Worst CPU ever

Any mobile chipset

8350

Pentium 4 was actually very based. It was the Pentium D that was terrible.

Bad CPU: AMD Phenom mobile processors. The mobile variants were fucking awful, battery draining hot plates.

>>56580455
>8350
it's bad, but not worst

9370, now that is a shitty shitty cpu

File: KL_VIA_C7_M.jpg (73KB, 1024x1032px) Image search: [Google]

73KB, 1024x1032px

VIA C7

i386

god the fx4100 was bad

Pretty much any atom cpu. I can barely run one as a next to bed fap machine.

>>56580425
The first gen atoms
The first celeron

>>56580530
Baytrail and newer are pretty good. Ifyour application is properly threaded, it can reach core 2 duo levels of performance

File: 220px-Intel_Pentium_D_Logo.png (34KB, 220x271px) Image search: [Google]

34KB, 220x271px

>>56580425

>>56580425
>P4
>Worst

So many people who haven't been cursed with the shit that is Cyrix MIIv, Durons, and socket 478 Celerons.

File: image.jpg (439KB, 1450x1314px)

439KB, 1450x1314px

>>56580425
Had 3 of these fuckers break down on me

The z520 in the Vaio P.
Goddamn Vaio, release a new one with an M processor. I'll happily give you my first born.

File: 8c4.jpg (13KB, 640x480px) Image search: [Google]

13KB, 640x480px

>>56580459
>>56580550

>>56580562
You say that, but the Durons were much faster than the first Pentium 4s.

>>56580459
>>56580550
is this a joke? i picked up a PD machine at one point and it was great

>>56580455

6/10. I would have taken you seriously if I didn't have one and know for a fact it's a fucking beast.

Bulldozer is most likely the worst uarch ever made if you look at how much damage it did to AMD's image and business.

>>56580580
Pentium Ds were nothing more than two Pentium 4s shoehorned into a single chip. Pentium 3 had better performance at lower clock speeds.

>>56580597
all i know is it felt much smoother than my next-best athlon64 3500+
that said, that pentium D machine was the first dual core machine i'd used, so maybe it was bad /for a dual core/, for which i didn't have anything to compare it with

>>56580588
dubs speak the truth

>>56580578
The Duron was barely faster than the Willamette and only the lower clocked versions, Northwood P4s were easily faster and overclocked far better too.

Fucking AMD shills and their Duron 800mhz back in the days, one of the worst purchasing decisions of that era.

File: h_1430240619_8975994_f17a717452.jpg (82KB, 600x800px) Image search: [Google]

82KB, 600x800px

> all these pedestrian opinions ...

actual worst CPU: random commieshit (slav or chink) that nobody here ever heard about

crappiest ISA that got popular: x86

crappiest microarchitecture that got popular: NetBurst

biggest flop by people who should've done better: Intel iAPX 432

The Datapoint 2200. Literally fucked the last 35 years of computing.

File: durpron.png (27KB, 464x350px) Image search: [Google]

27KB, 464x350px

>>56580578
lol no the Duron 1.6ghz was slower than the early P4 1.8A
You're thinking of the Celerons.

File: anandtech_benchmark.png (53KB, 843x632px) Image search: [Google]

53KB, 843x632px

>>56580621

You are being ironic, aren't you?

>>56580673
The 1.8A was a Northwood, therefore not one of the first Pentium 4s (i. e., Willamettes). What I said stands.

File: nstb9ePdOq1sjhtm9o7_r1_400.png (145KB, 300x300px) Image search: [Google]

145KB, 300x300px

>all the desperate damage control intel reps defending netburst ITT

Holy shit, I literally can't stop laughing!


>>56580459
No, the Pentium 4 was literal dogshit.

>>56580562
The Durons were faster, and the MII was just late to market.

>>56580580
Nope, they lost to the Athlon 64 X2 on almost every benchmark.

>>56580595
Replace "Bulldozer" with "Netburst" and "AMD" with "Intel", you shill.

>>56580616
No it didn't because it was slower, you're an obvious shill.

>>56580621
The Willamette was an embarrassment to Intel, it managed to lost in performance to literal Pentium 3 clocked over half 1GHz below, and it used DoA RAMBUS RAM and caused housefires too.

>>56580673
Nice cherrypicking, but anyone can see the whole comparison here: http://www.anandtech.com/show/1201

>>56580758
Why would Intel defend Netburst if all modern Intel processors use the objectively superior Core architecture? They pretty much admitted that Netburst was a failure.

>>56580688
While the Celeron cannot keep apace with its older brother or the Duron at the same clock speeds, it has to run much faster by default - so it not only offers fast performance, but it fits perfectly into Intel's "clock speed sells" strategy.
In the end, it seems to work out: The Celeron Willamette 1.7 GHz is currently the fastest budget CPU. At $83, it is even slightly cheaper than AMD's Duron 1300 ($84).

http://www.tomshardware.co.uk/good-old-newbie,review-468-12.html

Celeron Willamette is faster overall than the Duron and cheaper.

Amd e1-2100.
No other modern processors can beat that garbage

>>56580425
ITT: faggots who think "worst CPU" = "worst consumer x86 CPU made by Intel/AMD"

>>56580758
MII was late, couldn't scale well, had fucking heat issues and had that fucking 83mhz bus that fucked everything up. Also Soundblaster was incompatible.

any AMD cpu

not even trolling, i don't mind their GPUs, but every AMD cpu I have ever used has been total shit and was really slow

>>56580459
>It was the Pentium D that was terrible.
Hello sir the 805 will see you now.

>>56580850
I think the only reason AMD is still making CPUs is so that Intel doesn't become a monopoly. Gotta have those fair-trade laws,

File: 90ec5711f71b29c79a4e250b08d49c6e.jpg (90KB, 620x670px) Image search: [Google]

90KB, 620x670px

>>56580807
>Willamette Celerons were launched 15 May 2002, initially at 1.7 GHz, and offered a noticeable performance improvement over the older 1.3 GHz Tualatin-based Celeron part, being able to finally outperform an 1.3 GHz AMD Duron
>finally
https://en.wikipedia.org/wiki/Celeron#NetBurst-based_Celerons

Also, notice the launch date. Morgan Durons had been launched a full year earlier, and the 1300Mhz model was launched January that year.

Intel damage control shills literally grasping at straws at this point.

>>56580791
Damage control, son. They'd never admit their company has ever done something so embarrassing. They're literally trying to Correct the Record™.

>>56580578
>but the Durons were much faster than the first Pentium 4s.
So were the Pentium 3's

File: lmao.jpg (47KB, 634x447px) Image search: [Google]

47KB, 634x447px

>>56580880

So was literally anything.

>>56580507
Holy fuck I don't ever wanna see that fucking name again! First time I delta with one it took 2 hours to install office 2007! 2 fucking hours! That is a truly a masterful piece of shit!

>>56580870
>>56580850
Spotted the Intel damage control shills trying to red-herring-derail an embarrassing thread for their company.

>>56580887
friend of mine had one, only time i've ever used one
ran at "2GHz", apparently, and had a whopping 64k of L2 cache which puts celerons to shame, and struggled terribly playing mp3's in winamp

>>56580875
Yeah the Willimette Celeron is the same as the Willimette P4 which still makes the Willimette P4 faster than the Duron which was the original argument.

Now you're shifting goal posts to Tualatin Celerons.

>>56580518
you shut your whore mouth

>>56580894
cool shitpost

I currently use an intel cpu with amd gpu, amd cpus are total shit

File: stallman.jpg (127KB, 500x333px)

127KB, 500x333px

>>56580919

No I'm not, the Tualatins just happened in the citation like Pilate in the Creed, but the keyword there is "finally" (I even pointed that out for you). The point is: the first Willamettes were dogshit slower even than the Duron. The Celeron you mentioned didn't come up until much later.

>>56580962
This. I'm not a shill, I just use what works best.

>>56580962
>>56580975
₹2 have been deposited in your accounts ;^D

>>56580455
>implying 8350/20 wasn't the best price/ performance CPU on the market until recently

Got mine to 4.8 Niggahurtz, and it was the best experience I've had with any processor until very recently. I do love my 6700k, but damn, that 8350 was a little beast for awhile, there.

i5 6500 is a turd. upgrading to superior Zen asap.

>>56580491
Ever heard of the 9590? That shit melted a motherboard and boiled the liquid in a cooling loop.

>>56580597
>Pentium 3 had better performance

No shit, everyone knows that's why the i series was based off of the mobile variants of the Pentium III

>>56580758
>Replace "Bulldozer" with "Netburst" and "AMD" with "Intel", you shill.

P4 wasn't as bad as bulldozer was relative to competition at the time and it didn't cause Intel to lose 90% of their business' value on release.

Bulldozer killed AMD and has caused them to put out garbage products since it's release because they haven't had any money for R&D.

>>56580507
This. Mostly anything Via. Some exceptions but few.

File: Pentium 4 vs FX-9590.png (18KB, 730x363px) Image search: [Google]

18KB, 730x363px

>>56581017
The FX-9590 is faster than even the fastest Pentium 4.

>>56581105
It also destroys the computer it is in.

>>56581105
okay? and p4 is faster than the fastest amd k6

>>56580459
You're a fucking twat and also wrong

>>56580861
I had the Pentium D 805 and overclocked to over 9k Ghz. Was bretty good back then.

There are still plenty of P4s running out there in the world, too bad you can't say that for Durons har har.

File: 1331088514103.jpg (42KB, 504x520px) Image search: [Google]

42KB, 504x520px

i5-2500k.
At least mine is. Can't get past 4.3Ghz without 1.45 fucking volts.

File: 1464880364329.jpg (260KB, 1000x1000px)

260KB, 1000x1000px

>>56581045

>Intel releases product which is much slower and power-hungry that last generation
>business as usual for Intel

>AMD releases product which is just as fast as last generation
>AMD plummets and dies in the water

>th-there is no Intel monopoly guys, we swear!

You need to be wilfully blind not to see!

>>56581261
bulldozer was slower than 10h though.

>>56581209
In Eastern Europe people still game on Durons

https://www.youtube.com/watch?v=nUD44i5Aaq4

>>56581209
>There are still plenty of P4s running out there in the world
Unfortunately for our energy bills...

phenom ii x4 955 is the worst meme CPU ever made

>>56581280
that's why they still used them for cracking passwords.
stop spreading bullshit, faggots.

>>56580681
Thats not a northwood P4 you utter utter fucking terrible fucking utter cunt of a person

I always wonder what would happen if we pair run these old CPUs with DDR3 RAM. would it be just like the CPUs in phones, fast enough to play shitty games?

File: amd.jpg (20KB, 450x320px) Image search: [Google]

20KB, 450x320px

Anything with the letters "AMD" stamped on it.

>>56581438

I hope you're just pretending to be retarded.

http://en.wikipedia.org/wiki/List_of_Intel_Pentium_4_microprocessors#Northwood_.28130.C2.A0nm.29

File: Bulldozer.png (420KB, 545x902px) Image search: [Google]

420KB, 545x902px

>>56581280
Don't compare CPUs with the same number of "cores". Instead compare CPUs with the same number of floating point units like the FX-8170 and the Phenom II X4 940 BE. Then you will see that Bulldozer actually won.

>>56581474
Dumb Intel poster

File: via c6.jpg (2KB, 120x120px) Image search: [Google]

2KB, 120x120px

Why isn't anyone mentioning VIA. their c6 series was a torture. Most of my junior school, we had to learn basic and dos off it. It is still so bad that google can't find a decent picture.

>>56581182
9000ghz?
Not bad but my ti82 can oc well over 9000ghz.

>>56580758
At least i got paid for shitposting :^)

>>56581562
you didnt though

>>56580837

The Pentium 4 also couldn't scale and had heat issues.

>>56581619
Depends on which version Northwoods and late 6XX 64bit cedarmills were decent

Prescott was the super housefire.

>>56581105
It better damn fucking well be.
Coincidentally, the i7-6700k is faster than even the fastest Athlon 64 X2.

File: arm.png (8KB, 476x141px) Image search: [Google]

8KB, 476x141px

>>56581512

>Implying that I limit myself to only x86

>>56580507
I had a mini-ITX PC with a 1GHz VIA C3 back in 2010, and it was slow as shit. it was struggling to play youtube videos at 240p

>>56581484
Yea, don't compare CPUs based on sane criteria, compare them in a way that makes a bad architecture look okay.

Celerons.
Any celerons.
Celerons literally = cancer.

>>56580519
Used it for four years before getting a decent processor. despite it being shit i still like it cos mine was a fucking trooper, oced to 4.8ghz for most of its life, earned itself a cushy retirement in my storage pc.

>>56581855
the first ones that had no L2 cache were really slow, and the netburst ones were shitty, but apart from that, celerons aren't that bad.

the original AMD Phenom
>slow
>TLB bug

>>56580570
>The z520 in the Vaio P.
>Goddamn Vaio, release a new one with an M processor. I'll happily give you my first born.
This
A new Vaio P would be perfect

>>56581907
I've never had anything but a bad experience with a celerons but that's probably because the rest of the hardware it was embedded into were pieces of crap too, reflecting its budget market.

Still, pretty muc h the only celeron I'd even think about considering are the skylake-era which seem pretty well regarded.

I just hear the word 'celeron' and after so many years of bad experiences I hear 'cancer' back.

File: house-fire.jpg?format=1500w.jpg (643KB, 1034x689px) Image search: [Google]

643KB, 1034x689px

>>56581632
>tfw I had a 3GHz Prescott back in the day

>>56580425
probably the original 8120 or the late 9590

literally literally housefires

>>56580997
>i5 6500 is a turd
lolwut

>>56580425
Buy IBM Power 9.

>>56582109
But this is the best CPU ever made.

>>56582063
Lower single thread performance than a much cheaper i3-6100.

>>56580425
There were cpus that litterally killed ISAs because they were trying too hard and couldn't deliver. Things like SPARC, CELL and Itanium literally died because they were terrible. Anyone bringing up any x86 cpus are just consumer whores.

>>56582151
x86 is shit though.

>>56582190
>posted from an x86-x64 machine

>>56582145
interesting but i cant think of a single reason anyone would prefer a 6100 over a 6500

>>56582190
This x86 is pure trash. What has it done for us lately?

Athlon 64

>>56581105
Damn. And it only took them 5 years. AMD just keep waiting master race.

>>56581855
Celeron 300A was an overclocking beast. If you OC'd it the higher clock speed would easily make up for the lack of cache compared to the p2 300

P4 did suck, especially Prescott and pressler

File: Picture.png (31KB, 307x351px) Image search: [Google]

31KB, 307x351px

>>56582218
You sure about that?

>>56582190
So shit that it literally drives 90% of server and desktop computing in the world today.

>>56582350
>Something that was so shit it killed the entire ISA
PowerPC is dead, x86 killed it

>>56581531
there is mentioning of it

>>56582350
typical macfag

>>56582406
x86 winning doesn't make it better.

>>56582425
>lolxD Apple sux
Shouldn't you be in middle school?

>>56582406
>Something that was so shit it killed the entire ISA
Uh, no. Not even slightly true.

>>56582151
SPARC isn't dead and neither is Itanium.

>>56582406
But PowerPC64 live, and gonna kill amd64.
I mean IBM power 8 and power 9.
Also IBM developing quantum computers.

>>56582436
>hurr ima use a 12+ year old machine because fuck x86

#ThinkDifferent

>>56582436
shouldnt you be in a coffee shop or a bathhouse?

>>56582486
10 years old.
>hurr stop using your favorite laptop becuz I sez
Autism.

>>56582495
Never been to either. Nice shit stereotypes, quit spouting memes that are older than you are.

>>56582515
wow you're madder than the day you got your hiv test results.

>>56582549
>lolol he disagrees with me he's obviously irate
You couldn't be more of a twelve year old.

>>56582454
Both are only there to support the <1% of companies with legacy requirements.

>>56582483
POWER isn't PowerPC. Also IBM has pretty much abandoned their CPU business. POWER9 will be like POWER8, cool but not going to gain them any marketshare. All their AI and Quantum computing stuff will not make use of anything from POWER.

>>56582571
>POWER isn't PowerPC
It essentially is. The last POWER architecture that was actually POWER was POWER2. Everything after that's had a PowerPC µarch

>>56582559
>scream shout hiss pull hair i'm not angry guise waaahhhh

wishing you all the best and that antiretrovirals come down in price soon so you can upgrade to x86 and stop angrily shitposting.

>>56582613
I don't want any x86 trash. I'll be upgrading to the Talos workstation.

>>56582406
PowerPC may be dead, but POWER is still kicking.

>>56582515
>using outdated hardware for sentimental reasons
>calling anyone else autistic

>>56582657
>not using hardware just because it's old
>even if it works perfectly fine no ur not allowed to becuz I said
Autistic consumerist whore.

>>56582622
>Talos
figures you're a freetard as well

>>56582670
>running Mac OS X
>freetard
You're not very bright, are you?

>>56582571
>Both are only there to support the <1% of companies with legacy requirements.
Except they're still being actively developed, so you're retarded.

>>56582454
You're right, Itanium is UNDEAD. Zombie arch.

>>56582668
>perfectly well
>it can't play youtube videos without stuttering

File: intel.iapx_432_die_shot.102652367.lg.jpg (210KB, 500x446px) Image search: [Google]

210KB, 500x446px

>>56580645
Was about to suggest the 432

>>56582738
It can though, not that I ever use YouTube anyway.

>>56582145
The i5-3570K is $145 currently and faster than the i5-6500 in every way. This seems to be the best value CPU in the $100-$200 range.

http://www.ebay.com/itm/Intel-Core-i5-3570K-3-4GHz-Quad-Core-Processor-LGA1155-Ivy-Bridge-/322260210138

>>56582782
>huehuehue1

stop yer shilling

>>56582696
http://www.theinquirer.net/inquirer/news/1041690/intels-itanium-is-basically-dead
Even in 2005, people knew itanium is a dead end.
http://semiaccurate.com/2013/02/11/hp-and-intel-effectively-kill-off-itanium/
and in 2013 it was effectively dead.
http://www.itechpost.com/articles/21522/20160704/oracle-pay-3b-damages-hpe.htm
and when Oracle said it was dead and stopped development, it got sued!

Itanium is dead.


SPARC is still in active development but even fujisu is jumping ship:
http://www.nextplatform.com/2016/06/23/inside-japans-future-exaflops-arm-supercomputer/
>Fujitsu announced at the International Supercomputing Conference in June 2016 that its future exascale supercomputer will feature processors of its own design that implement the ARMv8 architecture. These processors will also implement extensions to the ARMv8 architecture equivalent to HPC-ACE2 that Fujitsu is developing with ARM Holdings

>>56582750
>I don't actually use my computer for anything so it's fine if it can't do anything

>>56582571
Power9 4-6 times more powerful than cannonlake, faggot.

>>56582905
Yea yea, nice links and all, but they still don't support what you said. They're still not dead.

>>56582926
>implying I only have one computer
Laptops aren't for power. Anything I do that's even remotely intensive is done remotely on a server.

>>56580580
Pentium D is just 2 Pentium 4's, so its just as shit as the Pentium and vastly inferior to the Athlons of the day

>>56582935
yet it still wont get them any marketshare against intel

>>56581952
Original phenoms were slow because of the TLB bug. They weren't that bad if you left the TLB enabled but if you did they could crash randomly.

>>56580459
Pentium 4 at its absolute best only managed to be mediocre compared to AMD's offerings. The Core2 series smoked Pentium 4 right out of the gate, It shifted the whole market practically over night.

All celeron crap

>People are actually defending the Pentium IV

It was literally dogshit the pentium III was far superior watt/performance wise. It is for a reason that Intel designed the pentium M around the P III and not the Pentium IV. Even Intel was aware that the Pentium IV was a bad chip and AMD was handing their ass to them with Athlon 64.

Stop being delusional morons. I prefer Intel to AMD at this point in time too but that doesn't change the fact that the Pentium IV was absolute horseshit (Eespecially the Prescott).

>>56583110
Let's not mention the incredible lengths Intel went to in order to try and stamp amd out if the market since they couldn't compete.

>>56582385
As does Windows.

>>56583018
The problem of IBM's CPUs it's the consumption. If they can fix it, surely will get more marketshare against Intel

>>56583334
The problem is they are too specialized and too powerful for most servers. That is why intel was able to beat them with economy of scale and eventually even compete at the high end performance spectrum.

IBM chips are monsters but the solutions that use them are expensive as fuck, especially if you consider the lifetime cost of locking yourself to POWER. That and the transition to GPGPU for extremely high end computing has made POWER's extra performance kind of pointless.

>>56583020
The biggest problem with the Phenom I other than the TLB bug was their low clock speeds (they could basically barely reach 3GHz).
The Phenom II was basically everything the Phenom I should have been.
> 45nm rather than 65nm
> ~10-15% increase in IPC over Phenom I
> Could reach mid-high 3GHz range, were way better at overclocking
> Compatible with DDR2 *AND* DDR3
> Could unlock dual cores into quad cores

The Phenom II's were basically trading blows with the Core 2's.

File: CELD351-R-unit.jpg (33KB, 300x300px) Image search: [Google]

33KB, 300x300px

The processor in my Power Mac G4 was better than this.

>>56580455
If you want to a faildozer, you want the 8150 or the ridiculously clocked 9590
>>56580595
Bulldozer isn't even close to the scale of damage that Netburst did
>>56580645
>random commieshit (slav or chink) that nobody here ever heard about
A lot of those are clones of the best DEC/Intel had back then, excluding the home grown Elbrus that could beat the TOTL Xeon's
>>56581045
>P4 wasn't as bad as bulldozer was relative to competition at the time and it didn't cause Intel to lose 90% of their business' value on release.
AMD didn't lose much more value than what it already lost when BD was released
Intel was forced to bribe every OEM and risk a court ruling in the EU that they eventually lost
Netburst was an actual disaster
>>56580816
tbqh all the E series are awful, but usable and pretty much what they're supposed to be
>>56581209
P4's are excellent BOINC-running space heaters
>>56581531
VIA isn't relevant to shills doing damage control against Zen
It isn't even going to compete straight ahead with Intel's HEDT, I don't know why they're so desperate shilling here
>>56582151
>Itanium
tbqh the first Itanium is easily the worst CPU ever made, so much hype, so many manufacturers jumping on board and abandoning mature products, and so much disappointment
It even feels like it was a plan to kill MIPS/SPARC/Alpha and lead to the x86 dominance
>>56582406
Apple killed PowerPC to get PASemi and their in-house SoC's

>>56580425
at least the SSE unit was bitchin'

Anything bay trail

bayt rail
bait rayl

With VP9, even Youtube lags, UEFI fuckups everywhere - no matter what brand, bad linux optimization, bad windows perfomance, worst case scenarios when you wish to use os x.

worst purchases ever.

>>56583954
This is a worst CPU ever thread, not a CPU I don't like thread. Bay Trail's great. Go use an Atom N270 for a week then try telling me Bay Trail sucks.

>>56583977
God the first Atom's where truly awful
They're still awful but at least they don't suck like they used to do

File: images.jpg (8KB, 225x225px)

8KB, 225x225px

>>56580425
AMD C-50

>>56584008
New Atoms are great, iirc they're on par with Core2 Duos now. That's not even remotely bad, it's just not amazing. Which is exactly what you should expect from a low power processor.

>>56580791
>all modern Intel processors use the objectively superior Core architecture

they don't use that in the current and modern atom line up (pentium and celeron included) are based on old ass P5 and P54C architecture with out of order execution additions.
They are basically based on designs first seen in 1993.

People pay up to $500 for that nowadays, TOP KEK

They are utter shit, it's only the extreme die shrinks that made them somewhat useful for some shit devices. Other than that, they are complete trash.

So stop lieing you fucking nigger.

>>56583977
>Bay Trail's great
Oh, yeah you convinced me, you are right.

Buy one, anon! They are super affordable!

>>56584058
Core was also based on designs from 1993 and earlier. The atoms have had extemely large changes with every generation until intel abandoned them. The quad core ones are more than powerful enough for every day tasks.

>>56584113
The fuck are you on about?

>>56584124
Buy it, there's nothing wrong with it.

They are
GREAT
R
E
A
T

Best CPU evurrrrr

>>56584143
Would you mind explaining why you started going full retard?

>>56584153
I didn't go full retard, I just hope for you, that you can get yourself one of these. Because they are so goood, mmmmhmmm good cpu.

>>56584176
You definitely did, and I have one.

>>56580425
Is Cyrix or Winchip still around? Didn't sell many of those, Celeron if it wasn't for the few overclockable versions. A lot of the SX chips seemed meh (386sx, 486sx)

>>56584197
Cyrix was acquired by VIA.

>>56584208
Looks like Centaur Technology (WinChip) was acquired by VIA as well. I forget WinChip even existed

>>56581474
Athlon XP was beast, if you bought Intel back then you were retarded. (Kinda wish AMD would do that again, Intel gets lazy and just sets themselves to cruise control)

>>56583977
Bay Trail? More like Gay Trail.

>>56582385
>>56583233
Dang, True though, Lots of apple phones and beat headsets out there, and honey booboo, MLP, Ford, and justin bieber are popular too, doesn't really make it right.

what does a CPU do anyway?

>>56584401
It processes things. It's the central processor

File: puma-on-rock1363319673958.jpg (54KB, 700x600px) Image search: [Google]

54KB, 700x600px

>>56580507
Hey! You think it couldn't handle windows 98 good enough?!?

>>56581105
>Physical cores:4
This angers my jingles

Baytrail is decent.

Heck I finally got rid of my Acer Aspire One which had the first gen single core Atom N270.

My server still runs on N330+ION and it can do 1080P H264 video via CUDA lol, has 2*2GB DDR3 and running on a SSD. up to 5,541,110 visitors per year.

Probably going to pick up something even more power efficient down the line, the N3150 based systems pull 36W~60W at load depending on configuration, that's like running a lightbulb or two all year round.

>>56584594
you can probably switch it for a pi if you want power efficiency

>>56580455

That is probably legitimately the most chip you can get for the least money.

>>56584594
I've got one of those niggers running my firewall. I wish I had gotten one of those purpose built firewall/router boards but they were more expensive

>>56584647
Older xeons are better but the boards cost a bit more if you want 2 sockets

File: 20160603_184827.jpg (1003KB, 2560x1920px) Image search: [Google]

1003KB, 2560x1920px

>>56580425
AMD E-350
Worst piece of shit ever created. These motherfuckers even tried to mount this CPU into 17 inch laptops. Holy fuck.

>>56584611
>average of 10 visitors per second
>raspberry pi

>>56584686
I had a E-350, it was great for normal use on ubuntu. Maybe you are just expecting too much for a low end CPU.

>>56584703
This, my school laptop had an E-350 and it was pretty decent.

>>56584701
You think a pi 2 or pi 3 can't handle that? The Pi 3 is more powerful than the N330 in pretty much every way except memory, a static web server would need much less.

>>56584611
>shitty 100mbs ethernet bus
>over USB 2.0
>which is also occupied with the shitty SD card based storage
>handling millions of requests at a time

>>56584703
E-450 was pretty desu, especially after undervolting.
E-350 was absolutely shitty.

>>56584014
Holy shit this.

Graphics performance was alright but have fun trying to install something. Fuck it took me like 2 hours to install MS office on it.

Netbooks were doomed to fail from the beginning.

>>56584724
It just seems like a bad idea to use a server with USB ethernet for a site that has a pretty decent amount of traffic.

>>56584686
E-350 was alright for what it was intended for, namely being a netbook processor that didn't choke on hi-def videos like its contemporary Atoms did.

Then laptop makers went full retard and used it in super cheap, crappy laptops of all sizes.

>>56584727
The E450 was just an slightly OC'ed E350 with small tweaks to keep the power consumption the same. The difference is less than 10%.

Atom n145 was utter wank.

To quote Blackadder, it ran like "an asthmatic ant with a heavy load of shopping".

>>56584611
Too lazy to use a Pi
I run a website, a FTP, seedbox, also occasionally used as a HTPC.

Website hits near 80 visitors simultaneously on average daily, with a record of 375 simultaneous visitors.

Killed 2 hdds for no apparent reason so it's on SSD now.

>>56580507
I used a router/file server with an Eden based mini-ITX board for several years in the mid-2000s. It wasn't that bad considering that no other fanless x86 CPUs existed at the time aside from Geode.

>>56584029
The new atom chipsets have shit Linux drivers.

>>56584735
Didn't help that Linux had shit amd support

>>56585110
Where did I say otherwise? I'm not talking about Linux support, I'm talking about processing power.

The N3150 @ 1.6~2.1ghz 6W is neck to neck with the C2D E6700 @ 2.7ghz 65W

It's pretty amazing how they were able to drop the power consumption so much during the 5 years between the E6700 2010 and the N3150 2015

>>56581213
implying thats bad.
end you're's rurself

>>56584666

Can't argue with those trips. Don't want to upset Satan.

>>56585459
>2010
wut? the E6700 is a 10 year old processor

>>56585811
Got the Pentium E6700 and the C2D E6700 release dates mixed up.

AMD E1 1200

Intel Atom N270 1.6GHz.

File: KL_Intel_Celeron_Covington.jpg (842KB, 3300x2000px) Image search: [Google]

842KB, 3300x2000px

Covington Celeron was essentially a 266 MHz Pentium II manufactured without any secondary cache at all.[7] Covington also shared the 80523 product code of Deschutes. Although clocked at 266 or 300 MHz (frequencies 33 or 66 MHz higher than the desktop version of the Pentium w/MMX), the cacheless Celerons were a good deal slower than the parts they were designed to replace.[5] Substantial numbers were sold on first release, largely on the strength of the Intel name, but the Celeron quickly achieved a poor reputation both in the trade press and among computer professionals.[8] The initial market interest faded rapidly in the face of its poor performance and with sales at a very low level, Intel felt obliged to develop a substantially faster replacement as soon as possible. Nevertheless, the first Celerons were quite popular among some overclockers, for their flexible overclockability and reasonable price.[5] Covington was only manufactured in Slot 1 SEPP format.

File: oh.png (42KB, 522x420px) Image search: [Google]

42KB, 522x420px

i7 6950X

Literally no point over an 8 core for the increase in price

Cyrix 5x86
Transmeta Crusoe
Intel i432/8800
Original Pentium 60/66Mhz
All Itanium shit
Pentium 4 willamette
Amd Bulldozer & co

>>56585965
fuck man I don't know how I lived with that CPU for as long as I did. trying to play games on intel GMA with that anemic single core with HT

>>56586036
Yeah they were a bitch to use, I had it in the Acer aspire one bakc in 2009 when i was 14, decent little linux machine/hackintosh too, due to the GMA 950 being compatible with everything.

File: lqb.jpg (4KB, 200x200px) Image search: [Google]

4KB, 200x200px

>>56580455

File: AAO.png (41KB, 745x500px) Image search: [Google]

41KB, 745x500px

>>56586052
>>56586036

>>56586072
Oh shit, that's sweet, I still wish I had mine, sold it to Gazelle in 2010 back when they still paid money for computers.

File: volari.jpg (95KB, 714x538px) Image search: [Google]

95KB, 714x538px

It.... hurts to live...

File: 0e7.jpg (32KB, 600x619px) Image search: [Google]

32KB, 600x619px

>>56586145
>XGI Volari

>>56585977
Covington wasn't that terrible for the price once you OCd it to 400 MHz.

Of course, it has nothing on the glorious 300A.

File: wizardry.jpg (121KB, 638x582px) Image search: [Google]

121KB, 638x582px

>>56586145
>XGI Volari

File: Untitled1.png (42KB, 674x532px) Image search: [Google]

42KB, 674x532px

>>56586072
check my DDR2 freqs

>>56586365
how is that even ddr2?

>>56582949
I've been using a PowerBook G4 12' with an 867MHz G4 for a few months as my main machine with no issues with a bit of tweaking and regular maintenance etc with Onyx it ran fine. Even played YouTube videos at 240p which was fine by me.

>>56586438
It's super advanced Asus powersaving faggotry. It took me a few tries to find a 2GB stick that doesn't shit itself at this frequency.

>>56580425
The older P4 era celerons with 128k of cache, at least the worst i've ever had to daily drive while still being a current processor.

Jesus christ that thing was slow.

>>56584978

Geode

:: shudder ::

>>56586595
Geode wasn't bad for embedded purposes.

If you were insane enough to run Windows on it however...

>>56586145
Holy shit that's amazing.
Also, that IDE drive is cooking.

>>56580570

I want this so much

I have a P530H and still deal with the sluggishness just because I love the form factor so much. If anyone who does laptop design is reading this please make a solid 8" notebook again, I will buy it

File: titfaces.png (748KB, 1282x483px) Image search: [Google]

748KB, 1282x483px

>>56580455
Bullshit.

>>56580459
>>56580550
Pentium D wasn't all that bad. It was quite cheap for a dualcore in times when that was very premium. Nobody here recalls Pentium D805, you fucking teenagers?

>>56580507
Yeah, they are pretty weak, but on the other hand they were fairly power-frugal in time when that was no yet in or common. SO it wasn't without good qualities.

>>56580540
Again, original Atom was very low power, it didn't really have competition.

>>56585977
First Celeron (Covington, without L2) was fairly shit, but still have have been stronger than Pentium. I have the 266 model somewhere, could test against 333 if I had working board and free time. Alas.

>>56582232
Indeed. Very well said.

>>56586145
Whoa, nice, nice machine.

Hard to pick really "absolutely terirble CPU", IMHO. Lots of them were worse than other thing, but that is not particularly terrible thing, is it. From recent time, I think I would pick the utterly terrible Broadcom BCM2835 SoC from original RPi. It was hopelessly slow and obsolete and yet so many unfortunate people ended up with that crap in the RPi/ARM hype. It had no excuse existing when Cortex cores with Neon already existed. Mentioning that, Tegra 2 was also quite bad, because fukcing Nvidie decided to not include Neon SIMD. WRYYYYYY

>>56586852
>original Atom was very low power

...except when it was paired with a 25W chipset on literally every ITX board.

>>56586917
Yeah, that was so shitty.

>>56582571
>Both are only there to support the <1% of companies with legacy requirements.
No they fucking aren't, Oracle is still maintaining a ton of software and Solaris on SPARC, developing new high-end SPARC chips to put in brand new high-end SPARC systems. That isn't dead.

Now, Itanium IS dying, HP is pretty clearly showing interest in moving HP-UX over to x86 (like they're already doing with VMS) and Intel has had that shit on the backburner forever, though it isn't totally dead yet as long as it's the only thing that runs HP-UX and you can still buy Itanium systems from their original vendors.

>>56582905
>b-but a bunch of clickbait sites said it was "dead" in the title!
Why does that mean anything? They said x86 was dead when RISC chips were kicking everyone's ass in the '90s, and they said the PC was dead when tablets and phones started picking up. Do you think either of those notions were true?

The Itanium was "dead" in a way that it didn't nearly have the adoption Intel and HP had wanted, it was a total failure in its ultimate goal of superseding x86, but it still kept trucking along, developers continued to develop for it, you can still buy them. That's not dead.

>SPARC is still in active development but even fujisu is jumping ship:
It's an interesting article, but it doesn't really mean that SPARC is dead, supercomputing isn't their only application (it never was) and I'm honestly surprised it's even still used there.

>>56581213
>i5-2500k.
came here to post this, overrated shit!

>>56580425
The original Willamette p4, the Prescott p4, the Pentium D, 8150, all of amd apus

>>56586852
pentium D throttled hard and had terrible performance, they weren't even really faster than the pentium 4 with ht

>>56580829
fedoratippingneckbeard.jpg

Remember when they put desktop P4s in laptops? They were probably great space heaters though.

>>56588162
I think I have one of those laptops floating around here somewhere.

>>56587875
>worst cpu ever thread
>i2500k

Cool story bro.

>>56587904
Wrong. If you had decent board, they were usable fine. And that D805 was used for OCing to 4,0 GHz because it had 533MHz FSB. It was great deal in its time. Sure. Athlon 64 X2 3800 was better, but also much more expensive.

>>56586033
>FX
this meme is getting seriously stale
if you use all 8 threads the FX8k are unmatched for dirty cheap work stations. yes the i7 is better but its also 4x the price.

6600k
haven't oc'd yet but it still chokes on massive dicks when I have too many Breitbart and RT tabs open because of all the fucking flash ads spam

Why can't a modern CPU handle Adobe's kikery yet??

>>56580459
is this bait? it's the other way around, P4 is terrible, Pentium D was decen and better value
>>56580455
1/10 made me reply, 8350 is #1 on CPUID for highest overclock, great chip

>>56582304
I had prescott. It was sheit.

File: cpuz.png (41KB, 813x402px) Image search: [Google]

41KB, 813x402px

this thing is shit

File: bench.png (884KB, 1024x574px) Image search: [Google]

884KB, 1024x574px

>>56589068

>>56589068
Course its shit, its an old Atom.

>>56589078
How long did that take to render?

>>56589092
like half an hour

>>56589106
Which in turn begs the question: why the fuck are you still running an Atom? Pretty sure there are pentium 2s that are faster than that shit.

>>56589135
Its a relic of times past. My actual computer is much more powerful.

>>56580425
fucking atom n270. fucking piece of shit.

considering buying a netbook with a pentium n3700. opinions?

>>56589240
n3700 is fine for a netbook

>>56589253
i've read it lags on some websites. how long do you think it will be useful?

for reference, that piece of shit atom is still somewhat usable today and it has 7 years

>>56589285
I have one and it does lag in

>tfw my only laptop's an Atom N270 netbook

File: amde.png (9KB, 920x500px) Image search: [Google]

9KB, 920x500px

AMD E Series

6850K

>>56589576
poor fag.

the 6840k is the CPU you buy when you need allot of PCI-E lanes, not allot of CPU cores.

>>56582483
I fucking hope so
>>56582350
Anon, you got an MDD?
I'd post from my Mini G4, but I don't feel like powering it on
>>56589240
My atom n270 plays Halo CE multiplayer, can't do anything else really

File: 1471215762605.png (437KB, 2586x1567px) Image search: [Google]

437KB, 2586x1567px

>>56580758
Those poor 939 owners tho

i guess they got their ride

>>56589649
>Anon, you got an MDD?
That's a PowerBook. I do have a graphite PowerMac G4 though (along with like 10 other PowerPC Macs)

>>56589675
Anon, I love you
I've got 40 G5s and 2 G4s
And 3 Powerbook G4s
And 1 G5 iMac
And 1 G4 Mini

I think I'm insane

>>56580425
>Every CISC chip on the planet

>>56589786
Underrated post

File: Screen Shot 2016-09-13 at 8.28.26 PM.png (119KB, 862x780px) Image search: [Google]

119KB, 862x780px

>>56589769
>40 G5s
Jelly as fuck. My only PowerMac G5 (dp 2GHz) just took a shit, it sucks but now I've got an excuse to upgrade to a G5 quad.

pentium 1 the first run with the FDIV bug

>>56589870
I'm so sorry. Do you know which part of it took a shit?

I accidentally bricked my DC 2GHz by changing OF variables

Oops

>>56589909
Not sure, it works for a day or two and just stops, then won't turn on for like a month.
>I accidentally bricked my DC 2GHz by changing OF variables
You can't just clear the NVRAM?

>>56580562
>Ctrl+F Cyrix

I'm shocked I wasn't collecting my social security checks by the time I finished installing Windows 95 on one.

>>56588162
I have a stack of those (Thinkpad Gs) on my bookshelf

>tfw I unironically used one of them as my DD in ~2011 because the battery life was actually pretty decent in comparison to my aging core2 T60
shit didn't even get hot either since they had a thicc ass solid copper sink and big vents

>>56589940
>Not sure, it works for a day or two and just stops, then won't turn on for like a month.
Seems to be the powersupply

>You can't just clear the NVRAM?
I've tried. I've even tried to remove the battery for a day. I think I need to manually reset the PRAM by finding the physical chip and sending the reset signal
cmd opt PR doesn't work either

I'd love to make a supercomputer using all 37 working if possible. The old DP 2.5s were all leaking
AltiVec was great

>>56580455
I own a 8320 and I really like it. OC to 4.4Ghz no sweat.
It cost a third of the price of an i7 and its awesome for rendering/multiprocessing.
Why so salty? I am hoping to buy an i7 in the next couple of weeks though.

>>56590023
I was looking into upgrading the PSU anyway, might as well try that.
>AltiVec was great
Still is, it's still present in POWER, just known as VMX because AltiVec's a Motorola trademark. A PowerMac cluster would be pretty sweet though

>>56590068
Anon, you on one of /g/'s discord servers?

>>56590086
No, I don't really use anything like that

>>56590104
Damn, it would've been nice to keep in touch. Not everyday you meet someone that likes PowerPC as much as you

>>56590130
you can hit me up at [email protected] if you don't mind using email

>>56583334
>If they can fix it, surely will get more marketshare against Intel
The /g/ collective really needs to stop treating POWER, SPARC et al. and the systems based on them as volume competitors to x86 servers, they simply aren't, and never will be. IBM isn't selling you POWER hardware to fill your entire datacenter and do absolutely everything, they're selling you a few high-end systems for a couple racks in the corner to augment certain services that benefit from high performance, availability, whatever while x86 boxes handle the "grunt" work.

>>56583437
>That and the transition to GPGPU for extremely high end computing has made POWER's extra performance kind of pointless.
POWER's modern advantages go a little beyond raw fp performance, other than the fact that it's the only thing out there capable of running AIX, there are various enhancements like hardware virtualization that are supposedly still ahead of even Intel, though I couldn't really go into details, maybe I can find the article I skimmed for it.

>>56590150
>powerpcfag
that email is fucking beautiful

>>56590196
kek, it's my old trip

>>56584790
It's funny how a lot of programs had special modes to run on atom systems with the justification of atom being so fucking slow
>>56585459
>The N3150 @ 1.6~2.1ghz 6W is neck to neck with the C2D E6700 @ 2.7ghz 65W
It's not, it runs about as fast as a E4300 ignoring usage of newer SIMD instructions
>>56586917
>dem' 945G chipsets on netbooks
JUST

>>56582046
I haz 8120 on a ghetto cooler and it runs fine. Haven't had anything apart from rendering make it hit 100%. Bought it for £95 when the 2500k was £170 back in the day

File: 1470683059569.jpg (51KB, 599x593px) Image search: [Google]

51KB, 599x593px

>>56581021

this is such an embarrassing post

File: Screen Shot 2016-09-13 at 9.58.26 PM.png (284KB, 1032x1152px) Image search: [Google]

284KB, 1032x1152px

>>56590214
I sent the email
>>56590192
The PS3 does 25% more FLOPS than a PS4
Just wanted to get that out there

Also, I just want to say that 90nm in it's entirety was a cooling nightmare
>G5
>PS3
>Xbox 360
>Nvidia 8800GTS
>Pentium 4

>>56589786
>>56589809
>muh CISC
The hilarious thing about these kinds of posts is that none of you can actually back up this claim with anything meaningful other than pointing at 25 year old synthetic benchmarks and masturbatory anecdotes, since the only people who still believe in RISC vs. CISC have no fucking idea about even basic CPU design.

Vote for Celeron mendocino

File: 800px-KL_Cyrix_6x86MX.jpg (90KB, 800x836px) Image search: [Google]

90KB, 800x836px

Lots of newfags here.

This glamorous piece of shit came basically overvolted and overclocked out of the box, so it ran burning hot and crashed constantly under any real load.

>>56590313
>The PS3 does 25% more FLOPS than a PS4
That may be so, but these kinds of nebulous comparisons like "operations per second" are not gospel, especially between vastly disparate architectures that do different things in different ways, not to mention the many other things that determine system performance like bus design, memory bandwidth, storage speed and so on.

>>56590451
I know enough that now they're the same thing, as most stuff is done through microcode. The problem now is that Intel has, what, 40 years of backwards compatibility? Isn't there like 20 different versions of the MOV instruction? Literally any new ISA would be so much better. Why do you think we're here stroking our dicks in wait of RISC-V?
Also we all remember Intel's shitty quality control

>>56590539
You mean Covington? Mendocinos could rape full-on Pentium IIs in applications that made effective use of that tiny but blazingly fast on-chip cache

>>56590583
You're not wrong, but you'd expect a console 8 years later to be much more powerful in every single aspect. That isn't what happened

>>56590625
blame muh parallelization meme

>>56590638
Both are parallel, no?
PS3 had one PPE and 7 SPEs
PS4 has 8 cores
Wouldn't that make the PS4 more parallel?

>>56590677
the PS4 uses what is basically an APU

>>56585977
ok that cpu was garbage

>>56590691
But it's still a full CPU, it just has an integrated GPU

>>56590586
>Isn't there like 20 different versions of the MOV instruction?
Don't think so, but even if it was the case, what is your point? How do these compatibility constraints hamper you or the work you're doing in any way other than violating some idealistic code of hardware-morals you hold so closely?

>Literally any new ISA would be so much better
Like the Itanium? Or the Alpha? Or any of the other different attempts to usurp x86 it on the desktop in one way or another that failed miserably because they were total shit no matter how "clean" and "elegant" they were?

>Why do you think we're here stroking our dicks in wait of RISC-V?
Because of its openness? AKA the entire point of its existence? Nobody's looking towards that shit as a performer.

>>56590807
>Or the Alpha?
Alpha was doing great until Intel memed Compaq into Itanium

>>56590807
>>56590451

You obviously don't write compilers.

>>56590839
Not as an x86 replacement, Alpha NT4 systems were absolutely useless and had next to no software, and the architecture as a whole was pretty much coasting on its past accomplishments by the time the buyout rolled around and all Compaq and HP did was continue to re-hash the same old pre-buyout design the same way SGI rehashed the R10K for over a decade.

>>56590854
Why should I give a shit about how easy compiler development is on a given platform if I'm not writing compilers?

>>56590925

Because compilers compile the software you run. If they do a good job, your software is better optimized. If it isn't, then it's dogshit.

>>56590807
>Don't think so, but even if it was the case, what is your point? How do these compatibility constraints hamper you or the work you're doing in any way other than violating some idealistic code of hardware-morals you hold so closely?
Look at what happened with Skylake and the SSE bug. More assembly instructions, more things to fine tune, and Intel is the one who makes a new architecture every year. Only now did they decide to stay on an "optimization" step from their usual tick-tock
>Like the Itanium?
Intel needed to do better to phase out x86. They were competing with themselves
>Or the Alpha?
Some ISAs are bad
>Or any of the other different attempts to usurp x86 it on the desktop in one way or another that failed miserably because they were total shit no matter how "clean" and "elegant" they were?
Had Apple stayed on PowerPC when they released the iPhone, I guarantee you we would be seeing some real competition against Intel

>Because of its openness? AKA the entire point of its existence? Nobody's looking towards that shit as a performer.
I'm pretty sure a lot of people are looking at it as a competitor to x86

PIV HT
>mfw at a gig, the cheapass bosses kept using this shitters
>mfw employees bark at me because their PC's are slow

Took me months to prove to those stubborn /&(&%/(/&(/ that their aging processor was at fault, not me.

So I have more reasons to hate P IV
Their shittiness, and making dumb users barking at me unnecessarily

>>56581855
Haswell celerons are good tho, they beat any core2quad core in everything besides multi core performance.

>>56590941
Yeah? And what's your point, that all x86 compilers are "dogshit" by virtue of the CISC meme and compiler writers are too dumb to read documentation to figure it out?

>>56591010
>Look at what happened with Skylake and the SSE bug
That was the first mentionable bug in... how many years, and what makes you think "simpler" designs are somehow immune to them?
>Intel needed to do better to phase out x86
Like what, better x86 emulation, i.e working all of your dreaded "bloat" right back into the new architecture? They never had a chance, because nobody gives a fuck about 5% more performance if it can't actually run any software.

>Had Apple stayed on PowerPC when they released the iPhone, I guarantee you we would be seeing some real competition against Intel
PowerPC was braindead garbage even when they were still using it, the later G4 and G5 chips were especially abysmal.

>I'm pretty sure a lot of people are looking at it as a competitor to x86
People look at ARM that way too, doesn't make them right. Most RISC-V proponents here certainly don't, it's about an open, secure platform, not a raw performance advantage that that shit is certainly never going to hit for a long time.

>>56591143
>That was the first mentionable bug in... how many years
Like 2? Maybe less.

>>56591165
What was the last one? Never heard anything of it.

>>56591143

>Yeah? And what's your point

I'll just leave this piece of information here: LLVM-PPC is about ~36,000 LOC big. LLVM-MIPS is about ~35,000 LOC big. LLVM-x86 is over twice both those figures big at about ~74,000 LOC.

File: cpu.png (33KB, 407x403px) Image search: [Google]

33KB, 407x403px

Pentium D 945 master race reporting in

>>56580540
>The first celeron

Those were great. Overclocked like champs. Like, over 50% easily.

File: passmark.png (39KB, 736x625px) Image search: [Google]

39KB, 736x625px

>>56590235

File: cpu2.png (22KB, 407x403px) Image search: [Google]

22KB, 407x403px

>>56591265
benchmark

>>56591143
>That was the first mentionable bug in... how many years
0-1 year
See the Low Activity hang in Skylake
>Like what, better x86 emulation, i.e working all of your dreaded "bloat" right back into the new architecture? They never had a chance, because nobody gives a fuck about 5% more performance if it can't actually run any software.
They had no reason to try to make Itanium better.
They could've made x86 emulation much, much better
>PowerPC was braindead garbage even when they were still using it, the later G4 and G5 chips were especially abysmal.
I mean, sure, the G5s were ridiculous to cool, but by no means garbage. G4s even less so.
The Pentium 4s of the same era weren't better by any means.
>People look at ARM that way too, doesn't make them right.
Fair enough, however
>Most RISC-V proponents here certainly don't, it's about an open, secure platform, not a raw performance advantage that that shit is certainly never going to hit for a long time.
That "openness" will allow it to become a competitor to x86. If manufacturers can make it cheap and powerful enough, it can compete

>>56591256
If you could show that performance difference of LLVM, that would be great

>>56591256
Who gives a fuck?

>>56591232
E5-2670 first stepping had broken VT-d.

>>56591265
I have a 925 lel

>>56591379
I also have a 920 and two 805's I got all of them including the 945 for 1$ each

>>56591256
36,000 lines of code is still plenty of room to fuck up and introduce bugs.

>>56591306
>See the Low Activity hang in Skylake
I remember that one now, it was patched pretty much immediately. Big whoop.

>They had no reason to try to make Itanium better.
Vendor lock-in is even more lucrative in the high end, which they surely did have a good reason to force the Itanium into. I don't really see how this fits in to the current discussion, though, especially considering a more diverse hardware market would be even worse in this regard. RISC vendors were incredibly stagnant especially on the desktop, you could blame it on complacency while they waited for the Itanium, but even companies that never shifted over like Sun were really just not improving their hardware either, definitely not at Intel's pace.

>I mean, sure, the G5s were ridiculous to cool, but by no means garbage. G4s even less so.
They did their job, but that was it. Linus Torvalds has a rather interesting rant about shortcomings of the PPC platform (and others) here that I found interesting to read http://yarchive.net/comp/linux/x86.html

>The Pentium 4s of the same era weren't better by any means.
On paper? Yes. But performance-wise? No, take it from someone who still uses PPC systems, they /were/ better. Even my dual-socket 1 GHz Power Mac slows to a crawl doing things my single-socket Northwood 2.2 workstation will do without a fuss. I can attribute some of this to slow RAM, but I feel you can only take it so far.

>That "openness" will allow it to become a competitor to x86. If manufacturers can make it cheap and powerful enough, it can compete
Maybe in some arenas if anyone ever does something with it, but there's no way in hell it will ever seriously challenge it. Hell, just look at the open SPARC and POWER chips that have gone virtually nowhere outside of a few expensive and unobtainable showroom pieces.

>>56591978
Anon, first of all, I'd like to thank you that I could have a serious discussion with you

To continue:
>I remember that one now, it was patched pretty much immediately. Big whoop.
As far as I remember, the patch was to never let it go into low activity mode. The point is, shit as retarded as low activity would cause it to fuck up. There's no quality control


>Vendor lock-in is even more lucrative in the high end...
I can't really argue with that. What I was trying to argue earlier was that because Intel was already in the lead with x86, there wasn't too much going for Itanium. But when I read the article, I saw that they sunk several billion into it, so clearly I was wrong.
>mfw Intel wanted to get to 5GHz with Pentium

>They did their job, but that was it. Linus Torvalds has a rather interesting rant about shortcomings of the PPC platform (and others) here that I found interesting to read http://yarchive.net/comp/linux/x86.html
It was a good read, thank you

>On paper? Yes. But performance-wise?...
If you don't mind me asking, what programs were you running? I work with PowerPCs too, although typically the later generation ones, such as the PowerBook and Mac Mini G4s and PowerMac G5s. I've found the PowerPCs to be much more reliable

>Maybe in some arenas if anyone ever does something with it...
If I recall correctly, SPARC hasn't tried to challenge the consumer area. Not sure how it's been doing in the server area recently
POWER has been expanding with POWER8. Hell, Google has been porting over many of their servers to POWER and ARM, away from x86. It's gaining marketshare in the server space. Even though it's one company, it's one hell of a company to be getting support from

excuse the ellipses, I was running out of characters

File: cyrax-mortal-kombat-dlc.gif (41KB, 440x396px) Image search: [Google]

41KB, 440x396px

*****Hey, everybody, heads-the-fuck-up...*****

File: CyrixCx486DLC-25GP.jpg (37KB, 456x445px) Image search: [Google]

37KB, 456x445px

...out of the way, runners-up, it's understood that the word "DLC" has a certain mystical creepy quality, so allow me set in the chill in preparation for my brother...

>>56592315
>Anon, first of all, I'd like to thank you that I could have a serious discussion with you
Same, it's always fun to discuss. Excuse my shitposting mood.

> Intel was already in the lead with x86, there wasn't too much going for Itanium. But when I read the article, I saw that they sunk several billion into it, so clearly I was wrong.
Yeah, HP and Intel threw a ton of money into it and hyped it up unbelievably, it's actually kind of sad that Intel ditched it so quickly. It was shit but not terrible in its high-end niche, but when I really think about it, I don't think it did really stand a chance since I don't think the integer performance was all that great.

>If you don't mind me asking, what programs were you running? I work with PowerPCs too, although typically the later generation ones, such as the PowerBook and Mac Mini G4s and PowerMac G5s. I've found the PowerPCs to be much more reliable
It's really web surfing that kills them the most, even with an AltiVec-optimized browser like TFF. I'll blame a lot of it on PC133 RAM being as shit as it is but hell, it's still sketchy when even a lowly Pentium 4 1300 is running circles around it without breaking a sweat. Still waiting to get my hands on a P4 with SDRAM to truly put it to the test though.

G5s are definitely a different story, but I think by then they really were just starting to lag behind. They weren't total shit, but they could have done better.

>If I recall correctly, SPARC hasn't tried to challenge the consumer area. Not sure how it's been doing in the server area recently
Sun released a lot of low-cost boxes like the Ultra 5/10 here and there, but no, it definitely hasn't. Still, POWER really hasn't either but it's gotten a ton of attention, I'd probably choose it over SPARC anyway for desktop use since SPARC is more optimized for extreme parallelism and other things that don't really lend themselves to that use case as readily.

File: CX486SLCe-33MP small.jpg (30KB, 422x422px)

30KB, 422x422px

*****...alrgiht, enough of you jokes that could simply just hop out of the socket if you were really that bad! Doesn't look like you're gonna try and contest this title against one that was so brutal as to have to be soldered in, one who was so nasty as to have been replaced right off the bat ... ... ... if the poor suckers actually had that option, MUWAHAHAHAHA!*****

File: Cyrix-6x86L-PR200+GP-CPU-150MHz-3,3V-Canada-1995.jpg (613KB, 1700x3037px)

613KB, 1700x3037px

>>56592546
C-c-c-combo breaker!!

>>56592573
You can shit on the DLC or 6x86 all you want (but those weren't really all that bad) but THIS is the real piece of shit, at least in that instead of using it as a 386SX upgrade like it was intended to be they used it as a cynical marketing tool to mark up shitboxes based on the 486 name alone.

At least the mainboards were cute as fuck, literally nothing but a CPU, NPU socket and a single-chip chipset.

>>56580425
Pentium 4

>>56592546
>Same, it's always fun to discuss. Excuse my shitposting mood.
Is k senpai

>Yeah, HP and Intel threw a ton of money into it and hyped it up unbelievably, it's actually kind of sad that Intel ditched it so quickly. It was shit but not terrible in its high-end niche, but when I really think about it, I don't think it did really stand a chance since I don't think the integer performance was all that great.
I'm pretty sure you're right on the shitty integer performance

>It's really web surfing that kills them the most, even with an AltiVec-optimized browser like TFF. I'll blame a lot of it on PC133 RAM being as shit as it is but hell, it's still sketchy when even a lowly Pentium 4 1300 is running circles around it without breaking a sweat. Still waiting to get my hands on a P4 with SDRAM to truly put it to the test though.
You're not wrong. Even the G5s take a shit on the web, even with TenFourFox. My guess is that a lot of stuff isn't optimized for PPC. It's still Firefox, and one guy is developing TFF
And I know that it's only a singly metric, and a bad metric at that, but AltiVec optimized bitcoin hashing puts the G5s on par with Intel's chips several years later

>G5s are definitely a different story, but I think by then they really were just starting to lag behind. They weren't total shit, but they could have done better
I completely agree that they could have done better, although I really do think that Apple didn't give it enough of a chance

>Sun released a lot of low-cost boxes like the Ultra 5/10 here and there, but no, it definitely hasn't. Still, POWER really hasn't either but it's gotten a ton of attention, I'd probably choose it over SPARC anyway for desktop use since SPARC is more optimized for extreme parallelism and other things that don't really lend themselves to that use case as readily.
This

I really wanted to buy the TALOS POWER8 workstation motherboard, but I can't afford to drop $3-4k on a motherboard and processor. I'd love to though

File: cu_1133_sl5b2_costa.jpg (162KB, 600x633px) Image search: [Google]

162KB, 600x633px

That time Intel had to recall a processor they only released to beat a record, by a paltry 33Mhz.

Performance was rated at the time to be slightly worse than the 1000Mhz version, even when there was a 1100Mhz one readily available.

>>56592546
I wonder what happened by the G5. Did IBM just give up?

>>56592932
PIII belongs in a slot.

>>56592653
I don't really get how that disagreement was actually possible, they were surface mounted chips. not socketed.

File: EvergreenIBM-Cyrix-286-upgrade-L.jpg (100KB, 640x472px)

100KB, 640x472px

>>56592797
>And I know that it's only a singly metric, and a bad metric at that, but AltiVec optimized bitcoin hashing puts the G5s on par with Intel's chips several years later
Interesting, I guess you'd expect the 64-bit PPC chips to have pretty good FP performance.

I still find my 2.0DC G5 pretty capable even for web surfing, though I mostly just use it for old systems emulation and SSH. Not sure about the older DDR/PCI models though, I've got one in storage (1.8DP) but it's fucked up and won't boot due to a RAM problem I haven't gotten around to fixing.

>although I really do think that Apple didn't give it enough of a chance
Maybe they could have extended it out a little further into the 64-bit era, but they did give it a pretty good run, the G4-based laptops weren't going to cut it forever while they waited for IBM to find a a way to put out the perpetual housefire that was the G5.

>I really wanted to buy the TALOS POWER8 workstation motherboard, but I can't afford to drop $3-4k on a motherboard and processor. I'd love to though
It's kind of exciting, too bad it can't run AIX though, not that I'd really want it on a desktop.

Anyway, the thread's going to die pretty soon and I have to get up early tomorrow. Sleep tight.

>>56593101
PPC, at least the later iterations past about ~500 MHz, just ran hot as fuck. I couldn't even use my Quicksilver in the summer because I had to reflow the piece of shit every week to fix the L3 cache, though I think I just got a dud.

Either way, IBM couldn't figure out how to fit the G5 into a laptop and Apple couldn't wait on it forever, though I think they'd already been planning the switch for years. Lots of "pure" RISC chips just couldn't evolve fast enough, at least compared to Intel.

>>56593229
Was never really sure how they worked since I've never seen one in the flesh, it seems 386/286 upgrades were basically like this 286 upgrade board (that featured IBM's way better 486SLC) that just fit straight over the chip.

ever atom ever made in that 2002 to 2010 era
it was marketed as windows 7 ready but can't even run the starter (lite) version and limited ram at only 2GB ram.

Laughing_whores.jpg

>>56580459
>>56580540
>>56580562
Pentuim 4's are shit they can't play 1080p movies (even AMD could do that on the same generation of cpu) and that is what it was marketed as on the 478 socket but celeron is shit on the 478 socket nothing runs right and most rams don't work on it.

the pentuim D was a fix to pentuim 4 but i only think the dual core pentuim D was okay at least but its still nothing compered to Athlon x2

P4's all run up to 90C even with good silver thermal paste
and also stop and shut down the PC when maxed on ram and will not flush the ram when full.

it also can't play movies at all and fails to run even Duke nukem 3d on max without stuttering like fuck
alot of games of that era could not play on pentuim 4's

Celerons of the p4 era where even worse they can't do anything right no movies no 90's era games or 80's trash it was hot and it failed so badly it makes a Duron 900mhz look good

>>56580595
nope not even bulldozer was as shit as pentuim 4 at least it did what it says it would , but your right mass effect one wasn't playable and shadows on old games looked fucked and some textures didn't show at all.

And the reason is because bulldozer didn't have many modern instruction sets
but
it could play 144Mhz display and 1080p movies

but that was only Gen1 of bulldozer known as zambezi

>>56593746
Know why I know your bait is lousy?
There are N270s in this thread running windows 10

Early 2000s when P4s were released was still a time of fucking DVDs and blurays were fucking released June 2006 a month before Core 2 Duo was released.

>>56580507
Fucking triggered