Friday, September 28th 2012
AMD A10-5800K Capable of 6.50 GHz over LN2: Company
AMD's new A10-5800K "Trinity" APUs, launched earlier this week, are capable of extreme overclocking, something similarly-priced Intel processors can't claim, according to Adam Kozak, desktop products manager with the company. According to Kozak, the roughly $150 A10-5800K are capable of 6.50 GHz overclocked speeds, when augmented with liquid nitrogen cooling.
Overclocking capabilities give AMD's sub-$150 chips such as the A10-5800K, A8-5600K, and A6-5400K an edge over similarly-priced Intel chips. The cheapest overclockable chip from Intel's current lineup is the $220 Core i5-2500K. Based on the "Trinity" silicon, the A10-5800K ships with clock speeds of 3.80 GHz, which go up to 4.20 GHz with TurboCore. The chip features an unlocked base clock multiplier, which makes overclocking possible.
Source:
TechWorld.com.au
Overclocking capabilities give AMD's sub-$150 chips such as the A10-5800K, A8-5600K, and A6-5400K an edge over similarly-priced Intel chips. The cheapest overclockable chip from Intel's current lineup is the $220 Core i5-2500K. Based on the "Trinity" silicon, the A10-5800K ships with clock speeds of 3.80 GHz, which go up to 4.20 GHz with TurboCore. The chip features an unlocked base clock multiplier, which makes overclocking possible.
157 Comments on AMD A10-5800K Capable of 6.50 GHz over LN2: Company
Businesses to me make up the biggest market of the Traditional PC because of the Number they Need at once. Businesses will always go with the lowest price parts for their general terminals and even servers. Only the science community goes for anything faster which makes for a smaller percentage of businesses. Enthusiast Market is very small in the consumer range because about 99% are average joes that just only know how to turn them on, open webbrowsers, office applications and play movies or the casual game, (No knowledge of how they operate- which in the end comes down to the gate- on/off or 1/0)
Ive been thinking of honestly making a APU based machine with the Highest expandable board possible (Which in this Case would be ones with the PEG 2.0/2.1 x16+x0 or x8+x8, x4 configuration boards- Def AsRock or Gigabyte)
overall Both Llano and Trinity based APUs are good for the Range they operate in.
At the end of the day both us X86 which is old and not the most efficient architecture.
But at the end of the day most people use X86 because its cheap and the games.
With out this love hate relationship and passion a lot of people wouldn't care.
A10-5800K FP SP GFlops: 4 Cores * 8 Flops * 3.8 GHz = 121.6 FMA SP GFlops
7660D FP SP GFlops: 6 Cores * 128 Flops * 0.8 GHz = 614.4 FMA SP GFlops
Total Aggregate APU SP FP GFlops: 736 SP GFlops
i5 2500K/3570K FP SP GFlops: 4 Cores * 16 Flops * 3.6 GHz = 230.4 MAC* SP GFlops
HD 3000: 12 EUs * 8 Flops * 1.35 GHz = 129.6 FMA SP GFlops
HD 4000: 16 EUs * 8 Flops * 1.15 GHz = 147.2 FMA SP GFlops
Total Aggregate APU SP FP GFlops:
i5 2500K = 360 SP GFlops
i5 3570K = 377.6 SP GFlops
*MAC requires two instructions one for the Multiply and one for the Add. While FMA requires only one instruction for the Multiply and Add.
Also, VLIW5 and VLIW4 have a huge utilization issue where Graphic Core Next doesn't. Even though 736 SP GFlops is possible with the A10-5800K. Until, Graphic Core Next the best you can get in an application is: 326.4 SP GFlops. I know a couple of people who tried to GPU accelerate their platform with OpenCL on the 6900 series and they were only able to get 1/3rd of the GFlops on the GPU. Till Graphic Core Next comes out and you are less likely to go above 330 GFlops.
I like the idea of apu's because if software actually used all of those 736 SP GFlops.
2500k at 4.5ghz = 288 GFlops for cpu only.
5800k at 6.5 ghz = 208 GFlops for cpu only.
Yields and silicon fabrication cost. Speed/efficiency of process ramp. Die size. Performance/watt (for OEM's to get away with the bare minimum PSU and cooling)
DIY market = largely immaterial, although I'd note that Intel users seem more disposed towards faster upgrade cycles. Many Intel users (including myself) seem to have graduated from LGA 775 Conroe to Yorkfield, to LGA1366 or 1156 to Sandy/Ivy Bridge(-E), and no doubt more than a few will move to 8/10/12 core Ivy Bridge-E when it arrives. All this would tend to indicate a healthy- for Intel-base level of adoption for any given platform.
AMD users on the other hand, while vociferous in lauding their Sunnyvale masters, seem to be more judicious in their expenditure...so, praise for the products...but happy to sit pat with their present system until the UltimateEarthmoverOver9000 arrives....translating into no cash in AMD's pocket....and less R&D available for the future UEO9000
Intel Penium 4 630
AMD Athlon 64 3700+
Intel Core 2 Duo E6600
AMD Phenom II 940
Intel Core i7 3820
Notice a pattern? :toast:
Sweet. Someone either:
1. who enjoys tech for it's own sake as much as any performance it may provide, or
2. a consumer locked into retail therapy-one step away from John Carpenter's "They Live!"
Strangely enough, many tech enthusiast sites have a high proportion of posters who subscribe to option 2. As someone who spends more on their watercooling than many people spend on their entire system, it tends to amuse me when people incessantly chant/plead for more people to buy an IHV's product, while they themselves upgrade on a 3-5 years basis.
Since you're drawing attention to your alternating IHV preference, I assume your next purchase is AMD....How long do you think you'll have to wait before an AMD platform betters the X79 system you already have (including the likely obvious 8 or 10 core Ivy Bridge-E/ 12-core Xeon drop-in replacement)? How long before AMD see's any of your cash?
An interesting exercise would be to chart the demographic of posters IHV affiliations with regard their present hardware fit-out and relative age of componentry...somehow I don't think my hypothesis would wilt under the scrutiny.
some of us have to aportion our finances differently is the truth in it, i got this 960T because Bd didnt perform and my Q6600 blew up, i didnt have double the ammount at the time for an intel platform and i wanted an upgrade path, so went in the direction i did, had i infinite money id also upgrade each platform release tho and it would be both vendors not one, in an ideal world id try them all for a bit:D
BTW, the answer is no.
Oh well.
I see this as a PR stunt aimed at .01% of the people.
Just as a question, How realistic is it to run LN2 and a 6.5GHz OC 24/7? How many will even be able to do this? NONE! So other than a PR stunt this means WHAT? I admit the price of these chips are great! I mean for 150 bucks and 500 bucks of LN2 parts you can hit 6.5GHz! Sweet!
I am not saying it is a bad thing but come on let us bring this into the world of REALITY! NO ONE can run LN2 24/7 not to mention a CPU at 6.5GHz 24/7! Would it be cool to have it? HELL YEAH! Would I find this a viable solution for some one? NO. So with that said this is fluff.
as has been said before ill look for a similar statement in the next intel world record Ghz statement from you, T:roll:
www.xtremesystems.org/forums/showthread.php?283218-Killer-Trinity-top-A10-5800k-confirmed-5.1GHz-on-air-and-7.3GHz-on-LN2
Binned chippies = who cares?
2D Benches are all at 4 GHz...:p
And again, those 2D benches are @ 4 GHz. :p
Don't get me wrong.. I just wanna squish any unrealistic expectations.