>Zhaoxin Shanghai to launch a 16nm x86 CPU in 2018 >8 cores

>>60868190
wow from backdoor to frontdoor, welldone NSA

>>60868190
Whoa...
So this ..
Is the power of ...

>>60868196
windows 10 is the frontdoor

this chip is just a chinese backdoor, which is probably preferable than a CIA backdoor because Chyna doesn't really give a fuck about you unlike CIA

As long as it has the single-core performance of an amd fx at the very least, I'd buy one.

>>60868190
This is a very bad thing for AMD. Please do not buy this if you care about the future of desktop computing.

>>60868292
If it's bad for anyone, it's Intel.

>>60868292
Only freetards will buy this.

>x64 compatible
>amd patent
How?

>>60868321
China.

True freedom.

>>60868303
You clearly underestimate the retardedness of Intel buyers. If Intel releases shit in a box they will still pay $400 for it, like what they're already doing with the 7600K.

>>60868321
Cant enforce patents in china

>there are people on /g/ right now who buy or don't buy based on brand logos
ha ha ha oh wow

>>60868345
I'm buying because it doesn't have a brand logo.

File: Mysterious Merchant.gif (4KB, 452x523px) Image search: [Google]

4KB, 452x523px

>chinese backdoor instead of nsa backdoor
DELET

File: 1473485265314.gif (869KB, 400x265px) Image search: [Google]

869KB, 400x265px

>>60868190
>Zhaoxin Shanghai

>yfw you run REACT OS on it
>yfw you are running a reverse engineered operating system on a reverse engineered chip in 2018
>yfw in 2018 the fucking x86 x64 isa and intel silicon design and fab process has been successfully been clean sheet reimplemented but windows 2000 has not......

>>60868321
It'd be funny if AMD OK'd this just so intel would have more competition to drive prices down

>>60868382
>but windows 2000 has not......

Wine can do everything Win2k can at this point.

>>60868386
It's not like they have a choice, unless they threaten to invade China.

>>60868321
>>60868331
>>60868344
>https://en.wikichip.org/wiki/zhaoxin
>Founder VIA Technologies
>Zhaoxin x86 ISA license statues is unknown although VIA has a cross-license agreement for x86 with Intel until at least 2018 (as ruled by the 2010 FTC order).

>>60868423
And the AMD64 and AVX instructions? Not in the same deal.

>>60868332
I AM an "Intel buyer". Intel CPUs of that size are way more expensive to manufacture than either AMD or this Chinese shit. These guys could sell their CPUs for peanuts which means zero margins for Intel.

Finally a cpu without CIA backdoors
We just need to suck Chang's dick instead

>>60868423
Surely its just a convenience to license from Intel now? Patent expiry date is 1997 so MMX should be free, pretty sure the x64 extensions were intentionally left open by AMD.

Are there specific no-gos or are they just avoiding incremental feature loss like noexec or VM extensions?

>>60868190
will probably end up in chinese tablets/nettops
but i've used a couple VIA desktop cpus before (many years ago, mind you) and have been pretty underwhelmed, so i won't get my hopes up

is it hardware backdoored too? whats the tdp?

Thank you based Chinaman for freeing us from the jew and the pajeet

>>60868190
Holy shit nice! A new CPU vendor that isn't intel or AMD, I'm glad to see this. Competition is good for everyone.

>>60868587
>new
anyway your post is half redundant, obviously a "new" cpu manufacturer wouldn't be intel or amd.

>>60868573
Most likely you are not getting better hardware than what intel/amd can provide. The news in that chinese companies are catching up really fast with them.

We're entering a new era or ricing

is there any information about what socket this uses? what motherboards would it work with?

™

>>60868604
of course and thats fine, im still curious about any remote management.

Also you people need to stop/ saying retarded shit

>>60868541
>pretty sure the x64 extensions were intentionally left open by AMD.
They werent. AMD used x64 to force Intel to license a bunch of stuff like the MMX and other SIMD instructions to them. It is why their 3DNow extensions died.

>>60868619
all their current CPUs are BGA (not socketed), and they make their own chipsets as well

>>60868247
Having all the manufacturing done in a single country

File: a.png (90KB, 1052x652px) Image search: [Google]

90KB, 1052x652px

you guys are talking about licensing issues as if they aren't already making x86 cpus
the news here is the move to 16nm, not them starting to make x86 cpus

pic related, a couple of their current cpus

>>60868749
>FSB
>2017

>>60868749
They don't use IMC's?

>>60868645
>can't find info in manufacturer's official website
>call other people retards
hint:
http://en.zhaoxin.com/ZXC.aspx?seriesid=5
for the ZX-E you will have to wait until they can actually make a working sample.

>>60868749
>8 core
>cache 4MB
cachelets

>>60868749
How powerful is this vs the Jew/Pajeet equivalent?

File: 1_640_355.jpg.png (157KB, 640x355px) Image search: [Google]

157KB, 640x355px

Lenovo "Zhaoyang" CF03 powered by KaiXian ZX-C C4600 2.0GHz

info: http://www.wanhuajing.com/d636784
cpu specs: http://en.zhaoxin.com/ZXC.aspx?seriesid=3

So, prepare your [s]anus[/s] wallet.

>>60868896
Apple btfo.

>>60868896
>http://www.wanhuajing.com/d636784
No L3, DDR3 ram

>>60868749
Fail.

>>60868958

gaymer?

C for cheap chink cpu.

>>60868749
>8 core chinese cpu
>35W
>intel 8 core
>140W

So now both NSA and MSS get to see my porn?

>>60868896
Are they intended for the international market?

>>60868749
This says they use ECC cache ....?

First i have heard of ECC cache. Who else does this ?

>>60869348
>First i have heard of ECC cache. Who else does this ?
Intel, probably everyone.

>>60869348
according to wikipedia it's not so uncommon:
https://en.wikipedia.org/wiki/ECC_memory#Cache

>>60868190
How many patents does this infringe on?

My guess is so many that it is only viable for domestic use.

File: 3_407_265.jpg (29KB, 407x265px) Image search: [Google]

29KB, 407x265px

>>60868444

confirmed AVX/2 instruction sets

>how ?

>>60868190
>Welcome back CYRIX

>>60869475
>welcome back shitx
Fixed.

>>60869022
Look at the clock speed though.

>>60869022
>>8 core chinese cpu
>>35W
>>intel 8 core
>>140W
https://ark.intel.com/products/91200/Intel-Xeon-Processor-D-1539-12M-Cache-1_60-GHz

And the Xeon-d has far better IPC than the chinkshit chip.

File: 1430236296498.gif (112KB, 255x231px) Image search: [Google]

112KB, 255x231px

>>60869515
>Recommended Customer Price $590.00

>>60869416
>bus speed 333mhz
Oh fuck boys, 2007 all over again.

If it has fully working vt-x, I'm buying the shit out of one.

>>60869489
Cryrix was better than both Intel and AMD at the time.

>>60868249
They do. You obviously don't know what commies are

>>60870169
>at the time
That's an interesting way to write "never".

File: .png (5KB, 487x170px) Image search: [Google]

5KB, 487x170px

>28nm
What year we in?

>>60870240
1991

File: btfo'dchinks.png (19KB, 1220x228px) Image search: [Google]

19KB, 1220x228px

>>60869022
CHINKS BTFO'D

>>60868332
the vast majority of intel buyers buy them not because they choose intel, but because the laptop they bought has an intel inside

File: 1438753303032.jpg (49KB, 455x224px) Image search: [Google]

49KB, 455x224px

>>60868749
>Maximum frequency of frequency: NO
Sorry I really need to have the maximum frequency of frequency

>>60870340
it was just a google translation

>>60870240
16nm in 2018.

>>60870412
we 7nm nao nigga

>>60870286
>release price $???

>>60870440
>7nm
in 2018

>>60868190
>>60870440
>>60870412
>not using 3nm subterrainean technology

>>60868332
> If Intel releases shit in a box they will still pay $400 for it.
So just like AMDfags with FX?

>>60870601
TSMC, Samsung, and GloFo are ramping up their differing 7nm processes now. They'll have risk production at end of the year.

>>60870210
Because the 6x86 was more efficient on an instructions-per-cycle basis than Intel's Pentium, and because Cyrix sometimes used a faster bus speed than either Intel or AMD, Cyrix and competitor AMD co-developed the controversial PR system in an effort to compare its products more favorably with Intel's. Since a 6x86 running at 133MHz generally benchmarked slightly faster than a Pentium running at 166MHz, the 133MHz 6x86 was marketed as the 6x86-P166+. Legal action from Intel, who objected to the use of the strings "P166" and "P200" in non-Pentium products, led to Cyrix adding the letter "R" to its names.

>>60870791
Unlike AMD, Cyrix had never manufactured or sold Intel designs under a negotiated license. Cyrix's designs were the result of meticulous in-house reverse engineering and often made significant advances in the technology while still being socket compatible with Intel's products. In Cyrix's first product, the 8087 math co-processor, Cyrix used hardware math multipliers rather than the CORDIC algorithm, which allowed the chip to be faster and more accurate than Intel's co-processor. Thus, while AMD's 386s and even 486s had some Intel-written microcode software, Cyrix's designs were completely independent. Focused on removing potential competitors, Intel spent many years in legal battles with Cyrix, consuming Cyrix financial resources, claiming that the Cyrix 486 violated Intel's patents, when in reality the design was proven independent.
Intel lost the Cyrix case, which included multiple lawsuits in both federal and state courts in Texas. Some of the matters were settled out of court and some of the matters were settled by the Court. In the end after all appeals, the courts ruled that Cyrix had the right to produce their own x86 designs in any foundry that held an Intel license. Cyrix was found to never have infringed any patent held by Intel. Intel feared having to face the antitrust claims made by Cyrix, so Intel paid Cyrix $12 million to settle the antitrust claims right before a federal jury in Sherman, Texas was to hear and rule on the antitrust claims. As a part of the settlement of the antitrust claims against Intel, Cyrix also received a license to some of the patents that Intel had asserted that Cyrix infringed. Cyrix was free to have their products manufactured by any manufacturer that had a cross license with Intel, which included SGS Thomson, IBM and others.

>>60870827
The follow-on 1997 Cyrix-Intel litigation was the reverse: instead of Intel claiming that Cyrix 486 chips violated their patents, now Cyrix claimed that Intel's Pentium Pro and Pentium II violated Cyrix patents—in particular, power management and register renaming techniques. The case was expected to drag on for years but was settled quite promptly, by another mutual cross-license agreement. Intel and Cyrix now had full and free access to each other's patents. The settlement didn't say whether the Pentium Pro violated Cyrix patents or not; it simply allows Intel to carry on making products under a license from Cyrix.

Which cpu did they copy if off of?

>>60870836
its really sad what happened to cyrix.
>In August 1997, while the litigation was still in progress, Cyrix merged with National Semiconductor (who also already held an Intel cross-license).
>The merger also resulted in a change of emphasis: National Semiconductor's priority was single-chip budget devices like the MediaGX, rather than higher performance chips like the 6x86 and MII. Whether National Semiconductor doubted Cyrix's ability to produce high-performance chips or feared competing with Intel at the high end of the market is open to debate. The MediaGX, with no direct competition in the marketplace and with continual pressure on OEMs to release lower-cost PCs, looked like the safer bet.
>National Semiconductor distanced itself from the CPU market, and without direction, the Cyrix engineers left one by one. By the time National Semiconductor sold Cyrix to VIA Technologies, the design team was no more and the market for the MII had disappeared. VIA used the Cyrix name on a chip designed by Centaur Technology, since VIA believed Cyrix had better name recognition than Centaur, or possibly even VIA.
>Cyrix's failure is described by Glenn Henry CEO of Centaur Technology as "Cyrix had a good product, but they got bought by a 'big smokestack' company and they got bloated. When Via bought Cyrix, they had 400, and we had 60, and we were turning out more product."[2]
>National Semiconductor retained the MediaGX design for a few more years, renaming it the Geode and hoping to sell it as an integrated processor. They sold the Geode to AMD in 2003.
they didn't get defeated in the market.... they were very competive against both intel and amd. they got cannibalized after they got bought out by national semiconductor.

>>60870896
Why is x86 such a horror show of broken dreams?

>>60870634
>AMD
>$400

Spotted the butthurt Inteldrone, go buy a 1700 faggot.

File: 60908953_p2.jpg (189KB, 1018x661px) Image search: [Google]

189KB, 1018x661px

>>60868190
>Compatible with x86 32/64 instructions, supporting extended SSE4.2/AVX/AVX2 instructions

Current VIA CPUs already support all that, it is not new. But AVX2/AVX is just there for compatibility, it actualyl worsens performance (for example in x265).

The IPC of current VIA core (28nm) is between Silvermont Atom and AMD Jaguar.

What these 16nm SoCs will bring is rise of the low 2,0GHz frequency, so realistically +50% singlethread performance.

Also they are now finally getting rid of the FSB and memory controller/iGPU outside of CPU. Finally.

In the big picture, this thing will only compete in the lowend and will likely still have bad performance/Watt ratio. Also bad iGPU.

Nobody really needs to fear Zhaoxin - the only harm it can do is by taking some sales on the basis of being chinese, for example in communist insitutions, schools, etc.

>>60868190
>16nm
How did they do that
There aren't any 16nm fabs in mainland China
There are plenty in Taiwan though

File: Intel is totally anal about Ryzen.png (257KB, 580x482px) Image search: [Google]

257KB, 580x482px

>>60868382
Zhaoxin is not a reverse engineered core. It is basicalyl continuation of VIA line of CPUs, which have cores designed by Centaur Technology, from USA. The company goes back to the 1990s x86 CPU market, they were providing the architecture for IDT WinChip. (and then VIA C3, C7, Nano)

>8 cores / 16 threads
Where did you find that they have SMT? That was not announced, it will still be 1core=1thread.

>>60871086
if the price is right they can take a good chunk of the low end market. who wouldnt buy a 8 core for under 100 bucks. in general this might be really good even for the people who wont buy them since it broadens the share of people with multi core systems even further(something intel has never done) and get better multi core performance for programms.

>>60868799
They are switching to IMC now, in ZX-D architecture, fucking finally. 16nm will be ZX-E.
Currently, they only have ZX-C on the market, 28nm FSB based.

>>60869348
ECC cache is norm, must for modern CPU. Appeared circa mid-1990s.

Only some ARM cores don't use ECC to save power. Pretty bad idea IMHO.

>>60868190
Cant wait to get a bunch of ebay prebuilts in for repair with this chink shit in.

Its bad enough with stripped mobos/psus that flood that market.

>>60869475
VIA abandoned all the Cyrix technology, they switched to CPU cores by Centaur/IDT which they also bought. It was a shame, maybe Cyrix could continue if somebody better bought it.

>>60868317
freetard here. why do i want chinese cpu?

>>60871299
Because they don't understand secure platform computing yet

>>60871299
You don't.
Stallman used chinese laptop at some point. But it was MIPS/Loongson. Also no good reason.

>>60869227
And the FSB

File: 1433276589653.gif (2MB, 388x356px) Image search: [Google]

2MB, 388x356px

>>60871299
BUY INTEL

>>60868292
>oi vey, delet

File: amdloo.png (316KB, 882x758px) Image search: [Google]

316KB, 882x758px

>>60868292

Usable, I think.
http://browser.primatelabs.com/geekbench3/compare/8377895?baseline=5973675

>>60871172
any low end mainstay would rip the shit out of that piece of garbage
even with all muh cores in action it barely ranks up against a fucking q6600 in >>60872195 and in most cases it will probably do worse

>>60870207
>You obviously don't know what commies are
The future?

>>60868249
yeah instead of distributing your information to law enforcement agencies they'll just sell it to private entities for a profit instead

>>60868190
>Cheap chink shit
Aka Chinese corporate spying stealing hard work of American on and and Jews of intel

>>60868749
So this is for their military equipment / space agency / government offices then?

>>60872682
it's just a shitty consumer chip for low-end systems

>>60871138
>from USA.
Into the trash with it.

>>60872724
So exactly what is used in military hardware.
>friendly reminder that the F-22 and modern missiles use computer chips designed in the 90's

Here are some benchmarks of their 28nm 2.0GHz quadcore. I suspect IPC won't change much, except for the effect of the integrated memory controller.

http://forum.cnews.cz/viewtopic.php?f=38&t=27768

>>60872968
nah not really, it's doubtful that thing is up to any military standards on things like rad-hardness and it's actually probably overkill for most of those kinds of applications anyway
>>friendly reminder that the F-22 and modern missiles use computer chips designed in the 90's
are you one of those retards that thinks you need a threadripper and 64 GB of RAM to do simple math?

>>60872510
DELET, PLS

>>60868190
>INTEL BTFO
No son, Intel gave them the license to make their CPUs x86 compatible.
Even as retarded as they are, Intel would not shoot themselves in the foot so badly.

>>60874011
They were more or less forced by regulatory bodies as well as their past obligations that go back to the IBM PC contract.

>>60872968
Reminder that the USS Ford is using an IBM supercomputer.

>>60868190
Do you have sauce on that info?

>>60868722
wut ?

AMDs new CPUs are made in the US in global foundries fabs
Intel only has a handful of fabs outside of the US, only one is in China.

This is more of a case of stolen corporate secrets/not paying licensing fees.

Does AMD or Intel dare potentially anger the chinese government if it tries to bring up a lawsuit ?

>>60872968
its milspec hardware, that is made to be very resilient and easily replaceable if needed and of course the vendor has to be able to service it for 20+ years


https://www.youtube.com/watch?v=55z_0BYb5is

>>60874181
google?
https://forums.anandtech.com/threads/via-zhaoxin-28nm-zx-c-2015-zx-d-2017-16nm-zx-e-2018-x86-processors.2471695/

>>60874199
>This is more of a case of stolen corporate secrets/not paying licensing fees.
like already said in this thread (like >>60871138
), this is not the case.

>>60871012
It's a quick-and-dirty solution that has overstaid its welcome by 30-odd years.

have the upstreamed any of the support?
is it competive? ie cheaper than amd but better than arm64?

does it have any hidden backdoors aka a soc within the die?


its useless otherwise

>>60874763
Fucking this. I wish it could die already. The only reason it's still relevant is because IBM happened to use it in a modular desktop computer system in the 80s that turned out to be easy to clone.

Can't wait for pajeets to start shilling this

>in b4 """""benchmarks""""" """""""leak""""""""""
>performs better than a 6950X
>finally released after counless setbacks (AMD much?)
>people try to do something convenient like play a game
>sandy bridge performance

Again if you want to cheap out just buy Sandy Bridge.

>>60868190
>Zhaoxin (Shanghai Zhaoxin Semiconductor Co., Ltd., also goes by VIA Alliance Semiconductor Co., Ltd.)

Just a few days ago I made a joke about via coming out with a multicore high clock x86 worldbeater. And now it happens

>>60868190
>trusting chinkshit
>implyng the specs aren't fake

>>60874763
>>60874862
it's not DOS you dipshits, the 8086/88 wasn't really all that different from any other 16-bit chip on the market in terms of "cleanliness"

File: wew.png (49KB, 615x173px) Image search: [Google]

49KB, 615x173px

>>60874912

>>60874893
Kill yourself, tripfag.

>>60868190
>16nm
Which foundry? Did China finally reach 16nm?

>>60868484
>Intel chips cost more to manufacturer
Sounds like a problem intel could solve with R&D. But they'd rather be lazy and just make you pay for it.

>>60874969
Intel is going to ride that ringbus right off into the sunset.

>>60874893
filtered

>>60874962
They're part of Via, so so all the silicon is subcontracted out to various companies that would have the capability of 16nm.

>>60874962
TSMC, it is fabbed normally on Taiwan.

>>60868277
this.
and if it's below $100.
would buy the 8 cores use it offline as encoding machine

File: 1413441555001.png (38KB, 430x721px) Image search: [Google]

38KB, 430x721px

>>60870827
>In Cyrix's first product, the 8087 math co-processor,
Cyrix didn't invent the 8087, which was an Intel product already nearly 9 years old by the time Cyrix released their kickass 80387-compatible FasMath chips.
>>60870896
Towards the end they really weren't that competitive, in and some markets (the PC high-end) they were never competitive at all. Shit happens when you build a business on someone else's proprietary technology that they can fuck you out of at any time.

>>60868190
u r fagget

>>60868277
>>60875566
It's slower than an Athlon 5350.

>>60868749
>2.0 Ghz
Looks like a VIA product to me!

>>60875618
That's the 2GHz quadcore. But yeah, better get used FX 4300 or something like that. X4 845 is quite cheap now.

>>60875503
Taiwan isn't (Mainland) China

>>60868444
>>60868321

The AMD64 extensions should come out of patent around 2020, the specification was released in 2000. So they may be running up to an international commercial release at the point it becomes possible. There hasn't been much more extension, the one that matters most for small systems is Noexec, and NX bit is from 2003 only a little later.

>>60870896
This reminds me of the DEC Alpha, buried(alive) by Intel.

>>60876391
nah the Alpha was dead by the time Compaq sold the IP to Jewtel after half-assing it for a few years waiting for Shitanium to come out

how many /csg/ retards are going to get chinked by this?

>>60877034
They're not a very bright bunch, so lots.

>>60868749
>Bus type
>FSB
current year

https://forums.anandtech.com/threads/via-zhaoxin-28nm-zx-c-2015-zx-d-2017-16nm-zx-e-2018-x86-processors.2471695/page-4

tl;dr they're shitty knockoffs from a new-ish company and aren't ready for serious production levels.

anyone want to set up honey pots with a few of these and some elbrus machines from russia? would be interesting to see the traffic going in and out.

>>60877034
I am.

>>60868599

That's some weapons-grade autism you got there.

>>60872343
Back to the Soviet Łagier, you communist scum.

>>60877034
You tell me

>>60868321
>x86 compatible
>intel
How? Because they are illegal in the US but its China. They don't give a fuck.

>>60868382
>clean sheet
Don't use that word if you don't know what it means.

>>60879958
>>x86 compatible
>>intel
>How? Because they are illegal in the US but its China. They don't give a fuck.
Nothing to do with it. X86 is decades old.

Holy shit I'm having a laugh here.
Most of you guys are fucking retards.

Have fun _smartasses_

>They can potentially be fashioned into non-volatile solid-state memory, which would allow greater data density than hard drives with access times similar to DRAM, replacing both components.[14] HP prototyped a crossbar latch memory that can fit 100 gigabits in a square centimeter,[73] and proposed a scalable 3D design (consisting of up to 1000 layers or 1 petabit per cm3).[74] In May 2008 HP reported that its device reaches currently about one-tenth the speed of DRAM.[75] The devices' resistance would be read with alternating current so that the stored value would not be affected.[76] In May 2012 it was reported that access time had been improved to 90 nanoseconds if not faster, approximately one hundred times faster than contemporaneous flash memory, while using one percent as much energy.

what the FUCK why aren't we using these?

>>60870286
>35W Atom
What's the point?

>>60868196

kekd.

>>60868190
>Compatible with x86 32/64 instructions,
>Compatible with x86 32/64 instructions,
>Compatible with x86 32/64 instructions,

Yeah... no thanks. Emulation will never be as good.

>>60880717
You are retarded on another fucking level, the likes of which /g/ only sees with microwave time memes.

File: 1024px-High_Bandwidth_Memory_schematic.svg.png (84KB, 1024x576px) Image search: [Google]

84KB, 1024x576px

>>60880472
It's research technology not necessarily ready for production yet.
HBM uses 3D memory stacks though, common in GPUs nowadays. Bandwidth is fantastic, but it requires the RAM to be on the same package as the GPU as well as a custom chip to interconnect the two.
https://en.wikipedia.org/wiki/High_Bandwidth_Memory

File: CPU_Z VIA CNR_zps46mqno20.jpg (104KB, 483x479px) Image search: [Google]

104KB, 483x479px

>includes S3 graphics
Oh boi

estimated price? if is cheap maybe i buy one

>>60881090
It's not socketed, you will only find it embedded in desktop/laptop motherboards.

>>60874199
>AMDs new CPUs are made in the US in global foundries fabs
Really? Because at least Wikipedia says most of GloFo fabs are in Singapore.

>>60881375
It has 3 fab plants in the US, the 200mm plant in Vermont, and a pair of 300mm plants in New York. Whether these plants actually fab 14nm remains to be seen as it appears the wikipedia article for Glofo is out of date.

File: matsy.gif (89KB, 483x451px) Image search: [Google]

89KB, 483x451px

>>60881039

https://www.youtube.com/watch?v=JotRC775YWg

>>60881375
>>60881405
Global Foundries Fab8 in New York is their 14nm, and 7nm facility.
All of AMD's 14nm parts including Ryzen CPUs and Polaris GPUs are made there.

>>60881662
Oh, okay. I didn't know that.

>>60868190
Dumb question but whats that black stuff around the die?

>>60881780
Glue?

>>60881780
Epoxy/resin. Protects the die.

>>60871086
>Nobody really needs to fear Zhaoxin
Why would anyone? If they actually do manage to produce better CPUs than Intel and AMD, wouldn't that just be a good thing?

>>60882000
Better CPUs not likely, cheaper CPUs maybe.

>>60882091
I know, but hypothetically.

>>60868190
If they release this as being compatible with 2nd generation Core i3 motherboards, this could become a real winner.

>>60882307
Ha no, not gonna happen. It uses a completely different interface for one, has no IMC, and intel would freak at a potential competitor's chips being used on their boards and chipsets.

>>60882307
>If they release this as being compatible with an outdated platform that isn't manufactured anymore...
If they were that stupid, they'd be bankrupt already.

File: 1489280723661.gif (2MB, 330x166px) Image search: [Google]

2MB, 330x166px

>Not even 18 cores before 2017

>>60880419
Intel holds x86 license regardless of whether or not its decades old.

>>60882583
>Intel holds x86 license regardless of whether or not its decades old.
The original x86 patent expired a long time ago. Whether or not they have a license is irrelevant.

>>60876275
That's not what I meant. I meant: at TSMC as any other mainstream chip, and not actually in the communist part. People assume that it is going to be fabbed in some chinese fab because, but that is not true.

AFAIK, neither Loongson, nor the similar russian "domestic" play were manufactured locally.

>>60881375
Singapore should be the site of the fabs received via the Chratered Semi acquisition. There are also the original AMD fabs in Dresden, Germany.

>>60882000
I was speaking about the competitors.

>>60882856
*Chartered

bretty good.
competition is what we need.

>>60882583
Intel holds the pantent forthe newer instructions only. AVX, AVX2, etc.

>>60868386
AMD licensed x86 IP to chinks last year, not this chinks though

File: Che_shirt_sells[1].gif (5KB, 179x184px) Image search: [Google]

5KB, 179x184px

>>60872343
hahahahaha

>>60884335
That's a good point. The chinks certainly care about a bunch of IP patents held by barbarians.

>>60885825

Doesn't matter there sure, but Intel would just get a ban on imports

>>60868190
And people think Intel has backdoors in their CPUs
God knows what these Chinks are gonna do.

>>60882371
>has no IMC
So it's going to be trash.

>>60868190
it has infinity dragon?

>>60868190
Actually a sub-company of VIA Technologies - who has been making x86 and x86-64 chips since the 1980s.

>>60868277
>the single-core performance of an amd fx
Performance on this one is a REALLY good question.

VIA released one generation in 2006 and just one chip in another generation in 2011. That's it.

>>60868321
>>60868331
>>60868344
Again, this is just a branch of VIA Technologies. They actually own important x86 patents themselves.

>>60880704
Is a 16 core in order CPU not everything you ever wanted?

>>60869515
>Xeon-d has far better IPC than the chinkshit chip

It doesn't matter, not even a little. For a whole lot of purposes IPD, Instructions Per Dollar, matter a whole lot more.

>>60870240
> 28nm
> 333 MHz Bus Speed
Wait a second. Is it 1999? Feels like 1999. I got a brand new Ryzen system today and that wasn't released in 1999 but now I'm not sure.

>>60870855
They didn't. They didn't in the past and they very likely didn't now.

>>60871086
>AVX2/AVX is just there for compatibility, it actualyl worsens performance
I really disagree with you on this, for special purposes like RAID6 you're looking at a pretty huge benefit when you're using AVX2. But.. it's not really that huge over SSE2 either.

>>60874199
Intel CPUs are made in Costa Rica iirc

>>60871172
>who wouldnt buy a 8 core for under 100 bucks

I'd probably buy a laptop with one just because it does x86-64 but I wouldn't buy it for a desktop. It could be nice for servers too.

As for cores, my Lenovo K3 Note from 2015 has a octa core CPU. It's about time this becomes the norm on x86-64 too.

And don't give me that IPC Intel Fanboy garbage. Nobody cares, not even a little. GNU/Linux has been using all available cores for all demanding tasks for almost a decade. Total processing power across all cores is what matters to me personally. Windows is trying to be ARM compatible so that joke of an OS will have to catch up to a multi-threaded world.

>>60874011
>Intel gave them
Actually.. Intel sued them for legally making 486 chips and lost big-time and were forced to hand over cash and some patents.

>>60872343
Let's fucking hope they stay in the past (and China/Russia as much as they like to draw "inspiration" from their "historical greatness" - I guess mass murderers sometimes became popular).

t. somebody who lived in that shit

File: tmp_11205-nanoha_8_12131560231.png (575KB, 672x512px) Image search: [Google]

575KB, 672x512px

>>60886369
>I really disagree with you on this, for special purposes like RAID6 you're looking at a pretty huge benefit when you're using AVX2. But.. it's not really that huge over SSE2 either.

You misunderstood. Sure that the instructions can bring benefit - but not in the VIA core.

Basically it has slower performance, if you take for example x264 and disable AVX2 (making the program to use two SSE2-4 ops for every AVX2 one), it encodes faster.

Basically they split AVX/AVX2 ops into 128bit halfs (if it even has 128bit SIMD unit and not 64bit) and do them in two steps. And apparently it clogs the chip so the performance isn'T even the same, it drops.

Of course, that is based just on few encoding benchmarks, maybe it doesn't harm in other tasks.

China is setting up a fabrication plant for these chips in Uganda.

>>60886517
>GNU/Linux has been using all available cores for all demanding tasks for almost a decade. Total processing power across all cores is what matters to me personally. Windows is trying to be ARM compatible so that joke of an OS will have to catch up to a multi-threaded world.

This absolutely isn't about OS, it is about application software you run on it.

If you OS hogs more than single-digit % of CPU on multiple cores, something is seriously wrong.

>>60879958
>>60868331
>>60868321
>>60868344

Intel, AMD, Via and IBM all hold the full x86 licence.

>>60875922
>x4 845
I had one of those.

>>60887515
>IBM
>x86
No.

>>60887619
IBM actually produced x86 CPUs at some points. Mainly in collaboration with Cyrix as the designer.

>>60887619
first x86 were created for IBM pcs. IBM even produced some x86 cpus.

>>60869385
Yeah, this is never going to be sold in the west.

>>60887335

Why are you poor fucks always against your own interests? Doesn't make sense.

>>60868190

Can't wait for my new 8 core $80 xiaomi cpu.

>>60868190
But will it support MMX?

>>60881481
Their single core performance is supposedly incredible, I wonder how well they do for emulation against Intel's Pentiums.

>>60870855
Atom/Celeron going for the moar cores strategy.

So, much talk about the CPU.

What about the chipsets?

Eventually we will have to learn how tell fake CPUs apart because the chink nature.

>>60872343
back to /pol/

>>60888992
*leftypol