You mean the only thing that CPUs are used for is games?
What are you talking about?
The graphs had a 3770k and 4770k both at stock speeds.
And stock 3930k.
That graph did, but look at pages 18 onward. Only compares the 9590 with an overclocked and more expensive Intel CPU. Basically huge waste of time and gives no useful information.
Admittedly, there are comparisons with the 4770 on super-relevant benchmarks such as USB speed, memory, and SSD performance... seriously?
It's a worthless review.
Unless "bash on AMD products" is in your top-10 list of "fun things I want to do with my day today" then why on earth would you spend time bashing on an AMD CPU at the expense of using that time to do any one of those other ten things you have on your top-10 fun things to do list?
317 watts; and that price; for that effort!!?
IDC:thumbsup: Well said!The results are pretty much as expected based on OC'ing reviews of the 8350 combined with the added expectation that AMD would be binning these as superior chips in terms of power-consumption and shmoo plot results.
I really don't understand the negativity surrounding AMD's decision to launch this product.
Its like being an orange lover but deciding it isn't good enough to just find great oranges to eat and leave everyone else to themselves, instead you also insist on loathing pear eaters and seeking out pear eaters to point out how much their pears taste nothing like your oranges. A pear is never going to taste like an orange, accept it and move on.
Its like being an orange lover but deciding it isn't good enough to just find great oranges to eat and leave everyone else to themselves, instead you also insist on loathing pear eaters and seeking out pear eaters to point out how much their pears taste nothing like your oranges. A pear is never going to taste like an orange, accept it and move on.
Yes, but that's easily 260 watts for the cpu alone, running cinebench and getting beaten by a stock 4770 using 1/3 the power. Hit it with linx and enjoy all types of sqealing from your motherboard and psu. What's funny is the clueless saying these chips will run at around 160 watts! Well, these don'k look like binned silicon, if one goes by the results in that review.Thats total system power draw. Still looks bad compared to everything else but not horrible in the context of the performance.
The FX9590 @ 5GHz is generally slower than a stock 4770K, costs nearly as much as a 3970X, and needs serious cooling.
This is sad.
Link: http://www.kitguru.net/components/cpu/zardon/amd-fx9590-5ghz-review-w-gigabyte-990fxa-ud5/
are also sad, because on amazon you can buy the 9590 by less than the price of two 4770kYou could buy two 4770k processors along with some memory and a good power supply for the price of a single FX9590
Again sad because in amazon the 3930k is slightly more expensive than the FX9590This is significantly more expensive than Intel’s 3930K (£450)
They did not overclocked the 9590 but run it at stock.how the highly overclocked AMD FX9590 system would perform
This guy seems unaware that worldrecords of the FX series have been achieved with eight cores enabled.Many people didn't realise at the time that 6 of the 8 cores were disabled to achieve the overclock. Its a meaningless chase for bigger numbers, translating to absolutely nothing in the real world. You show me an enthusiast user who buys an eight core processor to run with six disabled and ill show you a psychiatric hospital.
Some technical issues:
- Used Windows 7 and its bad scheduler for FX.
- Compared a FX at stock speed with the 3960x overclocked at 4.4GHz.
- Different brand and memory speed. The i7s used 13% faster memory.
- Use of benchmarks specifically optimized for intel such as Mediaespresso: Intel Quick Sync Video support.
That graph did, but look at pages 18 onward. Only compares the 9590 with an overclocked and more expensive Intel CPU. Basically huge waste of time and gives no useful information.
Admittedly, there are comparisons with the 4770 on super-relevant benchmarks such as USB speed, memory, and SSD performance... seriously?
It's a worthless review.
Oh snap, they used the most popular windows based OS? You sure they didn't hotfix it?
They overclocked FX to 5GHz, seems reasonable to overclock the 3960x to 4.4GHz
Does it really matter? AMD's IMC is nothing short of awful, same ram would still see a large throughput advantage for Intel even with the dual channel boards, quad isn't even going to be close so I'm not sure about the validity of the grievance there.
The fact that the more expensive AMD cpu doesn't come with a gpu to aid in things that support quick sync and opencl isn't a technical issue, it's a limitation with Piledriver.
You did pick the cheapest 9590 but not the cheapest i7-3930k, so it's not exactly fair to make claims that it is cheaper than the 3930k... but you bring up a few good points with Windows 7, memory inconsistencies, and Mediaespresso.
I'm curious to see if it still has more headroom to OC!
If now the rule is that one only can use the most popular OS why anandtech uses windows 8 in recent tests? In any case my complaint was not about popularity, but purely technical.
Well they did everything possible and more to favour Intel chips (OS, memory, selected tests...) therefore I agree with you one that seems reasonable that they overclocked the 3960x by a 33% and the FX by a 6%.
I am not complaining because one of tests utilizes an enhacement available on Intel, but not in piledriver. Don't misinterpret me, if one chip has a technological advantage over other that is great. I am complaining because everything was made to favour Intel.
Again sad because in amazon the 3930k is slightly more expensive than the FX9590
USB/SSD performance is really important and often overlooked, why is it bad that it's there?
The 9590 is basically overclocked by default
The results where they do show stock and OC results for all of the chips don't paint a great picture for the 9590 anyway.
Considering AMD doesn't even care enough to send out legit review samples to the review sites, this is a good review for a group that had to waste their own money on this pile of overpriced crap.
The 6300/8320/8350 are fairly good value when matched with the right workloads. The 9590 is horrendously terrible value..
How do you figure? It effectively matches the $1050 Intel i7-3960 in most of the tests. If that is a bad value, what kind of value is the Intel CPU, given it costs nearly twice as much and doesn't offer any performance advantage?
Comparing OC percentage is erroneous.
I'm not understanding your issue with QuickSync, it's not like the program has Cripple_AMD in it, it's just inferior hardware shown in a real world application.
Unless you limit a review to a very select few applications AMD is always going to be shown in a poor light because the hardware isn't as good. It's not bias, it's actuality.
We know that tomshardware review, made with many applications for which the fix makes nothing by design. Anand found 5% improvement in several games. Up to a 10% can be obtained in some game. In case this is not evident, a 5% improvement is about what one see in games between the i5-IB and the i5-HW.
I already explained my point about QuickSync and it is unrelated to Cripple_AMD (which is not even mentioned in the post that you reply). Once again: don't misinterpret me.
Are you serious? A 5GHz 8C PD barely matches/beats a stock clock i7 4770K. Both i7 3930K and i7 3960X would beat the FX9590 pretty much everywhere (at stock) and likely offer a lot more OCing headroom on air/water.