Tom's Hardware Wants You: CPU Benchmarks 2011/12

Page 4 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
I know it's suposed to be a cpu bench, but still i question the choise of those 4 games mentioned in the list. In the system builder marathon of december 2010, this site used an Athlon II X3 (oced to 3.8ghz) in combination with a nvidia 460 (768mb).
This config scored 38 fps in F2010 on ultra settings with 8x MSAA.
On those same settings my Phenom II X3 (oced to 3.2ghz) in combination with a 5770 i get just over 45 fps on average. With dips to 40fps and spikes to 52fps.
What i'm trying to say is, a 3.8ghz atlon II should be faster then a 3.2ghz phenom II and the same goes for the 460(768mb), it should kill my 5770.
My point, the games you picked all favor ati cards. Kinda make comparing cpu's hard.
 
I really only saw one other person mention this, but I'd really like to see World of Warcraft (dx11) added to the regular benchmarking suite simply because it's one of the most played games out there, and often these days people look at benchmarks for the specific games they play the most since it's pretty well established at this point that some games simply prefer one architecture to another.

Also while it's true a standard benchmark in WoW is generally not particularly taxing, a standard flight path benchmark (which I believe is really the only way the game can be benchmarked) is not what is taxing, it's 25 people with 25 spell effects going off as well as environmental scenery and effects and 50 npc models running around etc. on a 2560 by 1600 resolution with 4x AA that suddenly takes your i7 crossfire setup to 20 fps. The nature of mmo gameplay and how widely it can vary means that under some situations it can be extremely taxing, far more than the average FPS on a benchmark may indicate, but at least that flight path benchmark will show how well it scales on different architectures and newer cards.

This likely applies more to the GPU side of benchmarking than GPU with the new DX11 pathway, but I believe WoW and Blizzard's games in general (hi SC2) are notoriously CPU dependent compared to other modern games.
 
[citation][nom]justme1977[/nom]I know it's suposed to be a cpu bench, but still i question the choise of those 4 games mentioned in the list. In the system builder marathon of december 2010, this site used an Athlon II X3 (oced to 3.8ghz) in combination with a nvidia 460 (768mb). This config scored 38 fps in F2010 on ultra settings with 8x MSAA.On those same settings my Phenom II X3 (oced to 3.2ghz) in combination with a 5770 i get just over 45 fps on average. With dips to 40fps and spikes to 52fps. What i'm trying to say is, a 3.8ghz atlon II should be faster then a 3.2ghz phenom II and the same goes for the 460(768mb), it should kill my 5770.My point, the games you picked all favor ati cards. Kinda make comparing cpu's hard.[/citation]

Considering that Phenoms have a fair amount of L3 cache, and Athlons have none, I disagree with this statement.
 
I would like to see 3D Studio MAX with a professional renderer like Vray
I would also like to see some short of engineering software test too, like Solid works, Pro/E, Ansys, Nastran.
 
You should give more focus to CPU/Multi-threaded bound games. That's the real concern regarding games, how many cores to buy at any given time. And a Benchmark dedicated to single thread performance, just to see the evolution between architectures(to evaluate a potential upgrade). Keep up the good work.
 
I would like to see simultaneous benchmarking between different applications. Run a game while your running matlab / maya or some media encoding applications.

This would stress the ability of the system to handle multiple independent programs. With the way everything is going parallel, dual channel memory, dual GPU, SSD RAID, multicore CPU's and such I believe it's important to test out how a system scales while using multiple independent resources.
 
Battlefield 3 and Diablo 3 when they come out. for now, Battlefield: BC2!
 
+1 on the Matlab recommendation.

My game vote would be for NFS Shift2 Unleashed. It really highlighted the performance difference of Intel Sandybridge over AMD Phenom II. Developers (Slightly Mad Studios) even went on blogs to highlight the inferior "instructions per cycle" of the Phenoms and the subsequent lack of performance for some players.
 
[citation][nom]goodguy713[/nom]im sugjesting running a standard hard drive system and a ssd bases system in comparrison through your normal bench tests .. plus what ever you come up with there it would be interesting to see how negitive of an impact runing say western digital black 64mb cache drives or velosa raptor drives vs ssd drives on the new processors when they are finnally released i bet it would help a lot of people make there choices[/citation]

Not that im bashing WD or any other hard drive maker it just seems like it would be a great way to test the newer processors i firmly think that the standard hard drives are limiters.
 
1) less synthetics - who cares how much imaginary points are scored in 3D Mark?
2) More real life stuff: adding SC2 would be nice and adding compilation tests for Java and .NET are absolutely nessesary - people actually use these things every day, not just boast about the millions of some points scored. I don't care about how much points I receive in Vantage, but I do care how long does it take to compile a project
 
[citation][nom]stephenoh33[/nom]I suggest to simulate real-life experience simulation rather than synthetic performance benchmark. [/citation]


So... we should add benchmarks for how fast we can load porn? Because we all know thats all we use computers for... gaming and porn... lol
 
I will also add the motion for Starcraft 2. It was the first thing that popped into my mind when I read the first sentence.
You must also be able to show if physical cores are better or if Hyper Threading is better.
 
i think TF2 should be! why? well valves games are one of the most played games. since TF2 went free you will have a lot of people wonder can my CPU run TF2?
 
I'd like to see World of Warcraft in the list.

Suggested FPS killer tests:
1. CPU Crowded zone - Tol Barad battle
2. GPU Open world zone - Twilight Highlands

 
Run metro 2033 on max while defragging while running anti-virus scan while recording the game on fraps while having 10 youtube videos looping in the background in 10 different windows and while mincraft runs in the background.
 
0. single threaded things like copying text from a 3MB PDF file.
1. crysis 2 as it is quite cpu dependent.
2. +1 for matlab.
3.+1 for simultaneous programs running. like running SC2 with a video conversion running in backround.
4.some kind of antivirus scan. kaspersky internet security 2011 routinely uses 100% CPU for minutes at a time.
5.BF3 :)
6.k series SB CPU's at a "modest" speed of 4.2 hz.
7.real cores vs HT
8.games at resolution of 1080p as its more realistic than 800x600.
9. searching in windows. probably single threaded.
10.finding and replacing text in MS-word or libre office etc. again, single threaded.
 
[citation][nom]oxxfatelostxxo[/nom]So... we should add benchmarks for how fast we can load porn? Because we all know thats all we use computers for... gaming and porn... lol[/citation]

I was able to tug one out .534 seconds quicker when running the fusion system
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom