Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: Daniele on March 21, 2008, 07:48:21 PM
-
Hello guys,
I would like to know what difference exists between an acceleartor board with '030@50MHZ and another with '040@50MHZ.
Please let me know
-
An 040 50Mhz is around 3.8 times faster than an 030/50Mhz
-
030 @ 50Mhz = ~9MIPS
040 @ 50Mhz = ~40MIPS
-
Yes about 4 times. It is also a lot hotter of a chip. The only way to have an Amiga with a 040@50MHz is with the Sonnet Quad Doubler board. Not recommended. Just get a 060@50Mhz. About two times faster!
-
TjLaZer wrote:
Just get a 060@50Mhz. About two times faster!
To clarify: The 060 is about 2 times faster than the 040 at the same clock speed...
-
MIPS is not accurate measurement of processor speed. It's actually far from being 4x as fast.
-
I may be totally wrong here, but I seem to recall the 040 is more like 1.2x as fast. And, I seem to remember the 060 being more than 2x as fast as the 040.
-
As far as I remember, there is no 040@50MHz in any Amiga accelerator.
The maximum clock is always 40MHz (42 in overclocked models).
-
Piru wrote:
MIPS is not accurate measurement of processor speed. It's actually far from being 4x as fast.
I thought that MIPS for the Amiga were a more accurate measurement of speed? What was your reasoning? I guess it depends on what we are benchmarking, but I always found it to be about right in terms of how applications perform. The only reasons I wouldn't opt for the 040 are heating and price, so price to performance the 030 would probably be a better buy given the extortionate ebay prices on 040s, if you manage to land one.
-
MIPS only works (is accurate) when comparing same processors against each other. 030 vs another 030, 040 vs another 040, 060 vs another 060.
But then it's equally sensible to just compare the clockrate and forget about the MIPS.
If you want to compare different processors use SPECint and SPECfp, or so.
-
060.....
Kind of hard to find these days.
-
if anyone here believe that the 060 is 2 times faster than a 040 is because never had both cpus to compare
I had an apollo 1240/40 then I replaced by the apollo 1260/50...then I returned to the 040 because the 060 isn't compatible with anything old
I calculate 1.2x faster
The speed betewen both cpus is not noticeable on workbench and not noticeable on most applications or games...
for example...quake on a 040 about 14 fps...on a 060 maybe 15 fps
alien breed 3d 2 ...on a 040 18fps...on a 060 19fps
The 2x times faster comparison was done years ago on magazines..it was a lie for marketing purpose
anyways that comparison was made betewen the 040/25 and the 060/50
-
LaserBack wrote:
anyways that comparison was made betewen the 040/25 and the 060/50
Apples and oranges. Always good for a technical review.
-
for example...quake on a 040 about 14 fps...on a 060 maybe 15 fps
alien breed 3d 2 ...on a 040 18fps...on a 060 19fps
These are quite bad CPU benchmarks, the CPU can't make rest of the system faster.
-
yes they are bad cpu benchmarks
but that's the real-life diference I think
bye
-
Sure. But in other more CPU bound tasks 060 runs circles around 040.
-
And when doing benchmarks, don't forget the speed advantage (or disadvantage) of the RAM bus speed. It makes all the difference in the world. Even between 68030 boards, faster memory access makes for a much more responsive machine. Of course the most obvious historic example was when people switched from C='s A3640 board with the slow as molasses m/b memory to a WarpEngine040 with direct processor/memory access on-board the cpu card. This difference is not so noticeable between different brands of 68060 cards. Phase 5's were generally the slowest with respect to fast ram speed/timing (which is/was good for overclocking), but they also had built in logic to speed up the cpu card to chip ram access speed (mimicking as closely as possible 68030 cycles, iirc) so it was a wash overall no matter which card you chose to buy.
-
I reckon we should do a raytracing test to a stopwatch, without graphics accelerators. :)
-
LaserBack wrote:
if anyone here believe that the 060 is 2 times faster than a 040 is because never had both cpus to compare
Not true at all. The 060 is twice a fast at the same clock speed. And yes, I've had both to compare.
for example...quake on a 040 about 14 fps...on a 060 maybe 15 fps
alien breed 3d 2 ...on a 040 18fps...on a 060 19fps
What you're referring to is the main bottleneck of the system; chipram bandwidth. The CPU is a lot faster, but the RAM is the limit there. Doesn't matter if the CPU is twice as fast - the data still can't get stuffed into chipram faster. You're simply using the wrong kind of benchmarks.
The 2x times faster comparison was done years ago on magazines..it was a lie for marketing purpose
[/quote]
Not.
-
As I've always tested the accelerators I've had with ifx, c4d and lw, here are some numbers, all progs used the same scenes/images with all cpus, highest rendering modes for c4d/lw, gaussian blur for ifx, progs are latest versions, oxypatcher used for 040 and 060:
030@50
c4d: 28m 9s
lw: 23m 55s
ifx: 7m 15s
040@40
c4d: 11m 13s
lw: 9m 44s
ifx: 2m 58s
060@50
c4d: 6m 38s
lw: 5m 29s
ifx: 1m 40s
060@80
c4d: 4m 22s
lw: 3m 41s
ifx: 1m 2s
ifx times are approximate as ifx provides no time feedback and I had to calculate them with a stop watch.
-
I have an 040 Warp Engine card running @ 40Mhz with 128Mb of 60ns Ram. It`s definately faster around the Desktop than a MkII 50Mhz 060 Cyberstorm. However, this doesn`t mean it`s a faster CPU.
As Amithony said above, a real test would be to Raytrace but dont also forget, the 030 doesn`t have an FPU built in as most 040 & 060 CPU's have. Sure, some of these were LC versions & yardy yah yah yah, but if you want to Raytrace on an Amiga, you have got to have an FPU backing up the CPU or you will be waiting like you`re waiting for paint to dry whilst the Render finishes!
The 040 is a HOT chip @ 40Mhz but is floors the 030 @ 50Mhz even with an FPU alongside also @ 50Mhz.
An 060 @ 50Mhz wallops the 040 @ 40Mhz an 060 @ 80mhz just trashes the 50Mhz.
The 060 needed Dual FPU's, then it would have really rocked!
;-)
-
I think an 060 looks like it handles all of them fairly amicably. :)
-
Now a question that some of you will definitely not like,
what about an MC68LC060/75 without fpu ??? Does the lack of fpu justify the higher speed ??
-
Effy wrote:
Now a question that some of you will definitely not like,
what about an MC68LC060/75 without fpu ??? Does the lack of fpu justify the higher speed ??
IMO - a 060 w.o. FPU is kind of pointless. The FPU in the 060 is really fast, which means it's used quite a lot in 060 demos/apps. Later incarnations of the chip can run at 75Mhz (or even 100Mhz), and includes an FPU.
-
Raytracing etc isn't going to benchmark the CPU. Instead you'll be benchmarking CPU + Amiga hardware + OS + Misc software. To benchmark the CPU opposed to "the system" you'd need to write a benchmark that takes over the CPU and runs from memory that doesn't need any waitstates.
-
The reason for using an 68060/75 without fpu is not to use 68060 software but 68020 optimized ... does that make sense ??? :-?
-
@bloodline
MIPS = Meaningless Information Processor Speed.
But yes, a 040 is much faster than a 030 (If you don't use a crappy cbm 3640 because it will have slow mem access).
040/25 is usually at least as fast as a 030/50, but usually much faster. With a 040/40 the difference is huge! and a 060 is great.
-
@Kin-Hell
overclock your MKII 060 to something higher than 56Mhz (e.g. 60Mhz) and you will notice it a lot faster (I've heard it's due to the fact that the turbocard will synchronize better the motherboard)
-
Crumb wrote:
@bloodline
MIPS = Meaningless Information Processor Speed.
I don't see why it's "meaning less". It shows how many instructions can be executed per second. If those instructions are of equal value the processor with the highest value will be "faster". It'd be silly to compare two totally different processors of course.
-
Effy wrote:
The reason for using an 68060/75 without fpu is not to use 68060 software but 68020 optimized ... does that make sense ??? :-?
Nah, you'll just miss the whole point and want to upgrade to a "full" CPU later on. Take my word for it - I did just like you said (albeit at another platform - Atari).
-
Okay then, but who here can upgrade a Blizzard 1240 to 1260 ? I have got already an 68060 for it ...
-
:banana: 68020 rulez :banana:
-
I have had all of these cpu's at some time.
Speed often comes at the cost of compatibility: an 030 will run pretty much anything you want at good speed with an fpu, the 040 will run less stuff, and the 060 less still. TBH the sort of stuff 040 and 060 was made for you can do better and faster on cheap PC running winuae, unless you HAVE to have an Amiga to do it, its your money.
Getting to bench marks: if you run software optimized for it the 060 can be 4 times faster than 40 mhz 040, which is 4-8 times faster than 50 mhz 030.
If you do CPU intensive stuff that uses floating point arithmetic, then you need to have the correct libraries, the correct cpu command and a patch: cyberpatcher for 68004/60 made by phase5, 680040/60.library for GVP and oxypatcher for everyone else. There's also hsmathlibs which speed things up even more for stuff that the other patches don't patch. These patches make a HUGE difference to rendering and image processing. Some phase5 boards actuall speed up chip ram speed as well.
Memory speed does vary and can make a difference: I benchmarked my memory speed of an Apollo 1240@40 mhz in AIBB and it was faster then a published memory speed benchmark for a PPC/68040 board. And i noticed it too: booting was very fast as was copying to ram: disk, faster than my Cyberstorm 2 68060 on my A4000.
My advice: stick with an 030.
-
cantido wrote:
Raytracing etc isn't going to benchmark the CPU. Instead you'll be benchmarking CPU + Amiga hardware + OS + Misc software. To benchmark the CPU opposed to "the system" you'd need to write a benchmark that takes over the CPU and runs from memory that doesn't need any waitstates.
True and this can be done and is often quoted, and you get pretty meaningless results. IMHO its "benchmarking CPU + Amiga hardware + OS + Misc software" that matters to the end user
-
cantido wrote:
Crumb wrote:
@bloodline
MIPS = Meaningless Information Processor Speed.
I don't see why it's "meaning less". It shows how many instructions can be executed per second. If those instructions are of equal value the processor with the highest value will be "faster". It'd be silly to compare two totally different processors of course.
At what point do they become "two totally different processors"? Is an 030 "totally different" to 68060? Well other than running nearly the same 68k code as opposed to x86 code or alpha code, they share nothing else in common in terms of transistor number, pipelines, cache, fpu instruction set, design, number of executed instructions per cycle. Turn the caches off on an 040 or 060 and they execute the same instructions slower than 50 mhz 030. Turn the caches on and they may not execute at all!! AIBB is good (but not for 060) because it does things like rendering a beachball and timing it and yeah mips gives very misleading results.
-
Crumb wrote:
@Kin-Hell
overclock your MKII 060 to something higher than 56Mhz (e.g. 60Mhz) and you will notice it a lot faster (I've heard it's due to the fact that the turbocard will synchronize better the motherboard)
Yea and you then will get unreliable scsi if you have it.
-
stefcep2 wrote:
At what point do they become "two totally different processors"? Is an 030 "totally different" to 68060? Well other than running nearly the same 68k code as opposed to x86 code or alpha code, they share nothing else in common in terms of transistor number, pipelines, cache, fpu instruction set, design, number of executed instructions per cycle. Turn the caches off on an 040 or 060 and they execute the same instructions slower than 50 mhz 030. Turn the caches on and they may not execute at all!! AIBB is good (but not for 060) because it does things like rendering a beachball and timing it and yeah mips gives very misleading results.
I think we need to make a distinction between CPU(+memory) benchmarks and system benchmarks. The former is of interest to those using CPU intensive stuff like rendering, while the latter takes system bottle necks into the equation.
Personally, the mips figure makes sense to me since I get a rough idea of what kind of performance my code can get when only accessing RAM.
-
@stefcep2
It shouldn't cause problems at 60Mhz.
-
Crumb wrote:
@stefcep2
It shouldn't cause problems at 60Mhz.
i read somewhere that anything above 50 mhz is a problem. Its years though, but i remember thinking about doing it, and was advised against it for this reason
-
shoggoth wrote:
stefcep2 wrote:
At what point do they become "two totally different processors"? Is an 030 "totally different" to 68060? Well other than running nearly the same 68k code as opposed to x86 code or alpha code, they share nothing else in common in terms of transistor number, pipelines, cache, fpu instruction set, design, number of executed instructions per cycle. Turn the caches off on an 040 or 060 and they execute the same instructions slower than 50 mhz 030. Turn the caches on and they may not execute at all!! AIBB is good (but not for 060) because it does things like rendering a beachball and timing it and yeah mips gives very misleading results.
I think we need to make a distinction between CPU(+memory) benchmarks and system benchmarks. The former is of interest to those using CPU intensive stuff like rendering, while the latter takes system bottle necks into the equation.
Personally, the mips figure makes sense to me since I get a rough idea of what kind of performance my code can get when only accessing RAM.
Not a coder but i get your meaning.
-
Had a Blizzard 1260 at 66 Mhz, speed by made by DCE, no scsi problems ...
-
That might be true for the DCE A1200 boards. My experienmce was with the cyberstorm 2 for A4000 with cyberscsi module. i am pretty sure that with the cyberstorm boards overclocking caused problems with the the scsi board.
-
They mostly work with 60 MHz, but may fail if clocked higher.
-
stefcep2 wrote:
Crumb wrote:
@Kin-Hell
overclock your MKII 060 to something higher than 56Mhz (e.g. 60Mhz) and you will notice it a lot faster (I've heard it's due to the fact that the turbocard will synchronize better the motherboard)
Yea and you then will get unreliable scsi if you have it.
No, you wont if you overclock your CS PPC correcty.
These card shipped with only 2 of the 3 Oscilator Sockets populated. By changing a tiny resistor pack from pin 1-2 to 2-3, (under & to the right of Simm 2) it forces the 060 or the 040 on the CSPPC to use the Third Oscilator socket. This in turn leaves the SCSI & the Memory running on Oscilator #2 which is 50Mhz. The 1st Oscilator (most right) is the PPC Crystal.
Incidently, I have replaced the 50Mhz SCSI & Ram Oscilator with a 64Mhz Crystal on mine. This takes these buses from 25Mhz to 32Mhz. What a nice speed boost that was & with the 060/50A @ 80Mhz, she flys......
Plenty of Air required though & a good hefty PSU. ;-)
-
Just to add to the Original question, the real difference between a 030 @ 50Mhz & an 040 @ 50Mhz = Some Years of Progress from Motorola.
Any CPU without an FPU (Floating Point Unit) is gunno take far far longer to calculate Floating Point Instructions. These are used by Programs such as CAD Packages, DTP where Image manipulation is required & more specifically, 3D Modelling.
So taking the more common Motorola CPU's used by CBM for their range of Amiga puters, here they are in rough running order of speed.
68000, 68020 & 68030. I pause here because all these CPU's never came with onboard FPU's, but you could either fit a PLCC or a PGA Socket to take the required kind of FPU to be fitted.
Then we have the 680040 & the 68060. These came in a variety of speeds & also a variety of flavours. Speeds on the 040 were usually 25Mhz, 33Mhz & 40Mhz. 68040's were rare at 50Mhz due to Heat constraints within the silicon for using 5v on that Die size. When the 060 was released with a new Die Size, the Voltage was also dropped to 3.3v, meaning a 50Mhz 060 is actually a much much cooler CPU than a 40Mhz 040 & absolutely spanks the 040 @ 40Mhz when it comes to CPU & FPU calculations.
Going back to the Varieties, these were annotations used for what I class as "Deprived" CPU's, as they had either no FPU & worse still an MMU (Memory Management Unit)
Back in the traditional Days of Amiga, SCSI was always favoured because of the way SCSI Works. It off-loads all the CPU Transfer calls required by the hard drives to the hard drive controller itself. IDE does NOT do this on an Amiga. An MMU was required by "Unix" on the Amiga & the only after-market program I know of that required an MMU was "Giga-Mem". This was a Virtual-Memory program for the Amiga, where it used Hard Drive space as "Virtual-Ram". Try using that program on IDE drives instead of SCSI; - er...."Hello clockwork computer" when the Virtual Ram was being used! :roll:
EC & LC annotations in the CPU part number mean it`s a CPU that probably failed all full Tests when Soak Tested after the Silicon Wafers had been processed into their relevant Part. For Example, if a batch of 50Mhz 060`s were made, they would all be tested for CPU, FPU & MMU calculations. If the MMU failed, they went out as an LC marked CPU - (Low Cost). If the FPU failed, they were shipped as EC CPU`s or as in the case of the A4000, the 030's were always Badged EC030 anyways & the 030 never had an Internal FPU. CBM were numbering But there was provision for either a PLCC or PGA Socket to attatch an FPU.
Whilst Soak testing the CPU's, they were also checked for operating speed. Purer slices of Silicon would run faster & hence the variety of speeds. The faster & more feature rich ones being more expensive of course.
On top of this, the market also has demands for supply. On some occasions, you might be lucky to have an EC060/75 that will Under-clock to say 60Mhz & the FPU works. That makes it a faster CPU than an 060 @ 50Mhz.
Hope this all helps & makes sense as to which type of CPU is better to have. :-)