Sign in with
Sign up | Sign in
Your question

CPU Normalized Perrformance Rating charts

Last response: in CPUs
Share
May 3, 2008 10:41:09 PM

I created some supplemental charts for Tom's Hardware CPU and GPU charts.

http://emsai.net/reviews/

Who knows, maybe Tom's will add something similar too. At the very least they got easy access to a lot of cross-referencing data over the years. ;) 

(from my journal entry today)
A new Reviews section has been added, and two very new cool additions under that are the CPU and GPU normalized performance rating charts, with a very impressive upgrade adviser integrated into the chart.

The Normalized Performance Rating (NPR) charts uses the data of Tom's Hardware CPU and VGA charts, 31 of the former result tables and 43 of the latter result tables was painstakingly normalized, merged, verified, and can now be seen in all their glory. CPU Chart

These charts was made out of my own need as Tom's Hardware lacked anything similar, sure there is the FPS table in their VGA chart, and the price/performance index in their CPU chart, but neither was suitable for my purpose. I needed to compare all CPU's and GPU's against a baseline so that I could check possible upgrade choices, and see a range of potential upgrades.

It is also very interesting to see how similar the performance increase are for both CPU's and GPU's, and by looking at the charts it is obvious that new does not always equal better. Since my charts are based on re-purposed data from Tom's Hardware, my two chart tables are also available as CSV files if that is of any interest to anyone.

It would be really nice if the industry used a normalized performance rating system like this as it's so easy to see what hardware has x amount of improved performance over the other. Leaving the consumer to worry about cost and features only. Now if game/software developers hopped on board that would be even better.

Yes I know that there is the Windows Performance Index, but it's not as fine grained as my NPR (only two fraction points are shown, the CSV itself has up to 15), you may also think that there is no point with all the benchmark and chart sites out there. But look at my charts and you see it's all floating point math which can so easily be extrapolated, secondary baselines like the upgrade adviser does is very simple and fast to implement.

I will update these charts now and again, and if popularity gets really high I will obviously update them more frequently. I may or may not add older CPU's and GPU's, it all depends on if I can find some solid benchmark data someplace that I can normalize then re-normalize against the NPR baseline and finally add them to the charts.

Enjoy the new Reviews section, I have a few future plans that might be just as interesting.
a b à CPUs
May 3, 2008 10:46:09 PM

Good work. You might want to change the colors (esp. Brown) to some thing more visible.
May 3, 2008 11:30:08 PM

I applaud the efforts. Very neat idea for sure.

But I question the results and usefulness (as I do the charts themselves if you look at just the overall performance). I've long felt the charts can be very misleading, and it shows itself in bad forum recommendations. This, while a nice concept, may make it easier to get a bad impression of how the GPU's stack up in current games.

For instance, it tells me the only thing better than my SLI 8800GT OC combo is Crossfire 2900XT? Based on what, 3dmark06? Overall gaming total including low res no fsaa? For gaming, 8800GTX SLI is better than my combo, and the dual 2900XT is far worse.

Also, it has the GF7's ranked way too high for current gaming trends. The 7900GT is now far behind the X1950 pro if you look at UT3, Crysis, Call of Duty 4, Oblivion, etc. Someone could look at these and think a card that struggles in new games is a good used choice, when a brand new, cheaper, better card could be found for their needs.
Related resources
a b à CPUs
May 3, 2008 11:32:41 PM

Yeah, charts like those are just too difficult to keep updated as new drivers, games, hardware, etc. come out and mess them up. It seems that THG is only halfheartedly updating then now too.
May 3, 2008 11:43:23 PM

Instead of editing again, I'll say the same for the CPU's. Someone will bring it up if I don't. ;)  It's general performance, but doesn't necessarily equate to better performance for the individual needs.

For example, It tells us a QX6800 is a green step up from the e8500. If all a person does is play games a QX6800 is not a step up, and either will be excellent. Many understand this no problem, but surely a large number of Tom's readers do not.

Anyway, not trying to bash your efforts. I hope you understand where I am coming from. I actually like the idea and think it's clever and quite amusing. My only alarm is that just feeding noobs (no offense to anyone) overall scores is a dangerous thing.

Maybe there should be a warning about this at the top just to explain how to use/not use the info. At least stating more clearly specifically where the numbers came from (overall total score?)
May 4, 2008 12:39:31 AM

Good points. It actually does state what the numbers are based on.
But that line is at the bottom (along wit the CSV file which the displayed form is generated from in case anyone want even more specific number (14 decimal places)

Both the NPR charts use all the benchmark tables of Tom's, excluding a few as indicated in the text at the bottom. Mostly due to them being overall states as well or somewhat messed up. The HDTV one is really weird for example. So it should be roughly 50% synthetic + 50% game/apps.

I would never trust a purchase decision on a jumping +7 point like that. (there's one in the GPU chart too) and due to the way the advisor is adjusted, neither does it. That +7 one is never green so not advised at all.

If anything the advisor highlights (blue) that this may not be worth getting, ironically.

I really appreciate all this feedback though, and I'll see about writing up a brief explanation of the actual numbers/basis that isn't so wordy heh.

As to the data itself. I'd rather not remove a entry just because it scored so high on synthetic tests, I'd rather leave a note that certain entries score rather high on synthetic tests and may act differently under everyday use.

Just wish I did not have to emphasis that people use things like common sense, and think/look first.

As I said the idea is this is a quick lookup chart. People see a entry that is interesting then head to Tom's for further review. I did ponder linking each entry, but that would be a maintenance headache and I'm sure Tom's web/DB designers can do way better than what I can come up with :) 
May 4, 2008 12:46:32 AM

Maybe the days of synthetic benchmarks are dying?
A synthetic benchmark could be compared to a very well optimized application designed to do the same thing over and over again very quickly.

Most apps/games these days are rather on the bloat and not always that well optimized.

I don't plan to split the synthetic/normal data, if worst comes to worst I may just dump the synthetic data rather than split it. After all, Tom's has the split data in the first place.
May 4, 2008 2:08:30 AM

Well I certainly feel stupid now. It turns out that the GPU datafiles here had a duplicate of HD2900 XT CF, originating from Tom's own charts where this card is dual listed for some odd reason. The CPU charts seem unaffected, at least nothing as odd as this.

The csv builder has had a few checks added and a new GPU chart generated, after checking all benchmarks again, ratings now match what is expected.
a b à CPUs
May 4, 2008 10:20:22 AM

Well I think you have done some great stuff there ... so well done Rescator.

If you could get the Phenom B3's slipped in it would be great too.

Also, the table is narrow enough for another column ... perhaps the cache size would be a good addition??

Cheers :) 
May 4, 2008 10:47:33 AM

I find it interesting that the NPR for the e2xxx series is pretty much just the core clock/1000. That means that OCing the e2xxx series by boosting the FSB and keeping the stock multiplier should give an NPR equal to the new clock/1000. Thus, my e2160 OC'd to 3GHz should have an NPR of 3.0, matching a stock e67xx (w/2.66GHz core). Its NPR would be about 4% under a stock e6850 (NPR 3.12), which would have the same core and FSB clocks, so the NPR difference is pretty much solely attributable to having 4MB cache instead of 1MB. By contrast, an e4300 OC'd to 3GHz should have an NPR of 3.3 (higher than the e6850!), even though it has 2MB cache vs. the e6850's 4MB. Clearly, differences of 0.2 on the NPR aren't really significant, and cache size differences seem to make a 4-12% difference (in the range of 1MB to 4MB cache).
On a side note, I find it amusing that my e2160 OC'd to 3GHz and my e6420 OC'd to 2.66GHz end up with almost exactly identical NPR scores (the former cost $70 in early 2008 and the latter $175 in early 2007).
May 6, 2008 7:19:06 AM

The NPR charts for CPUs and GPUs have been updated.
7 new CPUs (AMD), and 6 new GPUs (ATi and Nvidia).

The charts can be found in the Reviews section as always.
http://EmSai.net/reviews/

New additions for CPU's are:
AMD Phenom X4 9850 BE
AMD Phenom X4 9750
AMD Phenom X4 9650
AMD Phenom X3 9550
AMD Phenom X3 8750
AMD Phenom X3 8650
AMD Phenom X3 8450

The CPU and GPU charts has been cleaned up as well,
OC variants has been removed, so has various stepping variants,
the latest stepping variant was kept in case of duplicate entries.

The information in the charts has changed to, no more Megahertz,
instead it's all about Cores now as this does not just affect performance but multitasking ability as well.

A feature column has been added as well as the features a GPU or CPU is just as important as it's performance.

The header of the Feature column has two lines in it called All: and None: this is a simple way to inform shared features or non-existent features that applies to all entries in the chart.

A small note about CPUs and SSE support. Recent AMD CPUs has SSE4a but this is not related to SSE4 nor SSE4.1 nor SSE4.2
Future AMD CPUs will introduce SSE5 but this is not related to SSE4 or 4.1 or 4.2.

To make matters even more confusing a few of the Intel CPUs support SSE4.1 but not SSE4, future Intel CPUs will include SSE4.2 thus SSE4.1 + SSE4.2 will equal full SSE4 support, it is unknown at the moment if Intel plan to support SSE5.
It is rumored that AMD might support SSSE3 and SSE4 later.

I guess the good news is that at least all CPUs in the chart support SSE3.

Enjoy the charts and as new CPUs and GPUs are released I'll update the charts as often as time allows.
!