CPUBoss Review Our evaluation of 4770K vs 9590 among desktop CPUs (over 75W)


Benchmark performance using all cores

PCMark 8 Home 3.0 Accelerated, PassMark and 1 more

Single-core Performance

Individual core benchmark performance

PassMark (Single Core), Geekbench 3 Single Core and 1 more

Integrated Graphics

Integrated GPU performance for graphics

Fire Strike

Integrated Graphics (OpenCL)

Integrated GPU performance for parallel computing

CompuBench 1.5 Bitcoin mining and 4 more

Performance per Watt

How efficiently does the processor use electricity?

Fire Strike, CompuBench 1.5 Bitcoin mining and 11 more


Are you paying a premium for performance?

Fire Strike, CompuBench 1.5 Bitcoin mining and 11 more


CPUBoss Score

Combination of all six facets

Intel Core i7 4770K 

CPUBoss recommends the Intel Core i7 4770K  based on its power consumption.

See full details

Cast your vote Do you agree or disagree with CPUBoss?

Thanks for adding your opinion. Follow us on Facebook to stay up to date with the latest news!

Differences What are the advantages of each

Front view of Intel Core i7 4770K

Reasons to consider the
Intel Core i7 4770K

Report a correction
Much lower typical power consumption 68.25W vs 178.75W 2.6x lower typical power consumption
Much better cinebench r10 32Bit 1-core score 7,718 vs 4,905 More than 55% better cinebench r10 32Bit 1-core score
Has a built-in GPU Yes vs No Somewhat common; A separate graphics adapter is not required
Much newer manufacturing process 22 nm vs 32 nm A newer manufacturing process allows for a more powerful, yet cooler running processor
Much lower annual home energy cost 20.24 $/year vs 53 $/year 2.6x lower annual home energy cost
Much lower annual commercial energy cost 73.58 $/year vs 192.72 $/year 2.6x lower annual commercial energy cost
Much better cinebench r10 32Bit score 30,095 vs 26,635 Around 15% better cinebench r10 32Bit score
Much better performance per watt 12.04 pt/W vs 5.51 pt/W Around 2.2x better performance per watt
Front view of AMD FX 9590

Reasons to consider the
AMD FX 9590

Report a correction
Much more l2 cache 8 MB vs 1 MB 8x more l2 cache; more data can be stored in the l2 cache for quick access later
Much higher turbo clock speed 5 GHz vs 3.9 GHz Around 30% higher turbo clock speed
Much higher clock speed 4.7 GHz vs 3.5 GHz Around 35% higher clock speed
More cores 8 vs 4 Twice as many cores; run more applications at once
Much better performance per dollar 5.85 pt/$ vs 4.08 pt/$ Around 45% better performance per dollar
Much more l2 cache per core 1 MB/core vs 0.25 MB/core 4x more l2 cache per core
Much better PassMark score 10,589 vs 10,016 More than 5% better PassMark score
Significantly better overclocked clock speed (Air) 5.07 GHz vs 4.47 GHz Around 15% better overclocked clock speed (Air)
Significantly better overclocked clock speed (Water) 5.06 GHz vs 4.53 GHz More than 10% better overclocked clock speed (Water)

Benchmarks Real world tests of Core i7 4770K vs FX 9590

GeekBench 3 (Multi-core) Data courtesy Primate Labs

Core i7 4770K
FX 9590

GeekBench 3 (Single core) Data courtesy Primate Labs

FX 9590

GeekBench 3 (AES single core) Data courtesy Primate Labs

Core i7 4770K
4,810,000 MB/s
FX 9590
2,790,000 MB/s

GeekBench (32-bit) Data courtesy Primate Labs

Core i7 4770K
FX 9590

GeekBench (64-bit) Data courtesy Primate Labs

Core i7 4770K
FX 9590


Core i7 4770K
FX 9590

PassMark Data courtesy Passmark

Core i7 4770K
FX 9590

PassMark (Single Core)

FX 9590

Specifications Full list of technical specs


Core i7 4770K  vs
FX 9590 
Clock speed 3.5 GHz 4.7 GHz
Turbo clock speed 3.9 GHz 5 GHz
Cores Quad core Octa core
Is unlocked Yes Yes


Has a NX bit Yes Yes
Has virtualization support Yes Yes
Instruction set extensions
AVX 1.1
Supplemental SSE3
AVX 2.0
Supports dynamic frequency scaling Yes Yes

power consumption

TDP 84W 220W
Annual home energy cost 20.24 $/year 53 $/year
Annual commercial energy cost 73.58 $/year 192.72 $/year
Performance per watt 12.04 pt/W 5.51 pt/W
Typical power consumption 68.25W 178.75W


Core i7 4770K  vs
FX 9590 
Architecture x86-64 x86-64
Threads 8 8
L2 cache 1 MB 8 MB
L2 cache per core 0.25 MB/core 1 MB/core
L3 cache 8 MB 8 MB
L3 cache per core 2 MB/core 1 MB/core
Manufacture process 22 nm 32 nm
Max CPUs 1 1
Operating temperature Unknown - 72.72°C Unknown - 57°C


Overclocked clock speed 4.47 GHz 5.07 GHz
Overclocked clock speed (Water) 4.53 GHz 5.06 GHz
PassMark (Overclocked) 6,694.4 10,860
Overclocked clock speed (Air) 4.47 GHz 5.07 GHz

integrated graphics

Label Intel® HD Graphics 4600 N/A
Number of displays supported 3 N/A
GPU clock speed 350 MHz N/A
Turbo clock speed 1,250 MHz N/A

memory controller

Memory controller Built-in Built-in
Memory type
Channels Dual Channel Dual Channel
Supports ECC No Yes
Maximum bandwidth 25,600 MB/s 29,866.66 MB/s
Intel Core i7 4770K
Report a correction
AMD FX 9590
Report a correction

Read more


Showing 25 comments.
Really I didn't know CPUBOSS was owned by Intel? My FX9370's Cinebench score is on par with a 4770k, I've noticed their bias towards AMD shits me off actually.. I'm rebuilding my AMD build soon into a new case and a better board for a new color scheme (MSI 990FX) GPUBOSS is owned by who then?
I find it to be just as bad.
BTW you guys should use this site instead http://www.game-debate.com/cpu/index.php?pid=1875&pid2=1187&compare=fx-9590-vs-core-i7-4770k-4-core-3-5ghz it's waaaay ,more accurate than this shit site
Yea dude, I have also confirmed CPUBOSS is INTEL property, the 9590 is not this bad, it's on par with the 4770k unless you OC it, not like the 4770k can OC good anyway. This fkng site keeps saying that AMD is crap but when buying the CPU's and testing them in real life, you will notice CPUBOSS is just a load of BS
I AM PLANNING TO BUY A NEW ONE. AFTER FINDING THE PROPER BUDGET PC AND SEEING ALL THE DISCUSSION I M STILL NOT SURE WHICH ONE TO BUY. asus readon r9 380 4gb ddr5 graphics card Msi 970 gaming 32 gb ddr3 amd motherboard amd fx-8350 8 core 4-4.2 ghz processor kingston (hyperx savage) red 8 gb ddr3 ram WD caviar green 1tb hdd deepcool tesseract wh sw mid tower cabinet corsair vs650 smps cooler master hyper 212x fan with heatsink NOW IS THIS CONFIGURATION GOOD ENOUGH OR SHOULD I GO FOR INTEL??? JUST HELP ME OUT GUYS THANKS
amd boost can get 1.6 peak performance for short periods of time so is get just as much as it need intel turbo get like 1.2 for more time but it can cause some problems, it mean lag in games, autocad slower . amd is faster for big problems, but it heat much more intel is slower but is a little bit coller than amd
not really top of power amd kill intel, graphics amd kill intel, raw power amd kill intel duration intel kill amd . but amd is even powerfull, faster . . if u understand RISC and CISC u should know what i mean
Thank you, I can't stand fanboys!
If only more people were like you.
I have a 9370 in my second rig and my cinebench score is on par with a 4770k @ 5.0ghz and costs nearly $150 less! Same goes for the 4690k, Its way below my score. I'm not a fanboy! I have an Intel x99 rig with Nvidia gpu's and an AMD rig with Radeon gpu's and have extensively tested both! People always talk shit about AMD cpu's but from my experience the 9370 is actually a really good and powerful cpu! It will be more than adequate to power through rendering and gaming and is much cheaper! Long live Intel and AMD!
I'm pretty sure it'll be down to Intel being too smug to do anything like this. I mean, Intel already turned down Apple when it wanted custom processors for its iPhone, back in 2007.
Why did the comparison only look at one type of motherboard for the Intel i7 4470? As the attachment illustrates the motherboard accessories interfaces, fans, pump, lights, LED's, drives, etc, have a considerable impact on the TDP rating. http://us.hardware.info/productinfo/benchmarks/1/motherboards?products[]=264808&specId=11787&tcId=329. The TDP is also dependent on the software load; the more complex software the more the CPU has to work and hence the motherboard has to perform; much higher TDP will result. Also note the the TDP is specified for the lower clocking frequency 3.5 GHz for the Intel i7 4470 CPU whereas the AMD FX9590 is documented with the much higher clocking frequency 4.7 GHz. Why author did not make comparison at the same overclocking frequency or several clocking frequencies to demonstrate the clear issue? From experience as the clocking frequency is higher the TDP increases significantly. The i7 4470 will have TDP of approximately 178 to 205 watts depending on the software load, motherboard type, and overclocking clocking frequency. The comparison is very biased by the author. Thank you.
The benchmark comparisons are noted with Intel's 4770K/4790K chip sets versus the AMD FX9590 CPU; overclocking capability. As such the TDP at the approximately same frequency, the cooling method, and skill sets of the user need accessing. We should also look at the overclocking TDP which is representative of the chip CPU set's characteristics. The FX9590 TDP is presented as 220 W @ 4.7 GHz where as the i7 4470K/4790K CPU is 84 W @ 3.4 GHz.. When the i7's are over clocked the watts rise dramatically. For rough comparisons, 3.8 GHz, the i7 CPU's draw about 172 watts with slightly higher overclocking the power TDP comparisons between the i7's and the FX9590 at about 4.6 - 4.7 GHz are not that much different. The AMD CPU are "factory set overclocked" base 4.7 GHz with upper level, user interface, of about 5.22 GHz.. Where as the i7's are "factory set" at 3.4 GHz with upper frequency, user interface, of about 4.2 - 4.6 GHz. For approximately the same clock speed there is that much difference between CPU's TDP (heat dissipation). The second point, as both the Intel and AMD CPU's will dissipate a lot of heat air cooling is not a option. Both the i7's and the AMD CPU's will require radiator/fan cooling. The AMD retail package the cooler is included. The cooler for the i7's have to purchased separately. The last point to note here is the AMD CPU is based at 4.7 GHz and turbo at 5.0 GHz, "out of the box", requiring basic to moderate skills in computer settings where as the i7's will require a more experienced person who understands the overclocking complexities; voltage setting, frequency, and temperature collation. Even with such skill sets not to say that the upper overclocking capabilities of the i7's can be achieved by every user (i7's are sensitive with voltage stability issues when overclocking). The heat dissipation of the TDP compared by Benchmark would be more relevant with 4470 and 4790 (non-clocking) CPU's comparison; people who purchase the 4470K/4790K will intend on using the over-clocking capability (otherwise why would you spend the money) and equally so, Benchmark should also noted the TDP comparison at the overclocking options (comparing apples with apples). For routine threading calculations the 4770K\4790K do very well with single task threading and the Hanswell architecture; the strength for the Intel's i7's CPU. The FX9590 strengths are likely with high task multi-threading software applications, Photoshop for example, multi-register arithmetic applications, and multi-tasking requiring multi-core applications. The single task threading are not the strength for the AMD chip set. For routine desktop work or mid-gaming the i7's are a better fit. Yes the Intel i7's perform better than the AMD FX9590 on "computer time and crunching in nanoseconds or few seconds" or pixels/sec more or a couple of frames/sec more, but for the average person the time or quantitative difference is not relevant; either vendors CPU's can perform the tasks quickly and efficiently. Cost comparisons for a desktop (assembled and tested); case, CPU, power supply, motherboard, 240 GB SSD & 2 TB HDD, 16 GB 1866 MHz ram, mid-range video card, and radiator/fan cooling about 1700 USD for i7 system, and about 1300 USD for the AMD system (+/-). Last, will be interesting to read how the FX9590 and i7's performs with the latest windows 10 operating system, and next generation application software (which hopefully will be available at windows 10 release), which is led to believe developed to take into consideration the new generation CPU's with multi-threading and multi-core applications.
they are using amd APUS as you said just because they are cheaper and they not need a dedicated gpu they are not usng amd because it is better (it isnt)
What do all the new consoles use? Ah yes AMD, most games are developed now for that system.
AMD over Intel any day of the week, someone has to support the under dogs, you Intel loving freaks should be happy that costs are kept lower because of competition, I'm a proud owner of an FX-9590, beautiful cpu.
If you've never seen a stock cooler above 45ºC then you've never seen a Prescott running a stock cooler. And if you've never seen a Prescott running a stock cooler, then you haven't been in IT for 17 years. Unless you're counting tier 1 script reading in an ISP call center as "IT". Just sayin'.
TDP 220W is a wonder, AMD! ¬¬
AMD----if you dont want to spend much money and want medium power INTEL---If you can spend lots of money to get hands on Power with quality SO----just decide for yourself
AMD is better because its price is more affordable. If you do not think that throw money in garbage with the intel.
People don't seem to understand the difference between what the TDP value actually stands for, for Intel and AMD. It's not the same thing. And I guarantee you, clock an i7 up to 5.00 GHz and you too will be pumping out over 200 watts.
less performance, 220w tdp, water cooling necessary, 1000 w good quality psu recommended by amd, espensive motehrboard, in the end fx 9590 is too espensive.
Calculating all of the benchmarks and dividing by 6, you get 6,113 for the i7, and 5,746 for the FX. However, CPU Boss is extremely bad for anything of real use. I mean, there's one comparison where a Pentium 4 is outshining a Core2 Duo, or something like that (suffice to say, it shouldn't be outshining anything!).
Does no one else notice that if you actually calculate the averages that the 4770k gets an 8.4 and the FX 9590 gets an 8.6(rounded, it's actually, 8.55). So unless CPU Boss is using some really weird ass weighting system, they intentionally changed the 4770k's score to be higher.
comments powered by Disqus