Amiga.org
Amiga computer related discussion => General chat about Amiga topics => Topic started by: ElPolloDiabl on December 16, 2009, 08:15:01 AM
-
Hi,
Just a roundup of the CPUs over the years that you loved or hated for whatever reason.
Hate: 8mhz 68000
It didn't have enough power to show off the Amigas awesome operating system.
Love: 68030
Gave you enough power for flight sims, racing sims and multitasking in the OS. Plus very backwards compatible.
Hate: 68040
Ran hot and the speed up was choked by the Amiga chipset. Although math operations were a lot faster. Not as backwards compatible
Love: Celerons with on die cache.
First time that gamers had an affordable CPU.
Love: PowerPC G3
A fast processor for a Mac or Amiga. I haven't had any software to push its limits though.
Like: the Pentium M/Pentium D
Put an end to the gigahertz war and its 2MB cache gave it awesome performance.
Hate: Current dual-core/quad core cpus that give you no noticeable performance increase. Better off spending your money on faster graphics cards and faster hard drives.
-
Love: 7447/7448
Badass performance. The fastest Altivec unit per clock. (I have 4 of these, ranging from 1GHz to 1.67GHz)
Hate: Pentiums
Ran way too hot and had poor performance per clock.
Love: Current intel dual-core/quad core cpus
Insane overclockability even with passive cooling. Very fast per clock.
-
Love: Current intel dual-core/quad core cpus
Insane overclockability even with passive cooling. Very fast per clock.
or just a good marketing trick to brand CPUs lower than their optimal frequencies to give impression of "OMG I'm an insane overclocker" to some users :lol:
-
Hate: 286 (because I nearly brought one for gaming)
Love: 486DX266 (I was too poor to buy one)
Love: IBM Blue Lightning processor 75Mhz (I did get one of this babies though)
Hate: P60 POS (I avoided this!)
Love: 030's & 060's but hate 040's
;)
-
A CPU is a CPU, they all do the same thing (although obviously with different degrees of performance, etc), so Im not too concerned what it is. Standouts in recent times though would have to be the early Athlon64's,... sure theyre outdated by now, but it was that cpu that forced Intel to pull thier thumbs out and probably plays a big part in how cheap and powerful hardware is these days. Core2 (duo/quad) is the other stand out. Low power consumption and heat dissipation (for a cpu of its class especiialy), and very overclockable, even with modest (retail cpu box) cooler.
-
For purely nostalgic reasons, 50mHz 68030 :knuddel:
-
Hate: anything from Transmeta (complete slugs)
Hate: AMDs (especially in laptops, I'll fry my eggs on the stove, thanks)
Hate: Desktop-class P4 chips in laptops (same reason as above)
Like: Pentium M/Centrino (not horrible performance, alot less overheating)
Love: Intel Dual-core and up of the last few years
Currently enamored with: Intel Atom (quick enough, hardly gets warm)
As far as Amigas go, 68060s are nice.
Other old favorites: 6502 (ubiquitous,), i8080 (IMSAI anyone?)
-
@ Fanscale: Pentium M/Pentium D - these two CPUs together??? You've got to be kidding...
-
When frying eggs - my ex ex laptop P4 3.4 GHz.
Ok, 68060 is the long time favourite - especially in A500T.
-
No votes for the 6502 or Z80? :(
You guys suck.
Seriously though, if you profess to love or hate a CPU you are either slightly weird or you are a coder. Assuming you are the latter and of the age range of original Amiga users, you probably cut your teeth on a 6502 or Z80 and moved up to the 68000 with the Amiga.
-
No votes for the 6502 or Z80? :(
You guys suck.
The 6502 is fine, but my vote would go for the 68k. Assembly programming on that thing it is just pure poetry.
Intel's never did much for me, though, which is probably why it's down there somewhere on the bottom of my list.
-
Love: 68000 - the first CPU I learned to program... I still have the 68k architecture in the back of my mind when ever I program anything...
Love: The first Athlon64 - the x86 done right, probably done more for the modern computer industry than any other.
Hate: PPC 603 - cost me a lot of money and offered very little.
Hate: Pentium4 - total dead end...
-
Love: POWER5+, Z80, 68000
-
Love: 68000 - the first CPU I learned to program... I still have the 68k architecture in the back of my mind when ever I program anything...
Love: The first Athlon64 - the x86 done right, probably done more for the modern computer industry than any other.
Hate: PPC 603 - cost me a lot of money and offered very little.
Hate: Pentium4 - total dead end...
To be fair, it wasn't the 603 that let you down, it was the hardware it was plugged into and the whole context switch thing. No L2 cache, slow memory bus. Any code you wrote that would fit into the 603's instruction cache and didn't need to spill registers onto the stack could run a lot faster than the 68K CPU on the same card.
Writing efficient code for WarpOS (and presumably PowerUP) is an interesting challenge.
-
Hate: Current x86 micros. I am obliged to use them though, but they are bloated stuff. Have you seen what compromises they had to make to achieve backwards compatibility?. You can fry eggs without problems on them, and you need your own power plant to drive them.
Love: Z80, an 8 bit beauty and efficiently designed. Easy to program, and dead easy to interface with watever components you may find.
-
I love all cpus, but only one at a time
-
love 040/060... even 25Mhz versions made using Amigas a true pleasure. I still remember my Falcon040/25 A1200 accelerator :-)
hate: 030... it was an accelerator but miggies don't look fast with it, even decompressing JPEGs takes ages
love G4: altivec was an elegant idea.
hate any 603 derivatives: crap ppc performance, makes me remember the old A1200/030 dark times when users who only played games were ignorant about 040 superiority
love 68000: it was incredible how many things you could do with a 7Mhz cpu!
hate 68010: an useless upgrade (more or less like 030 boards only used to launch whdload games due to availability of extra ram)
-
The 68010 was an important processor. It was never intended to be faster than the 68000, it was intended to better support P&G virtualisation (http://en.wikipedia.org/wiki/Popek_and_Goldberg_virtualization_requirements). To that effect, it was the first 680x0 to properly segregate supervisor and user modes and be able to recover from various bus errors.
You can't knock it for not improving the experience on a machine where none of these features really mattered.
-
MC6809 - First processor I coded assembly on. Easy enough for an 11 year old to program on!
MOS6502 - Disliked because of lack of 16 bit index registers and didn't like zero paged addressing. Apple2's weird video memory mapping didn't help me feel the love either.
i8088 - Yuck! Hated the instruction set and the stupid segment registers.
MC68000 - Absolutely loved this CPU. Is such a joy to program on. Great instruction set! K-SEKA and Dr. Pepper go hand in hand.
I Stopped programming assembly around 94, but other Memorable CPUs for one reason or another:
Original Pentium 90 -- quite a work horse and I paid way too much for this cpu!
* Celeron SLOT-1 CPUs - I remember burning on out with a peltier-cooler (condensation killed it). Was able to overclock like crazy!
AMD Athlon "Thunderbird" - got many years of great performance out of this chip.
* Pentium 4 - stupid lame chip that under performed at any Mhz and ran too hot.
Mac Mini's 1.4ghz PPC seems to be a real work horse running OS X as a media server for 3+ years!
Currently loving my 68060 CPU in my 1200 and 4000T
-
This Q9450 isn't bad, I suppose.
From a number crunching perspective, I'm more into GPU's at the moment.
-
Love: Z80 - I´ve learned to code with this baby. Easy to use, very flexible.
Like: 6502 - Quite interesting, but I never really coded it for "real" just to play a little with my C64. I need some time to play around with it (huahuahua)
Love: 68020 - Enough power, way better than the 68000.
Hate: Anything that came before the Pentium II - They are slow, and some can fry eggs...
Love: ARM - A hell of a processor, cheap and nice to play with. I've been playing with an ARM7 microcontroller, but I would like to play with the bigger guys
Have to live with: Current x86 derivatives.
-
Prettry much anything NOT Intel I like, well the 486's @ 100Mhz (and some other X series ones) were pretty good and robust from my experiance.
Mostly I like the... Z80, a real classic processor that really brought the British computing industry to our homes.
All the 680#0 Moto ones (Especially the ceramic ones), for much of the same reasons as above execpt the world. :)
Socket A Athlon AMD's and the latest AMD Quads (well the good ones anyway).
Also liked some of the RISC ARM and StrongARM CPU's in the Acorn's for their relability.
Wasn't keen on the 68010's though, almost totally pointless especially in the Amiga world.
-
It's funny. It seems most members of the Amiga community do not like Intel CPUs including the modern ones :) not surprising, just a bit silly.... maybe...
-
Loved: 6502/6510
Indifferent to: Z80. Had one in my 128, I think I used CP/M mode exactly...Once?
Loved: M68k @ 14mhz, namely an AdSpeed I put in my A500. Made it a wholly different machine.
Hated: M68020 @14mhz in my A1200. What was the point in all those colors and screenmodes when the stock box ran slower than my A500 did?
Loved: m68030 @ 28mhz. Ahh...Now this made my Amiga experience. I was given an 030 card for my 1200 by a friend (a DKB1204, IIRC) and it really flew, with 4mb fast RAM.
Loved: 486/100mhz, in my first PC. The whole board lasted all of a week until I blew it up...sniff...miss ya buddy!
Hated: '486/80, powered the 2nd PC. Ugh. Only served to remind me what I'd lost with the death of the first one :(
Hate hate HAAAAAAAAAAATE: Cyrix 6x86 series "budget" x86 compatibles. "Hey let's make a pentium clone with '286 FPU performance!" God I could tell some stories about the 6x86 pr/166 :puke:
Loved: Pentium 200mmx. Pretty much the pinnacle of the wonder years of P1 design (the 233 was a shade faster, of course, but I never had one :P ), and it came at a time when PC game innovation was at its peak. Go back and look at classic PC games and that's what your target CPU is/was.
Loved: Coppermine PIII 750 - I ran this guy for like two, three years. Loved every ounce of gaming goodness I got out of it.
Probably would have loved, but had a crappy motherboard: Intel P4 2.2ghz. Again, had a badly performing board. If I'd replaced the board rather than jumping to the Athlon 64, I might have had fonder memories...
Lovin': AMD Athlon XP3000 x64. Ooooold in terms of processor life but seriously? It's still going, it runs Win7 64bit (pro version) without a hiccup. Definitely gotta say this is a good CPU.
(I should point out that these are all user experiences; I haven't coded since the 64 days and that was in BASIC LOL)
-
I loved the 486/4x100 great for DOS, coming from a Commodore 64 to an Amiga A600 I thought the 68000 was fast but my Apollo Turbo Mk-3 33/40 is good enough for what I do realy. I hated the Pentium 4 I had nothing but trouble with that sod, The AMD range of CPU's gave very very good perfomance, Athlon dualcore 64bit 4600+ is great fast smooth never ever had any problem with that CPU. How ever having said all that I must admit the Pentium MMX233 was one of the best CPU's to be made for its class.
-
Seriously though, if you profess to love or hate a CPU you are either slightly weird or you are a coder. .
You made some typos here:
Seriously though, if you profess to love or hate a CPU you are slightly weird (aka a coder). .
Much better ;)
-
Love:
all 68k - an assembler freak's wet dream
6502 - an assembler freak's nightmare, but the heart of Apple ][ & C64
Athlon 64 - starting off several evolutionary steps simultaneously
G4 - really kicks ass in Altivec
K6-2 - great upgrade for Pentium boards (thanks AmigaPixel!)
Hate:
Pentium 4 - worst performance per Watt ever, complete nonsense
8086, 286 - coder's bad dream, dumbest memory model ever
-
I love the 6502,6510 processors they helped form my sweet childhood memories. :)
Nowdays I could not care less what's inside my box.
-
I enjoyed my 030 @50mhz in my A2000, with the 50mhz fpu. But my 060 in my A1200 was a gem,definitely hot though especially when I use to render Lightwave animations. I had to take off the trap door and raise the A1200 up an inch or so and put a house fan right next to it.
By the way I have an old HP with a P4 1.5 socket 423 (A real slug!) that I want to upgrade as cheap as possible.The consenus seems to be the P4 was/is a bloated oven. I already have a Compaq with a Pentium D 820 cpu, and the mb supports Core 2 Duo up to the E6000 series. I seen few socket 478 P4 2.8 with Hyper Threading 1 meg L2 for as cheap as $15 on Ebay and Asus mb less than $20 also with HT support. I figure for the money it wouldn't be bad as long as I get a good heatsink and fan. And a good quality power supply. I have't used an AMD since the K6-2 400 so I am not very knowledgeable with newer AMD cpus. Any advice on this?
-
LOVE/Hate = TI TMS9900 Gads, no one else here used these things?
Love 68030 & 68060 - used to heat my room with an '040
non-committal to most other CPUs.....
-
6502
I first learned to program on the 6502, and it was the CPU by which all others would be judged - I thought it was fantastic. Some of the euphoria attributed to the 6502 is probably due to the incredible feeling of getting my first programs to work though.
Z80
I then moved on to the Z80. Again, fantastic CPU and actually easier to work with than the 6502 I thought. In fact, it was so easy to code with this CPU that if you weren't careful it could give you some bad programming habits! The sheer number of registers you could work with was the main reason I found this an easy CPU to work with.
Some brief CPU nasties
I then briefly moved on to x86, but absolutely loathed it. Coding for intel's processors in assembly is akin to stabbing one's eye with a sword. It was horrible. I took an instant dislike, and moved away.
....to the 68K? Actually, no. I never coded for the 68K. I moved on to Microcontrollers with the Microchip PIC. I coded in Assembly for a while, but to be honest this processor is a bit of a nasty little thing to code on in Assembly as well. The bank switching is the main thing I dislike about it.
Microchip PIC
These days I still use the PIC as my main processor of choice, but I tend to write most of my stuff in C unless there is a specific requirement for Assembly (which, to be honest, is rare).
But I guess my favourite CPU from a users point of view would be the 68K since with it, through the Amiga, I have enjoyed many many years of good fun. I actually bought quite a nice 68K Assembly language book but never really indulged it.
Applehammer.
-
I should add that although I despise the x86 processors (in terms of writing assembly for them), Robert Noyce is my one true childhood (and adult) hero.
AppleHammer
-
Although not a cpu, I love the 555 timer. She gives me a buzz. :-)
-
Love:
M6800
All 68k
AMD K6 family
PPC g4 and g5
Intel P4 HT (plenty of power for today)
Hitatchi Sh2/SH4
Hate:
MOS 6502 (stupid lawsuits)
Emotion Engine
Intel Core Duo and higher
Pentium 2
Intel 8008
Intel 80088 ( 16-bit bus)
Intel '186 ( no performance increase)
Intel 386SX
Intel 486SX
Intel Overdrive
PPC 601
Whatever chip in Efikas
Motorola Dragonball
MiPS (PS1 CPU)
-
Xeon “Nehalem”
Phenom X4
Basically any processor with 4 or more cores....
-
I don't like the Pentium 4. And after trying to reverse engineer some code for the CDP 1802 I'm not too sure about that one. But what I HATE is crippled CPUs. For instance, the Celeron 766, perfectly fine CPU cursed with a ridiculous locked multiplier for 66MHz FSB long past the time when 133MHz-capable boards and memory were available. Intentionally making things slow is not cool. (Video cards with decent name brand GPUs and horrendously slow RAM also annoy me)
-
I do like the 6502/6510, the 68000, the Pentium Pro, the K6-III+, the ARM3 and the G3.
I more or less dislike the Pentium 4, the Celerons derived from the P4 and the slot A Athlon.
-
Love:
Z80
68020
68030
68060
PPC 603/604
G4
Socket A Athlon
Core 2 Duo and derrivatives
Hate:
Pentium family from original up to 4 - what a load of crap!
68040 - so hot and incompatible at times
G3 - underpowered for what it was advertised
G5 - inferior to G4 (higher clocks though)
-
No votes for the 6502 or Z80? :(
You guys suck.
Seriously though, if you profess to love or hate a CPU you are either slightly weird or you are a coder. Assuming you are the latter and of the age range of original Amiga users, you probably cut your teeth on a 6502 or Z80 and moved up to the 68000 with the Amiga.
Only just started reading the thread. I would have said Z80 as that is what I cut my teeth on in an Amstrad CPC 464 then 6128
-
Love: 7447/7448
Badass performance. The fastest Altivec unit per clock.
You have any experience of the 8610?
-
Love:
Z80
68020
68030
68060
PPC 603/604
G4
Socket A Athlon
Core 2 Duo and derrivatives
Hate:
Pentium family from original up to 4 - what a load of crap!
68040 - so hot and incompatible at times
G3 - underpowered for what it was advertised
G5 - inferior to G4 (higher clocks though)
I think I've read somewhere that 68060 and the original Pentium are quite similar in design? Or something like that...
-
I can't believe people are complaining about multicore CPU's. I find my quad core indispensible, especially when transcoding my media collection. Really, why convert one stream at a time, when you can do three concurrently and still have ample horsepower left to do all your normal stuff?
-
my all time sweetheart...030@50 :-)
-
@ Karlos: you won't find much logic when Amigans start bashing "competing" platforms :D
-
@ Karlos: you won't find much logic when Amigans start bashing "competing" platforms :D
I'll let top answer that:
Tasks: 155 total, 4 running, 151 sleeping, 0 stopped, 0 zombie
Cpu(s): 72.2%us, 0.4%sy, 0.0%ni, 16.9%id, 10.3%wa, 0.1%hi, 0.1%si, 0.0%st
Mem: 4053804k total, 4026464k used, 27340k free, 8892k buffers
Swap: 20000884k total, 18008k used, 19982876k free, 3396568k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
8631 karlos 20 0 175m 32m 5096 R 100 0.8 13:58.72 mencoder
8652 karlos 20 0 170m 27m 4652 R 97 0.7 6:05.80 mencoder
8642 karlos 20 0 170m 27m 4652 R 96 0.7 7:52.02 mencoder
(...snip...)
-
I take it that top reports every thread as a separate CPU? an octocore, nice :D
-
I take it that top reports every thread as a separate CPU? an octocore, nice :D
If graphs are more your thing, the standard gnome system monitor shows a separate trace for each core (as does windows task manager). Or if you prefer the console but top is too spartan, htop shows a bar for each core and one for memory usage.
-
Hi,
Like : 6502 ( Apple IIc rulez :p, my first(s) asm coding lines )
like 68000 ( Amiga 500 )
like 486DX2 266Mhz ( My first PC was with this CPU :p Duke nukem power)
love 68060 (superscalar, fpu all in 1 cpu :p)
don't like : Intel / AMD cpus (too poor in registers for coding comparing to 680x0/6888x cpu(s)/fpu(s))
don't like : Dual/Quad core : they show the lack of progress in CPU technology
Regards,
AmiDARK
-
don't like : Dual/Quad core : they show the lack of progress in CPU technology
If you think they are less evolved than single core CPU's running at higher clock rates then you must have a pretty one dimensional view of processor development. Just recall the P4; an exercise in marketing the "MHz myth". One core on my Q9450 at 2.66 GHz does more work per MHz than the 3.2GHz P4 at work. And there's four of them.
It astounds me that amiga enthusiasts, who, let's face it evangelised pre-emptive multitasking as the holy grail of OS design fail to appreciate how multicore CPU's are the hardware equivalent. Four cores means that 4 threads can each run at any instant, unlike a single core CPU wherein (even given HT) only one process is running at any given moment.
You don't even need specially written massively threaded software to appreciate multiple core, just run a few big CPU hogging apps concurrently and you see the benefit immediately.
-
It is rather astounding that Amiga folk are so disparaging about anything invented after 1990. When I bought my Amiga back in '85 it was *because* is was top of the line hardware, has this community became a hangout for luddites? The Amiga was created out of cutting edge hardware pushed insanely beyond it's limits. That's what I bought into in 1985 and that's what I still want now.
The multicore chips were the best thing to hit the market this century. The old single core P4's are dogs compared to even the lowest end dual cores. I've got a pair of quad cores in my main computer and am astonished at the rendering and video capabilities of it. Try a lot depends on the GPU, but as Karlos says, even basic multitasking is a heck of a lot faster and smoother. I will never go back to a single core for my main machine again.
-
I even notice a massive difference on my lowly AMD64x2 especially in some games that seem to make good use of dual core .
-
Actually when I first started using multi core CPUs I didn't think it was a quantum leap compared to single core (I don't use multithreaded apps often). It was when I was forced to go back to single core a month ago when I really felt the difference. Even when doing simple tasks such as web browsing or burning a DVD the computer can become unresponsive and generally not pleasant to work with. Funny thing that before I first got a dual core CPU I had never noticed that :)
-
Borcholli core - love.
Over clock or over cook!
-
Single core CPUs were never a good idea, especially when we started to demand multitasking... But technology just didn't allow us more than one chip as complex as a CPU in a home computer... but now it does, and I would say two cores is the minimum any computer system should have... my iPhone is the only computing device I use at home with a single core... and hopefully Apple can sort that out soon ;)
-
Single core CPUs were never a good idea, especially when we started to demand multitasking... But technology just didn't allow us more than one chip as complex as a CPU in a home computer... but now it does, and I would say two cores is the minimum any computer system should have... my iPhone is the only computing device I use at home with a single core... and hopefully Apple can sort that out soon ;)
Well they got you to shell out for a locked piece of hardware that doesn't belong to you and needs their permission to do anything with, so I'm sure they can do just about anything!
:D :D :D
-
Well they got you to shell out for a locked piece of hardware that doesn't belong to you and needs their permission to do anything with, so I'm sure they can do just about anything!
:D :D :D
Maybe,
I'd like to help clear up that little myth about the locked nature of the iPhone, if you use the iPhone SDK you can install any app that you compile yourself (and you can give that app you compiled to up to 100 friends) on your iPhone.
It is only if you want to sell/distribute an app to unknown people that Apple then require approval... which is fine with me, I'm having fun with my mobile device... :)
-
I've jailbroken my 3GS, it works so much better that way. Gotta be careful about backgrounding things though, they still suck power in the background...
-
which is fine with me, I'm having fun with my mobile device... :)
Is the sock feeling jealous yet?
-
Surprised noone mentioned the Phenom2, has great price/performance and very good upgrade shcemes where you didn't need a new motherboard.
It also seemed to have solved the so called performance problems of the original Phenom which a bigger/better cache or whatever.
It's also almost like a favor in the name of competition you're doing in buying these too...
I haven't used many other processors, let's see: 68000, too slow, 020, too slow, 040 good price/performance back in the day, but could have been better, 060 was very good, almost like a pentium but expensive..
I had a Phillips MSX VG8020 back in the day, but I was just a kid and mostly played cool games:) It was graphically better than the ZX :)
-
I even notice a massive difference on my lowly AMD64x2 especially in some games that seem to make good use of dual core .
And when they can have an operating system that is able to perform operating system tasks in parallel thats when we'll really feel the benefits of multiple cores.
-
Single core CPUs were never a good idea, especially when we started to demand multitasking...
It never bothered me when multitasking on my Cyberstorm 68060 single core..
-
Single core CPUs were never a good idea, especially when we started to demand multitasking...
Well, who could know that we would need multi cores some day when some people said 640KB is enough RAM anyone could ever need :)
-
The fact that the Amiga had coprocessors tells you all you need to know about single core CPUs :)
-
like: don't particularly have one
hate: most of Intel x86 because the product line changes too fast.
Single core CPUs were never a good idea, especially when we started to demand multitasking...
For intensive computer tasks (like high-end games, data-crunching of any kind), I prefer multi-core. For non-intensive computer tasks, I prefer single-core.
Both have pros and cons.
-
What are the advantages of a single core? Other than power consumption and price. When we have multi cores with totally independent cores, which can be completely turned off and on as needed (is it already possible?) the power consumption also won't be an issue.
-
Well, who could know that we would need multi cores some day when some people said 640KB is enough RAM anyone could ever need :)
LOL- why would you ever want more I ask you? :D
-
What are the advantages of a single core? Other than power consumption and price. When we have multi cores with totally independent cores, which can be completely turned off and on as needed (is it already possible?) the power consumption also won't be an issue.
On my Q9450, the clock speed of each core can independently switch between 2GHz or 2.66GHz as needed. That might not sound much but as the current draw is proportional to both the voltage (which increases with clock) and the number of gates switching per second, this would imply a pretty large difference in power dissipation between the two states.
-
Just the furthur karlos's post, I have a 2.2Ghz Core2Duo here that on average uses less power than my old 1.5Ghz PPC 7447 ( aka G4)... But is many many times more powerful...
-
Well, who could know that we would need multi cores some day when some people said 640KB is enough RAM anyone could ever need :)
Urban myth was never said in that context
-
The fact that the Amiga had coprocessors tells you all you need to know about single core CPUs :)
Well, to be fair, the coprocessor argument is slightly different. A dedicated GPU is almost always going to be more efficient at graphics operations than a CPU.
However, it is worth noting that modern GPU's are so efficient at what they do due to using a many-core solution to the inherent parallel data operations common to graphics processing.
If you heavily multitask, a multicore CPU is always going to outshine a single core, no contest. I was able to use 3 of my 4 cores for encoding video (three separate mencoder processes, see above) and get the job done in 1/3 the time it would take a single core. And all without losing any speed at all, since the fourth core was more or less free to run all the non-cpu bound stuff.
Of course, it would probably have been even faster if I were to use the GPU for the job :)
-
To address the above point about each having their respective strengths and weaknesses, I honestly can't see any advantage to a single core CPU over a multi core one. Unless you are not in a position to leverage the more than one core, that is, in which case there's obviously no benefit of more than one core.
-
Well, the only situation now where multicore is a burden is when using an operating system that can't use more than one core... So I guess that is rather ironicaly probably only AmigaOS now... :)
-
Well, the only situation now where multicore is a burden is when using an operating system that can't use more than one core... So I guess that is rather ironicaly probably only AmigaOS now... :)
Depends on what your definition of "an operating system using more than one core" is
What would interest me more than just having one program crunching numbers within one core, another crunching numbers in another core, a third crunching numbers in a third core etc is if the operating system itself could could be split across the cores running a different task in each core, or i could get one program to split itself across each core. I hardly ever need to encode more than on video/dvd at the same time, but I'd like to have four cores all working to decode that single DVD at the same time. At the moment this doesn't happen particularly well, if at all, as most software and the OS is not designed to detach into parallel tasks like that. Most benchmarks I've seen from dual core systems are at best 30-50% faster than a single core at the same clock speed, and for some benchmarks with some dual core cpu's the dual core can even be slower. (Windows 7 might be better, i haven't had any experience with it)
-
@ JJ: In what context? I'm sure SOMEBODY said it, I'm not saying it was Bill. Actually, he is/was quite the visionary and it doesn't sound like something he would say at all and I'm not attributing the line to him.
-
Hi,
> Just a roundup of the CPUs over the years that you loved or hated for whatever reason.
> Hate: 8mhz 68000
I dont hate because the first Amiga 68K because the custom chips let it run full speed - in color: sorry Apple!
Remember: A 68000 @ 7.159 Mhz (on an Amiga, obviously!) CREATED the entire "demo scene" with what you_could_do on a stock Personal Computer.
I love how simply adding a 68010 or an ICD (or Aminet!) 14Mhz Accelerators so cheaply and totaly upgraded the base machine bought in 1985.
Thats right: 1985!
...Were ya borne yet?...
*heh*
I will respond to the rest late...
-
Depends on what your definition of "an operating system using more than one core" is
What would interest me more than just having one program crunching numbers within one core, another crunching numbers in another core, a third crunching numbers in a third core etc is if the operating system itself could could be split across the cores running a different task in each core,
I dunno where you are getting your impression from but the OS does do exactly this. If one core is busy, any newly created processes will be put on a different core. Also, most kernels allow a process to move between cores whenever the task scheduler decides to redistribute the load. For example, on my machine, a single-thread process using 100% cpu load is actually distributed across all four cores (although only one one of them at any instant), spending approximately the same amount of time on each one. That's probably better thermally for the CPU than having one core at 100% and the other's permanently idling along.
or i could get one program to split itself across each core. I hardly ever need to encode more than on video/dvd at the same time, but I'd like to have four cores all working to decode that single DVD at the same time. At the moment this doesn't happen particularly well, if at all, as most software and the OS is not designed to detach into parallel tasks like that. Most benchmarks I've seen from dual core systems are at best 30-50% faster than a single core at the same clock speed, and for some benchmarks with some dual core cpu's the dual core can even be slower. (Windows 7 might be better, i haven't had any experience with it)
Programs must be written to support this. Presently, the only way for an application to do this would be to create several threads and give a portion of the workload to each one. This assumes that the job they are doing supports this type of parallelism. Many jobs don't. There's no magic way to automatically fork a process into subthreads and divide work between them.
The next generation of C++ aims to include language-level support for writing multithreaded code, which will certainly help.
Incidentally, I was using 3 of the 4 cores to encode a single DVD. Each one was working on a subset of the total number of titles on the disc. I could have used all four cores but the number of titles was an exact multiple of 3 and I was doing other stuff at the same time, so I launched three processes. So, in effect I was getting exactly what you were asking for.
-
The next generation of C++ aims to include language-level support for writing multithreaded code, which will certainly help.
Apple have added blocks to C...
http://arstechnica.com/apple/reviews/2009/08/mac-os-x-10-6.ars/10
this goes hand in hand with their grand central dispatch subsystem... :)
-
Maybe,
which is fine with me, I'm having fun with my mobile device... :)
Oh no, no, don't misunderstand me. I think it's absolutely wonderful for you and Apple that you've agreed to behave yourself while you lease their hardware! No, seriously, that sounds like you guys have a great arrangement.
-
Actually you own the iPhone, so you can do anything you like to it, but Apple won't help if you break something, that is Jail breaking voids warrantee. I can live with that.
-
Apple have added blocks to C...
http://arstechnica.com/apple/reviews/2009/08/mac-os-x-10-6.ars/10
this goes hand in hand with their grand central dispatch subsystem... :)
Actually, I find that absolutely disgusting. New syntax should not be added to C for any reason whatsoever, unless it is an official ISO standard that can be adopted openly. Which pretty much hasn't happened since C99. For multithreading, I would much prefer a standard library of functions. Actually changing the language syntax breaks portability in a way that's incredibly awkward to work around. Redefining the ^ operator for passing a block of code isn't even readily ported to C++ where operator overloading is a language feature.
Now, C++ added further syntax to C but it didn't change any meaning of existing C syntax (ok, there are a few small gotchas) but whenever something new is added to C++ nowadays, it's via the STL.
-
Actually you own the iPhone, so you can do anything you like to it, but Apple won't help if you break something, that is Jail breaking voids warrantee. I can live with that.
well, jailbrake doesn't void warranty... ofcourse you have to restore with apple firmware before claiming warranty ;-)
but the thread was for cpus so:
LOVE: 030,060,ppc750,ppc7455,Duron,athlonXP,K64,Core,core2
Hate: 68000,040,Celerons ALL!,Pentium4 ALL,ppc970fx
athlonXP and durons are loved because of their age, otherwise they have almost the same performance per tick with G5 (ok G5 is better on vectors, have THE fastest BUS till now, but also newer)
@piru
fastest G4 was 7455(even it was older) and not 7447/7448, because it supports L3 cache
and you know the problem with bus speed on G4....
-
to be quick and less complex
liked
moto 6800 classic and 030
the slot pentiums for ease of throwing in and workin for long periods cheaply
and my amd 64x2 cause its current works and is ok
worst cpu
to be avoided at all cost is the cyrix i have under clocked them many times to work at all !
-
worst cpu
to be avoided at all cost is the cyrix i have under clocked them many times to work at all !
This, oh god, this.
I want to invent time travel specifically for the purpose of going back in time and stopping the Cyrix company from being formed, and the 6x86 line from coming out. Damn my mid 90's needs for a cheap Pentium clone.
-
Actually you own the iPhone, so you can do anything you like to it, but Apple won't help if you break something, that is Jail breaking voids warrantee. I can live with that.
You do not own the iPhone. (http://scitech.blogs.cnn.com/2009/02/18/could-jailbreaking-your-iphone-land-you-in-jail/)
-
What most tend to forget is the fact that a single core would always be faster than a dual core running half its speed (or quad respectively) - 1x 6 GHz > 2x 3 GHz.
There is no 6 GHz? Yes, that's the point. The industry has changed over to multicores as the only way to considerably raise performance since there's no way to increase clock speeds any more. There are gentle increases as well as gradual architecture improvements, but nothing to show off to any potential customer.
-
What most tend to forget is the fact that a single core would always be faster than a dual core running half its speed (or quad respectively) - 1x 6 GHz > 2x 3 GHz.
There is no 6 GHz? Yes, that's the point.
I know I personally tend to forget situations that don't exist.
-
You do not own the iPhone. (http://scitech.blogs.cnn.com/2009/02/18/could-jailbreaking-your-iphone-land-you-in-jail/)
Sorry, but what does a year old editorial piece have to do with ownership of the iPhone...
I bought a device with an operating system, that I can use on a daily basis... I can buy apps for and I can write my own apps for... It's pretty powerful for the price and quite innovative... but if I modify it beyond the manufactures specification (like swap out the installed firmware) they have stated that they wouldn't provide me with any support...
I am of course referring to my A1200... but the above equally applies to my iPhone...
-
I dunno where you are getting your impression from
from this:
For example, on my machine, a single-thread process using 100% cpu load is actually distributed across all four cores ****(although only one one of them at any instant)****, spending approximately the same amount of time on each one.
and this
***Programs must be written to support this. ****
and this
The next generation of C++ aims to include language-level support for writing multithreaded code, which will certainly help.
hence my statement that multi-core CPU's aren't used particularly well at this time.
-
Wait, hang on. That was given in answer to your earlier point. Didn't you say you'd prefer it if the OS distributed the work between the cores? Well, that's exactly what the kernel does do. To the extent that even a single thread will move from one core to another depending on the existing load on each core and for improved thermal distribution (if nothing else).
A single thread of execution, by definition cannot run on several cores concurrently. Do not confuse multi core with superscalar execution. They are not the same thing and for what it's worth, each of these multiple cores is superscalar and nicely pipelined too. Each one is thus executing more than one sequential instruction in whichever thread they happen to be executing at any instant.
Now, as for the issue about writing stuff for multi core, you have to accept that not every computational task is parallelisable. Therefore there will be some things that cannot on their own take advantage of more than one core. However, you must bear in mind that on all multitasking operating systems there are usually many threads that are ready to run and the scheduler has to pick just one of them per quantum. Not so on a multicore. It will allow as many threads to execute per quantum as the hardware will allow. From a throughput perspective, running four threads in one quantum is up to 4x faster (assuming no interdependencies) than running one each for four successive quantums. Better throughput means less latency for all tasks.
What I am saying is, that unless you use a single application on a single-tasking "OS" (or MacOS classic as it's more commonly known :lol:), it's impossible to not reap a performance gain from multi core in general use.
The principal exception to this rule is when you want to run a single, very CPU intensive task that cannot ever be optimized for parallel execution where a single core processor may offer a price/performance benefit over a more costly multi core part that can't run that one task any faster.
Now, amiga users are fans of multitasking, after all it was always our OS's strength. Seems a bit daft to try to ignore the clear advantage of a multi core CPU in an SMP capable multitasking OS.
-
Having used a quad core, there is absolutely no way I'm moving back to a single core for a main machine, unless it can match the same overall throughput. However, it'd have to be at least 10GHz to manage that.
-edit-
Actually, looking back at it, I've been enjoying multicore computing ever since I got my first PPC board. A comparatively simple task (by current standards) such as playing back MP3, on the 68040 was a joke. It could manage mono 22kHz and there was no cpu time left to do anything else. That all changed when the PPC decoding engine was running in AmigaAMP. That whole arrangement was far less efficient than SMP (unavoidable considering they are two totally different processors) but the benefit of having a separate CPU core to dump the job onto was clear even then.
-
Wait, hang on. That was given in answer to your earlier point. Didn't you say you'd prefer it if the OS distributed the work between the cores? Well, that's exactly what the kernel does do. To the extent that even a single thread will move from one core to another depending on the existing load on each core and for improved thermal distribution (if nothing else).
A single thread of execution, by definition cannot run on several cores concurrently. Do not confuse multi core with superscalar execution. They are not the same thing and for what it's worth, each of these multiple cores is superscalar and nicely pipelined too. Each one is thus executing more than one sequential instruction in whichever thread they happen to be executing at any instant.
Now, as for the issue about writing stuff for multi core, you have to accept that not every computational task is parallelisable. Therefore there will be some things that cannot on their own take advantage of more than one core. However, you must bear in mind that on all multitasking operating systems there are usually many threads that are ready to run and the scheduler has to pick just one of them per quantum. Not so on a multicore. It will allow as many threads to execute per quantum as the hardware will allow. From a throughput perspective, running four threads in one quantum is up to 4x faster (assuming no interdependencies) than running one each for four successive quantums. Better throughput means less latency for all tasks.
What I am saying is, that unless you use a single application on a single-tasking "OS" (or MacOS classic as it's more commonly known :lol:), it's impossible to not reap a performance gain from multi core in general use.
The principal exception to this rule is when you want to run a single, very CPU intensive task that cannot ever be optimized for parallel execution where a single core processor may offer a price/performance benefit over a more costly multi core part that can't run that one task any faster.
Now, amiga users are fans of multitasking, after all it was always our OS's strength. Seems a bit daft to try to ignore the clear advantage of a multi core CPU in an SMP capable multitasking OS.
be that as it may the fact is that today's OS's and more importantly most applications (afterall people care more about running apps not operating systems) don't take full advantage of multicore cpu's, its still an area that is evolving.
-
-edit-
Actually, looking back at it, I've been enjoying multicore computing ever since I got my first PPC board. A comparatively simple task (by current standards) such as playing back MP3, on the 68040 was a joke. It could manage mono 22kHz and there was no cpu time left to do anything else. That all changed when the PPC decoding engine was running in AmigaAMP. That whole arrangement was far less efficient than SMP (unavoidable considering they are two totally different processors) but the benefit of having a separate CPU core to dump the job onto was clear even then.
Gee that is really poor for a 68040. I remember playing back mp3's in 14 bit 44 khz on a 40 Mhz 68040 Apollo A1200 (with executive) and it was fine (on a dblntsc screen). I had a little shell script that opened up a requestor to select the mp3, beacsue from memory GUI-driven mp3 players seemed to significantly increase the cpu load when playing the same song.
-
68060 or 6502
68060 becuase of the giant leap forward performance wise of the 68k family, without lots of heat and so many great machines used a 6502 - sure it had drawbacks but the alternative is a Z80!
-
be that as it may the fact is that today's OS's and more importantly most applications (afterall people care more about running apps not operating systems) don't take full advantage of multicore cpu's, its still an area that is evolving.
Most current operating systems do take full advantage of multicore processors. It is pretty much impossible not to if you actually support SMP (for example) and your OS pre-emptively multitasks.
Individual applications may not, that I agree with. Even so, to suggest that they don't benefit from multicore is a bit of a fallacy since they'll never be the only thing competing for CPU time, assuming your OS actually multitasks.
On that basis, I don't agree that people don't care about running OS. After all, people gripe that windows runs slowly on their single core CPU. Well guess what? Windows likes to run dozens of processes. Many processes on one core means each process can experience significant latency and the scheduler has to do more work to ensure they all get serviced fairly. This issue is mitigated by having more than one core. If a core is busy and another task needs to run, it can do so without waiting for the previous one to yield simply by executing on whichever core is free. Since you aren't likely to be running four compute bound processes concurrently, on a quad core there's almost always one core that's free at that moment to take the job on.
It would be nice if every application were naturally multithreaded but it really only makes sense for certain operations, which is why multitasking is the key way to leverage the benefit of multicore, at least for now.
-
Gee that is really poor for a 68040. I remember playing back mp3's in 14 bit 44 khz on a 40 Mhz 68040 Apollo A1200 (with executive) and it was fine (on a dblntsc screen). I had a little shell script that opened up a requestor to select the mp3, beacsue from memory GUI-driven mp3 players seemed to significantly increase the cpu load when playing the same song.
You've just said it yourself. It can play the mp3 but asking it to render a gui at the same time is a stretch too far. I'll wager you didn't do much else that required any CPU time when it was playing too (Just think, if you had two 40MHz 040's and an SMP version of the OS, you could easily have the GUI and run several other things besides).
Now, consider that my 040 was 25MHz and I was running GoldEd and gcc. I needed executive just to prevent the music from breaking up just scrolling golded's display. Which meant that the editor suffered instead but that was more bearable.
Once the PPC went in, 44kHz mp3 playback 14-bit no problem, even used the graphic equalizer too to improve the output a bit.
Another clear win for multiple CPUs, even an example as asymmetric as this one :)
-
Love: 68K cpus, because they're fun to program on :)
Meh: Most other cpus, because they're not fun to program on :D
Dislike: All 8 bit load-store style cpus, such as 6502, because they suck big balls to program on :(
Hate: None, I don't hate cpus :D
-
Hate: None, I don't hate cpus :D
Quote of the day :D
-
Anyway, not sure I actually did my round up, so here goes.
Liked the Z80 as this was the first CPU I actually wrote anything for. Pretty big instruction set for an 8-bit processor too and a fair few undocumented instructions.
Never really did much on the 6502. Appreciated the design though, where the Z80 took at least four clocks (and potentially a lot more) to do anything at all, the 6502 was pretty much hardwired in comparison and had a much more uniform execution speed. Unlike the z80 the machine was register sparse but the zero page made up for that. In those days, accessing memory or an internal register made very little difference speed wise.
Loved the 68000; from an assembler programmer's perspective it was almost the perfect design. Each major revision brought something to the table beyond extra speed. The 68020 was even more fun to code on, you learned a bit more about pipelining your code and how to leverage an instruction cache (tiny as it is). The 68030 introduced a datacache that was equally tiny but also fun to try and use. The 68040, for me, was almost a perfect CPU by then. Shame about the heat but having everything on chip and a performance that laughed at the 68030 more than made up for that. Caches were finally big enough to be really useful and the move16 operation made a difference when shoving data to chip ram.
The 68060 was a lot more challenging to code for, since you finally had to take superscalar execution into account to get the best performance. Alas, this is one M68K I don't currently have :(
x86
Everything prior to the 386 was a complete joke and unworthy of mentioning. The 80386 wasn't bad though, once you got out of real mode and into flat memory space. From an assember coding perspective though, the instruction set remained a bit of a dog's dinner. The 80486 was better again and finally started to reach (at the time) silly clock speeds. One thing they had that I wished the 68000 had was the bswp instruction. Having to use three instructions on the 680x0 to byteswap one longword seemed very odd, given that it's not an uncommon thing to have to do. The pentium mmx introduced SIMD but I never really did anything x86 since.
Contrary to many folks here, I rather liked the PPC. The 603e on my Blizzard was a whole new toy and one that had a programming model that was quite fun. It's probably the only RISC architecutre I can say that about (save ARM, which has some cool features too). Writing effective code for WarpOS was (and is) quite a challenge in it's own right. Architecturally, 32 GPR registers are a lot of fun from an assembler programmer's point of view, as are 3 operand instructions. I also like the fact that you can basically decide which arithmetic/logic operations should update the condition codes, since this allows you to tweak loops and the like in a way that was much harder on 68K. Actual syntax, not so nice. Bit ordering seems bizarre if you are used to M68K, since bit 0 is the MSB and bit 31 the LSB. Never really understood the purpose of that.
Never did get into PPC beyond that. I have a G4 machine that I need to get into coding. Particularly interested in the altivec stuff. The permute features in there seem rather nice and a bit unique maybe.
x64: Well, this was a surprise. An x86 derivative that is actually fun to code for. The 16 register model is a nice compromise IMO. I'm still learning it though, so Jury's out.
ARM: Quirky but cool. Especially liked the conditional execution of instruction feature, far more effective than having to branch somewhere just to execute a single instruction.
MIPS: never really got into this, so can't say.
Alpha: Did some computational chemistry coding on this beast. It was all in C though, however the compiler had vector extensions for the chip. Once you figured out how to use them the speed increases for all that n-body charge force calculation could be sped up nicely. Alas, since then, GPU coding has shown where the real future for that sort of computation lies.
With the exception of the 8086/186/286 it's probably fair to say I like most CPU's :) It's hard to hate any, they all have their character.
-
@Karlos
Hmmm, If this were a discussion about CPU archs then I'd have quite a bit to say... but I'm pretty sure we were supposed to pick actual, physical implementations and justify our reasons... :)
-
I just did. My reasons as to why I like or dislike CPUs tend to be architectural.
-
Hate: None, I don't hate cpus :D
I may not technically hate it, but I have a deep dislike for PPC. If the PPC never existed, odds are Amiga would have moved on to a economically viable arch. Instead, I get to watch another years worth of rerun-like episodes of Cargo Cult Island.
-
Gee that is really poor for a 68040. I remember playing back mp3's in 14 bit 44 khz on a 40 Mhz 68040 Apollo A1200 (with executive) and it was fine (on a dblntsc screen). I had a little shell script that opened up a requestor to select the mp3, beacsue from memory GUI-driven mp3 players seemed to significantly increase the cpu load when playing the same song.
cough http://www.amiga.org/forums/showthread.php?t=50831 (http://www.amiga.org/forums/showthread.php?t=50831)
I'd be willing to wager that was using a much more optimal mp3 decoding library than I was back in 2000 when I was using an 040 for the job and probably wasn't doing 14-bit playback either.
22kHz mono decoding was the setting you needed on 68K mpega.library based decoders back then if you actually wanted to use your computer at the same time as listening to anything.
I wouldn't say it was poor performance from a 68040, just practical use of available resources. Some of us used to use our systems for more than just trying to prove to PC users that we could do things they were taking for granted.