by this? Their proprietary drivers work, but they're pretty bad for everything else. Switched to a 580 recently and it just werks out of the box.

>> No.66256360 [DELETED] 

>Mfw the future is going to be Keller vs Keller.

Can't wait for Keller uploads his mind and becomes skynet

>> No.66256390
File: 776 KB, 1111x597, 1524456728026.png [View same] [iqdb] [saucenao] [google] [report]

>yfw those fabs start making AMD chips for more volume

>> No.66256391

They should just make an AI from Keller's consciousness and call it a day.

>> No.66256432
File: 292 KB, 814x640, 85627D1E-E387-4A5F-9EDE-8819EDEE9AD1.jpg [View same] [iqdb] [saucenao] [google] [report]

>29 (twenty nine) phases

>> No.66256451
File: 5 KB, 229x220, Meat Mountain.jpg [View same] [iqdb] [saucenao] [google] [report]

>4x 8-Pins

>> No.66256454

about to kill myself when the price of all this hardware is gonna shoot right up in australia post 1 july. fuck

>> No.66256487

It's 28+4 phase for the cpu

>> No.66256520
File: 1.83 MB, 700x5000, 1520948519741.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.66256535

Actually googled Intel and nuclear power. Very witty.

>> No.66256622


>> No.66256639

google for Hope Solo nudes

>> No.66256664


>> No.66256675

>AMD is going to overtake intel/nvidia this time!

Like and subscribe if you've heard this before

>> No.66256676

you will change your mind very fast.

>> No.66256720
File: 260 KB, 814x640, 1528376685690.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.66256856

True, but my professor for example uses stock GPU's to run his Jupyters. Times are a lot different now than when TensorFlow first popped out.

>> No.66256905

Not the guy in post; just a canuck boy.
JBP is actually a fairly worthy figure and most people don't give him credit where it is due.
He's by no means perfect and he goes on about muh postmodernism a bit much, but he presents some very interesting thoughts; at least to me.

>> No.66256917
File: 54 KB, 858x424, jPeterstein.png [View same] [iqdb] [saucenao] [google] [report]

>Jordanian Peterson

>> No.66257275

>Intels 28 core Xeon is $8000+
>AMD TR 2 32 core probably around $2000 ish

what will it be, Anon?

>> No.66257290


>> No.66257332

well I did say 8000"+" so technically I'm not wrong

>> No.66257915

isnt it frequency* voltage^2* capacitance?
read it somewhere, if you know your shit tell me more

>> No.66258271


Vacuum tubes get ridiculously hot, unless you're talking about some newfangled design.

Name (leave empty)
Comment (leave empty)
Password [?]Password used for file deletion.