I remember a more or less recent x86 processor manufactor that

>>61351062
https://en.wikipedia.org/wiki/List_of_x86_manufacturers

File: vortex86.png (400KB, 963x520px) Image search: [Google]

400KB, 963x520px

>>61351085
>rise technology -> vortex86
http://www.vortex86.com/

Why didn't they just list it right on the page? Anyway, thanks.

>>61351146

>Vortex86DX2
>DX2

aah, brings back memories

>>61351062
I'll never forgive HP for killing DEC

>>61351435
Alpha/PRISM was dead in the water.

File: shockednegro.png (26KB, 140x133px) Image search: [Google]

26KB, 140x133px

>tfw x86 still hasn't died

>>61351498
So what Architecture are you posting from? Are you an ARM phone poster?

>>61351518
I hope my current desktop will be my last x86. I would loved to have the Talos Secure Workstation, but that never got off the ground. It would probably also be a quite a bit more powerful than I need, ARM in desktop format would probably be nice.

>>61351560
>I would loved to have the Talos Secure Workstation, but that never got off the ground

If you join their IRC, you can find out that they're working on a second one. They're tight collaborating with IBM and already have access to the upcoming POWER 9. So they will release the Talos II (nice Star Trek reference) with POWER 9 "someday"

Keep in mind that system would be 5 digits though. Not sure if your parents allow you to spend that kind of money to run Arch.

>ARM in desktop format would probably be nice.

ARM is a bigger closed-source mafia than x86. Don't get fooled. It's not better just because there are more manufactorers. I work for a company that licenses Rokchip and Mediatek chips. A ton of NDAs to see some basic specifications make the long-term support of a product virtually impossible.

>>61351581
>If you join their IRC
Haven't been on that since they had to cancel it, I will surely check up on that as soon as I get home though.
>Keep in mind that system would be 5 digits though
The first version was not gonna be cheap either. I'm just happy that there are alternatives popping up again.
>ARM is a bigger closed-source mafia than x86. It's not better just because there are more manufactorers
Oh well, I just think it's a bit depressing to go through the screenfetch threads and see everyone having the same CPUs.

File: ARM Intel.png (1MB, 1556x816px) Image search: [Google]

1MB, 1556x816px

>>61351498
Not sure what will happen

>>61351864
This will either kill intel or make them competetive again

File: EPIA-M10000a.jpg (57KB, 640x480px) Image search: [Google]

57KB, 640x480px

What is your opinion of Via, /g/?

>>61352907
I buy a few everytime I go visit my wife's parents in Taiwan. Nice toys

>>61352927
What do you do with them?

>>61352945
Everything people do with their Raspberry Pi, just better

>let them catch dust
>door stopper
>show off to geek friends

Also I sell OpenBSD NAS and other small server devices based on VIA boards.

>TFW had a Cyrix CPU which caused a specific bluescreen in Windows 95

That's when I stopped buying knockoffs (including AMD) and stuck to genuine Intel.

>>61354877
>i only buy genuine intel(tm) AMD64 chips
uhuh.

>>61352927
Is your wife pretty?

>>61354940
I'm a fat neckbeard so she's the best I can get. Also sometimss I pretend that she's Japanese because I'm a weeb

>>61355048
Post

>>61355048
>Philosophers and poets have given many wonderful definitions of love, but to keep it simple, there are only two main components:
>1. Mutual attraction
>2. Lack of other options
Truly.

>>61351435
I agree DEC was badass

>>61351498
It did, a long time ago.
Now its just a translation shell, since the pentium pro.

>>61352907
Trash. Literally 100x slower clock for clock than AMD. I benchmarked.

>>61352009
ARM is dead in the server market. It has worse performance per watt, bad multicore peformance and bad single core performance. Xeon D killed it. AMD EPYC prevented it's resurrection and now Intel is releasing even more powerful servers.

The raspberry pi is basically the only usable SBC because every other board uses ARM GPUs with proprietary drivers written for some old as fuck linux kernel version.

I don't know but maybe ARM found a niche for underperforming laptops for less than $100. Anything higher than that and you can just buy a x86 laptop.

>>61355261
I've been wondering, what's the benefit of doing it like that vs just building a cpu that runs x86 code natively?

>>61355635
The x86 architecture is bloated and inefficient to actually implement on silicon. It's more efficient to build another architecture and emulate x86 on a single abstraction layer

>>61355800

That's also a good description of your mother.

>>61355635
There's a lot of highly technical reasons behind it but the basic explanation for why RISC has largely taken over is that RISC-like instructions are smaller and simpler to handle so you can implement them with less hardware and the space savings from using RISC-like hardware means you can add more hardware to handle more instructions at a time in parallel. One way they increase the amount of hardware is registers, RISC architectures will have way more general purpose registers on average, modern RISC based x86 processors have like 70+ general purpose registers. The "named" registers that x86 uses exist purely in a simulated form. Tasks are assigned a named register and that name is assigned to a free hardware register. The same name can be assigned to any register that is free. If you rearrange tasks before they're sent to the registers you can assign tasks that do not depend on each other to different registers and execute these tasks "out of order".

There's just a lot more ways to optimize RISC-like instructions which is why they've basically taken over.

>>61356030
>If you rearrange tasks before they're sent to the registers you can assign tasks that do not depend on each other to different registers and execute these tasks "out of order".

You can execute the tasks at the same time basically. It significantly increases efficiency at the cost of power and heat generated.

>>61351062
You are talking about VIA.

>>61351062
Transmeta? Wasn't recent, though and they emulated x86, though.
>>61352907
Their license for creating new x86 chips expired afair. That's good because their drivers were Win-only botnets.

>>61356030
>>61355800
>RISC has largely taken over
Taken over? In which sector exactly? (Except mobile). Heck, even consoles moved to x86.
>You can add hardware and handle intructions in parallel
Only in certain scenarios where instructions don't depend on each other. Usually CPU tasks tend to be more sequential than, say, GPU ones.
It's of vital importance for a CPU to have good single-core performance, and x86 delivers.

File: Xeon.jpg (65KB, 681x519px) Image search: [Google]

65KB, 681x519px

>>61355300
>Xeon D killed it.
With those price? Lol

>>61357536
>With those price? Lol

Oh no, he's retarded.

File: how-to-fix.jpg (138KB, 697x419px) Image search: [Google]

138KB, 697x419px

>>61351146
>Why didn't they just list it right on the page?

>>61357536
Xeon D is more expensive than Xeon E. It's DOA for low end stuff in which ARM reigns supreme.

>>61355300
>ARM is dead in the server market.
But muh Scaleway...
https://blog.online.net/2017/04/27/scaleway-disruptive-armv8-cloud-servers/

File: man of culture.jpg (51KB, 1024x576px) Image search: [Google]

51KB, 1024x576px

>>61357613
>Snagit

bump ax, ax

>>61351146
seems like they have no bios licence. they make x86 socs because of that.

>>61361980
>bios license
from who? we've been using clones since forever
there's even an foss (LGPL) bios implementation available (SeaBIOS)