Discussion in 'Hardware' started by Bloody_Pete, 14 Dec 2011.
The majority speak.
I can understand the argument, but I've just fought tooth and nail to achieve the score I have, should I be passed by someone with a worse processor, at stock, using Linux? Should they, on the other hand, also run the benchmark within Windows so that it's a fair and relative example of how fast their machine is in relation to mine? You know; there's a reason benchmarks like 3Dmark and the like only run on Windows (Besides DX support.)
It's a complete unbalancer. Windows machines are not being compared to mac here. Although in theory; they can, that's the point of Geekbench, but at the same time; Running windows, as we've show, gives the system a different performance balance than Linux. Linux produces higher scores, but at a different offset. I believe it's down to the kernel, but all the same; It's an unfair advantage. Eventually the winner would be whomever had the fastest machine (Give or take 5%) that managed to get Linux to run at command line, with only networking and Geekbench in order to record the absolute best score they could.
People like me, who currently can't go back to their best score, Or no-longer have the hardware, would be unfairly unseated by this new horde of people going "USE ALL THE LINUX." while benching, I can accept being outdone by better hardware, or worse hardware clocked much better, but not surpassed because i'm unwilling to run Linux on my system, simply because of the timesink it would be. I can hit a button on boot and be at 5ghz. I can't hit a button and have linux without having to first download the OS, install, configure, refine, then bench. The majority of people use windows for their tasks, why should they suddenly up to Linux to bench?
Cant we split the table to show linux systems as well? Then you could have individual table for each, and maybe a joint table?
Il be quiet now
No.. it's enough work as it is. We've split Windows into 32 and 64 bit already. To have separate tables for Win 32, 64, and Linux 32 and 64 and Mac... that's just a nightmare.
Seriously... if we're all using the same OS.. it really doesn't matter.
If you want to take on the responsibility of maintaining the boards... I'll pass it over to you... but like I said., it's often a lot of work, and to make it more work for the benefit of what seems to be 2 people is not something I'm prepared to do.
Its cool i dont care if its done or not, i was just suggesting something. I understand its a lot of work and i do appreciate it.
I plan to try my system out when ive finished re-building it
but all the objections already mentioned support my case
you dont want to incorporate linux because it gives an "unfair" advantage, yet your willing to allow OSX - which is essentially a fancy GUI over a BSD kernel.
in a lot of regards OSX is more like linux than windows, but your still allowing it whilst claiming that this is a purely hardware based comparison.
You still don't get it.
OS-X isn't giving an unfair advantage to any PC users is it.
In order to enable anyone to participate, we are using the de facto OS for each platform - Win 7 for PC, and OS-X for Mac. That way, anyone who own a PC, or a Mac can participate without having to install a different OS.
No one will be making direct comparisons between PC and Mac, but having one common OS per platform DOES allow people to compare Mac against Mac, and PC against PC. Simple.
Whether Win 7 is the "best" OS is neither here nor there, as everyone will be using it in this thread, so like for like comparisons are still valid. If you want to see how your i7-920 on a Asus P6 board compares to others, you can, easily. However, if there are 920s on there running a Linux distro, it skews the figures and no direct comparison can be made to yours.
I know you like Linux... but because you like something doesn't always mean it should be included. Linux is for a minority, and this thread is for everyone.
If you're so bothered, I can send you the excel spreadsheet, and you can take over the administration of this thread if you like, but I'm not adding more and more categories to it just to please a couple of people.
LOL@ new tag addition
You should be thankful there's no Linux listings anyway, because now Simon is in the land of laptop, I'd dry butt**** you all in this thread if I did a geekbench run under Linux!
[edit, edit, edit]
BTW... No one is against posting Linux runs up in here... they just don't go on the boards. I've posted several, as have others. It's interesting to see them, and no one in here is against Linux per say. It just makes no sense to confuse a standardised system by allowing them on the scoreboards.
OMG! Maybe benching my macbook wasn't such a good idea. Talking about can of worms, cat among the pigeons etc.
But anyhoos none of it matters because my new laptop is #1 when it comes to 32bit efficiency! It may not be as fast as some of you guys with your hexa-core bios tweaked beasties but when it comes to running pointless synthetic 32bit benchmarks and cooking the numbers to compare score/clock speed while strangely ignoring core count it's currently unbeatable.
Wow! That's so cool! It means absolutely nothing, proves absolutely nothing and does absolutely nothing... which is kind of fitting as I did absolutely nothing to this machine other than unbox it and start it up.
Well I'm glad this my mac being at the top of this one chart has finally put a lid on the whole "Macs vs PCs" debate once and for all and confirmed that macs are better*
*Joking. My PC is still a PC, and even this Mac has a win7 boot install for gaming.
Edit: By the way, macs usually give higher geekbench scores running windows than OSX. And while Macs come with OSX installed, they are designed to run Windows too and come with all the software other than the OS licence and media.
So in the terms of 'like for like tests' wouldn't my Mac running windows be the same as a custom built machine with self installed OS? Neither OS came with the hardware, both are supported with device drivers from the hardware manufacturer etc.
Surely my laptop running on it's win7 x64 partition would be a better comparison against say a comparable spec Dell / Alienware running windows? If we're going to bench on different OS platforms.
So in the interest of fairness, I say... Allow mac on your list but they must be benched in OSX and in windows! They get their place on the table based on the highest score (but the lowest score must also be published)
So each person with a Mac goes on there twice? LOL Then the PC owners will say "So I should have my Linux score up there then"... and on it goes.
No. Macs run OS-X, PCs run Windows in this test. No one is comparing Macs to PCs in these benchmarks, so why insist they run Windows? Not every Mac owner owns a copy of Windows, so again, how is it fair? Those with a copy of Windows they can install will always beat someone else with the same Mac who doesn't.
And why should it factor core count into the equation? What possible use would knowing how fast each individual core is be? We had this argument months ago when someone else upset that they couldn't beat 6 core chips spat their dummy out, and I ran Super-Pi (a single threaded benchmark) and still beat the vanilla SB chip anyway. so what's your point? What next, let's have a 4 core leaderboard, then a 6 core leaderboard... and when IB-E comes out shortly, a 8 core leaderboard?
Where do we stop? Each individual CPU type having it's own leaderboard?
To be honest, there's little intrinsic value in having even the efficiency scores either, but they're there, and removing them is as much work as keeping them to be honest as when I input a new score, it automatically works it out.
...and that's all you ever will do to it... seeing as it's practically impossible to upgrade
Now quit bitching all of you because at the end of the day they're only.....
No, definitely not. One machine, one placing.
Because the 'efficiency' score is currently biased towards multi-core processors. It's not a measure of how 'efficient' a system / processor is. The most 'efficient' dual core machine imaginable could never beat a slightly less efficient machine if it had more cores. If you're going to have two scores one for 'fastest' and one for 'most efficient' then surely the efficiency one should be configured to identify efficiency? (We're playing word games here, so please tell me to shut the eff up and climb back under my rock!)
None at all!
My point exactly!
32bit and 64bit ran with version 2.3.4 CPU @ 4.2GHz (Until this Heatsink turns up...)
Will update Tuesday... will be away for a few days.
Score went up for no reason apparently:
Are the specs in your signature correct? Are you really pushing that on a 330W power supply?
It's a laptop
I stopped using Geekbench to measure anything when I released how incredibly inconsistent and biased it is. Other benchmarks (and common sense) don't tally with the results.
Yes, all from a 330W PSU that was supplied with my laptop, which is what Dell gives with any Dual GPU laptop.
So yes very impressive all that power from a 330W PSU.
It all depends on what you have running in your task manager, which is why you probably fine it fluctuates so much.
i got geekbench purely so i could post itt
such is life
The score can vary a little, which is why it is best to do three runs. I find that results taken in that manner are pretty consistent.
Separate names with a comma.