• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

hardwareheaven shows BD > 2600k for gaming

thing is though for gaming, I would like to see if the CPU helps. Ie use 1920x1080, give us a 6970, then cf, then 6990, and with the nvidia side, 580, 580 sli, 590.

this way the gpu can be seen if its limiting or not, and with the more taxing systems if the CPU comes into play.


so, for example:

2500k vs fx8150

6970, 55fps vs 55fps
6970cf 100 fps vs 70
6990 90 fps vs 70 fps.

this would show real world gaming, and that if down the line you upgrade your gpu or add another if there is a bottle neck.
 
Last edited:
It's not just HardwareHeaven - here's another result that shows BD ahead of an i7-920 in most games, albeit overall it's close -

http://alienbabeltech.com/main/twen...-6970-6970-crossfire-vs-phenom-ii-and-core-i7

FX-8150 does do well in games. At stock it smokes my i7-920 at 3.8GHz and the Phenom II 980BE at 4.3GHz.

Charts:

http://alienbabeltech.com/main/wp-content/uploads/2011/10/PerfSummary-1.jpg
http://alienbabeltech.com/main/wp-content/uploads/2011/10/PerfSummary-2.jpg

So why are his results so good as well?
 
PS: I was under the impression the i7 920 is triple channel and FASTER than an i7 2500k? Or is it the other way around?

The one above is clocked at 3.8GHz, so no slouch either. :confused:
 
The i5 2500K is faster than the i7 920 (the triple channel RAM doesn't really matter as sandy bridge handles it's dual channel memory very efficiently) - please see this comparison. However looking at those charts he is running a single 6970, some high-end CPUs and using high resolutions/setting - so it should be no surprise that the results for the i7 920 and Bulldozer are about the same (within the margin of error) as it looks like he was GPU limited (much like the HH test) so he isn't able to max-out either CPU and see how they truly compare.

If he reran the tests with three 6970s in Crossfire or with the same GPU and much lower settings then I would put a lot more faith in his results. I see he has results for 6970CF at least, and in these tests the results for i7 920vsBD are either about the same (GPU limited still) or the i7 920 wins by a fair margin.
 
Last edited:
It's not just HardwareHeaven - here's another result that shows BD ahead of an i7-920 in most games, albeit overall it's close -

http://alienbabeltech.com/main/twen...-6970-6970-crossfire-vs-phenom-ii-and-core-i7



Charts:

http://alienbabeltech.com/main/wp-content/uploads/2011/10/PerfSummary-1.jpg
http://alienbabeltech.com/main/wp-content/uploads/2011/10/PerfSummary-2.jpg

So why are his results so good as well?

His results are GPU bound, several of his game results are exactly the same regardless of CPU, or clock speed.
By his tests, if you just game with one one card, then then bulldozer is pointless if you already own a phenom.

Serious sam, no difference regardless of CPU or speed.
Crysis, no difference.

When you look at the crossfire results, with more GPU headroom, the BD trails behind by quite a margin, sometimes only giving 50% of a last generation I7.

To all intents and purposes, its a perfectly good CPU, it just isnt as powerfull as an I7, by quite a distance, in almost everything.
 
Last edited:
He went to a lot of trouble to prove the systems were GPU locked! :D

I think he's using the 'real world' excuse where 'no one games at 1024x' ... I can see the point and true in higher resolutions with a single card you might as well have ANY modern CPU because your games will be GPU bound anyway. But as you say above it's not a true analysis of actual processing power.
 
so.... does BD utilise your GPU more effectively?


What the tests are saying is that if you use a certain GPU with certain GPU intestive games.... BD gives a better result that I7?

Help me understand!
 
He went to a lot of trouble to prove the systems were GPU locked! :D

I think he's using the 'real world' excuse where 'no one games at 1024x' ... I can see the point and true in higher resolutions with a single card you might as well have ANY modern CPU because your games will be GPU bound anyway. But as you say above it's not a true analysis of actual processing power.

Yet "low res" gaming isn't a true analysis of Bulldozer's power, or why can Bulldozer beat not only a 2500k but a 2600k in winrar, 7zip, par 2, x264?

Bulldozer isn't epic, its bigger than it should be and uses a lot of power(though, in some ways its better than Intel, when the gpu isn't being run, in high cpu loads the 2600k can use around 800million transistors(around 160-180mil are in the gpu) that uses 95W, Bulldozer uses 125W, but thats 2 billion transistors, so powering around 2.5times the transistors it only uses 30% more power), but due to consumer cost and the lack of a lower end HT chip, its really very competitive with a 2500k.

Which chip I end up getting will still probably boil down to two things, will any of the lower end Bulldozer's "unlock", ie £90 or £130 ones can unlock to a full octo core, or if Intel put the 2600k down in the £170-190 price bracket which would make the 8120 not fast enough or good enough value to buy instead.

Just a pain we don't know when the 2700k is coming, how much the 2600k will drop in price and afaik no one reviewed anything other than octo core Bulldozers so no one had the chance to test unlocking options(every FX4100/fx6100 score I've seen was emulated by disabling cores on a full chip).
 
so.... does BD utilise your GPU more effectively?


What the tests are saying is that if you use a certain GPU with certain GPU intestive games.... BD gives a better result that I7?

Help me understand!

No, its simple that, as ALWAYS 99.9% of games are gpu limited, so the "extra" framerates a faster chip can provide, simple end up with lower cpu load being used in gaming.

IE 2600k gives say 150fps at stupid low res, Bulldozer gives 110fps, so does Phenom 2's, at a middle res, the bulldozer still gives 110fps, the 2600k gives 120fps, but at high res, you're gpu limited, the gpu can only give you 80fps, the Bulldozer is capable of that frame rate, the 2600k is capable of that, the Phenom is aswell.

Look at the charts, most of the wins and losses are within half a frame in speed, no matter which chip you had you would not feel that difference in frame rate.

Anyone using a bit of sensible thinking has been recommending on these forums for, 2 + years ,that there hasn't been a real CPU upgrade for gaming for years if you are above a certain threshold in performance. If you have, basically any mid 3Ghz or above quad core, from a Q6600 up or Phenom 2 up, then gaming is covered and a new gpu will give you more performance.

Gaming ISN'T a massive cpu load, its not cpu dependant, it doesn't require the best cpu ever, never has and likely never will.

I've told people this for years, and most well informed Intel or AMD or Nvidia "fans" have been saying the same thing for years. Plenty of reasons to upgrade a CPU, gaming isn't one of them.
 
Reading Drunkenmaster's post there I noticed something he said about no-one has tested the FX4100 cpu, just emulated that by disabling cores on a full full chip. Does anyone know of a review done by anyone of these "true" quad chips? They might have a reasonable price/ performance ratio.
 
Last edited:
He went to a lot of trouble to prove the systems were GPU locked! :D

I think he's using the 'real world' excuse where 'no one games at 1024x' ... I can see the point and true in higher resolutions with a single card you might as well have ANY modern CPU because your games will be GPU bound anyway. But as you say above it's not a true analysis of actual processing power.

Yes to say that "Bulldozer rocks for gaming" is highly misleading, a more fair conclusion would be that at typical high resolution settings the GPU will become bottleneck sooner and Bulldozer's inherent poor performance is not exposed.

It was the same story back in the day with Phenom (the original) v i7 920, everybody knew the 920 was a faster processor but games rarely exposed it and it was decent enough outside of gaming and cheap enough to make it a worthwhile purchase.
 
Reading Drunkenmaster's post there I noticed something he said about no-one has tested the FX4100 cpu, just emulated that by disabling cores on a full full chip. Does anyone know of a review done by anyone of these "true" quad chips? They might have a reasonable performance/ performance ratio.



its the only chip out of the bulldozer lot I'd be interested in buying. so i wish someone would bench it properly. £95 with plenty of overclock headroom!? Gimme some benchies please!!! :D The other processors are way too over priced.

My november purchase is either going to be a phenom 2 1090t, a 2500k or a 4core BD(depending on OC'd benchmarks)
 
The low end cous will not unlock extra cores. AMD stated they were intentionally blocking it. Whether someone can get around this is another question (hi2u pencilmod?)
 
dm, then thecpu does matter for gamers, at least ones that that have high end cf/qcf or sli/qsli set ups, as most likely bulldozer would hold the gpus back, when you have that much power.
 
Reading Drunkenmaster's post there I noticed something he said about no-one has tested the FX4100 cpu, just emulated that by disabling cores on a full full chip. Does anyone know of a review done by anyone of these "true" quad chips? They might have a reasonable price/ performance ratio.

Id like to see this chip benchmarked aswell. Would be nice to see what sorta overclock it could reach on air.
 
Back
Top Bottom