• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Idea for rating Processors.

L

LeeTJ

Diamond Member
I was just thinking of the whole Mhz thing and how inaccurate it was especially in relation to each other.

for example a 1 ghz P3 vs a 3.06 ghz p4. i can't imagine many applications where the 3.06 p4 will actually be 3 times faster than the 1 ghz P3.

so here is what i was thinking.

the 1 ghz P3 and the 1 ghz AMD (pretty even in performance) to me was the most signficant speed increase in a long time. here is my reasoning, a 1 ghz P3 / AMD processor now, even tho 1 or 2 generations behind the most current processors available can carry out a larger percentage of tasks for typical users and moderately high end users than ever before, and can carry them out reasonably well.

I could flesh that out a bit better but i'll let you guys fill in the blanks.

here is my idea, why don't we establish the 1 ghz P3/amd processor as the BASE, give it a rating of 1 for example and then rate all other processors based on their ability to performs tasks in relation to that base of 1.

i hope this makes some kind of sense.

what do you guys think?
 
Originally posted by: Harabecw
i can't imagine many applications where the 3.06 p4 will actually be 3 times faster than the 1 ghz P3.
Click to expand...

Well...

Almost!
Click to expand...

Yeah, but in the first case the benchmark was run with a first generation GeForce 256 DDR and in the second it was run with a Radeon 9700 Pro.....not exactly isolating the CPU.
 
PR ratings is flawed in general.

You would need a standard system setup consisiting of standard Harddrive, Memory, Video, Sound, Motherboard, Drivers. Once you get sorted through all that crap, you have to pick the software in which to benchmark.
 
I remember reading/hearing somewhere the Microsoft is working on a "rating system" for machine spec's, and that instead of going to buy, say, UT2k3 and being presented with:

System Requirements
Operating System WIN 98/ME/2000/XP
CPU Pentium III or AMD Athlon 733mHz processor
*Pentium or AMD 1.0GHz or greater (RECOMMENDED)
Memory 128 MB RAM
*256 MB or greater (RECOMMENDED)
blah blah blah ...

You would instead have:
System Requirements
MINIMUM Class-III Computer
RECOMMENDED Class-I Computer
(or whatever)

as part of their push to make the PC a more viable gaming platform ... this is along with them wanting to come up with a standard controller (ie, have six buttons, two triggers, and a d-pad) to help in that area as well. I WISH I could find/remember where I read about this, but I can't seem to. If someone else can, I would appreciate a link being posted here! 🙂
 
The PC is a viable gaming platform. I do better on a PC than a console, and it's more fun/possible to tweak. This is more of MS's dumbing down of computers IMO
 
Originally posted by: LeeTJ
I was just thinking of the whole Mhz thing and how inaccurate it was especially in relation to each other.

for example a 1 ghz P3 vs a 3.06 ghz p4. i can't imagine many applications where the 3.06 p4 will actually be 3 times faster than the 1 ghz P3.

so here is what i was thinking.

the 1 ghz P3 and the 1 ghz AMD (pretty even in performance) to me was the most signficant speed increase in a long time. here is my reasoning, a 1 ghz P3 / AMD processor now, even tho 1 or 2 generations behind the most current processors available can carry out a larger percentage of tasks for typical users and moderately high end users than ever before, and can carry them out reasonably well.

I could flesh that out a bit better but i'll let you guys fill in the blanks.

here is my idea, why don't we establish the 1 ghz P3/amd processor as the BASE, give it a rating of 1 for example and then rate all other processors based on their ability to performs tasks in relation to that base of 1.

i hope this makes some kind of sense.

what do you guys think?
Click to expand...

The question I think here.. is Processor performance can be *drastically* influenced by chipset,

The origonal palomino. Put it on a KT133A, then score it. Then move up to an Nforce2. What does it score now? IN the majority of gaming cases, 30% faster.

There is no standard way to compare processors is because there is no fair way to compare processors.

What happens when you bump a P4 from an i845SD to an i875? HUGE performance increase. Even if you just move from the i845 to the i875 it'd be enough to change the score by quite a margin.

What graphics card do you use to score it with? What harddrive?

My point is, a systems infrastructure has a far greater impact on performance than the CPU in use. Watch an AthlonXP 2100+ provide the beatdown on a 2600+ if the 2600+ is using a KT133A and the 2100+ is on an Nforce2.

There is no way to score CPUs like the way you're talking about.
 
I think it should just be a set of a few numbers designating how fast the cpu performs in specific standardized benchmarks. They would measure the major aspects of cpu speed such as FPU performance. You'd end up with something like 10,15,5,25 each number is for a specific benchmark.

If we did something like that instead of comparing a new processor to old ones, we would have a much better idea of what performance it has. The one major problem with this I see is how to standardize the platform the cpus are tested on.
 
could there be a mathematical approach to this??

design a benchmark(or benchmarks) that is CPU only and record scores for each processor
pick the highest end processor and test it on each platform (nforce2, kt600, 865, 875, etc.) and record scores for each platform
pick the highest end processor and platform and test a range of video cards and record scores
pick one hard drive and test it on each platform to test each platform's IDE performance and then use the fastest IDE platform to test individual hard drives and record scores

when u finish, u have a mathematical system with which u can look at your machine's CPU, platform, video card and hard drive and be able to generate some kind of idea of how it performs. this kind of approach will let buyers know what they're getting. i get peeved when i see a company advertise a computer based solely on its processor. they don't mention it has crappy integrated graphics, a 2 year old hard drive and a slow platform on which to run.

using this approach, we can see what kind of total computer one has. however, designing benchmarks for this system would be tough, but i think it could be done semi-accurately.

ideas for processor benchmarks: software 3d renderers, prime number computations, scientific benchmarks
ideas for platforms benchmarks: content creation, gaming tests
ideas for video card benchmarks: in-game real-time scores on latest games using a wide variety of IQ settings
ideas for hard drive benchmarks: current synthetic benchmarks, file copying benchmarks

these are just rudimentary ideas, but i think a system like this could work. your guys' ideas?
 
Of course the ratings don't make sense, and naming in general doesn't either (e.g. GFFX 5200 Ultra slower than GF3 Ti200). The purpose of every company out there is to name their product specifically so as to decieve the buyer, make their product seem better than it is and sell more units. End of story. That's what makes computer building and tweaking an art moreso than a science. As long as there is competition none of these companies will standardize anything, ever.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top