Well actually the correct way should be to calculate the points per MHz te get proper picture on IPC So for Ryzen 1888/3700=0.51027 for @Anarion 3770K 1900/4400 = 0.43181 for @Laci 2500K 1739/4600 = 0.37804 for @justdoge 6600K 2286/4500= 0.508 for @Only Intruder 4690K 1770/3900=0.45384 for @Warrax 2600K 1706/4500 = 0.37911 for @EL1TE 4670K 1909/4200 = 0.45452 which means that if all is scalling in linear fashion with clock on this bench the results will look like this if all the CPUs were runing at 4GHz Ryzen 4000*0.51027 = 2041 6600K 4000*0.508 = 2032 4670K 4000*0.45452 = 1818 4690K 4000*0.45384 = 1815 3770K 4000*0.43181 = 1727 2600K 4000*0.37911 = 1516 2500K 4000*0.37804 = 1512 Can someone of you guys clock to 4.0GHz to test this math ?

Have no idea whether this thread concerns literal information, of course, but what surprises me is why AMD upsetting the Intel apple cart should surprise anyone. AMD shutout & shut down Intel several years back with the A64--just blew the doors of poor ol' Intel's plans...;) It was even more surprising when it happened then, because no one else had managed such a feat. Indeed, Only AMD alone among all of the would-be Intel cpu competitor companies lived to survive and prosper. Intel has a nasty habit of eradicating competition through whatever means possible. AMD alone thus far is the only competitor to survive and best Intel in terms of technology and marketable products. In its "response" to A64, aside from paying retailers not to sell AMD cpus and threatening motherboard manufacturers who might dare to make and sell Intel competitor products, Intel cancelled the original Pentium architecture and went back to the drawing board only to come up with Core/2--made possible *only* because of a cross-license deal with AMD whereupon Intel obtained x86-64, the foundation for Core 2 and all else that has come after from Intel (Today's "Pentium" is not the actual Pentium architecture--it's merely another offshoot of Core2.) But what Intel wanted to do was to put the world on Itanium, and run the world's ram from Rdram. Rdram was many times costlier than DDR SDRAM, and going with Itanium--the only way Intel wanted to do 64-bits at all--would have meant dumping *all* of your x86 hardware & software..! Starting over from ground 0! Intel's vision, of course, was horrible--at least from the consumer's point of view. I'm sure the Itanium future looked very bright from Intel's corporate view, however...;) Thank goodness for AMD arriving like the proverbial calvary...! If AMD did it once I see no reason to think that AMD cannot do it a second time. I *hope* this is a case of AMD wanting *maximum impact*--people expecting less but getting a lot more at release time. The original Athlon was a complete surprise--and a delightful one. We shall see in a few days. And then it will be possible for people to see just how badly Intel has been gouging them over the last few years...;) We will also get to see what Intel's been hiding in reserve for a day like this--if anything at all.

This is the kind of benchmark I've been looking for months! CPU-Z benchmark is a good reference in my opinion and it's quick and easy to use. Bench results are very impressive, about the same IPC and higher Multithread performance. This mean one thing: it will be all about MHZ. So if Ryzen overclock well, it will be a winner for everyone. Now I wonder if 5+ghz as the norm will be a thing, because Intel might push it that far if competition is fierce. My results (2600k @ 4.5ghz): ST: 1706 MT: 7095

Looking good, so lets hope the availability will be adequate and there will be no problems with it.

This is the kind of benchmark I've been looking for months! CPU-Z benchmark is a good reference in my opinion and it's quick and easy to use. Bench results are very impressive, about the same IPC and higher Multithread performance. This mean one thing: it will be all about MHZ. So if Ryzen overclock well, it will be a winner for everyone. Now I wonder if 5+ghz as the norm will be a thing, because Intel might push it that far if competition is fierce. My results (2600k @ 4.5ghz): ST: 1706 MT: 7095

CPU-Z is good but it only tests one type of workload. So we still don't really know yet if it's a winner for "everyone" or just gamers. Granted Guru3D is mostly, if not all gamers - but yeah. Regardless its definitely going to push Intel's pricing down and help AMD gain some marketshare back.

4670k @ 4.2Ghz: http://i.imgur.com/DHNPxGU.png I'll pass, will wait for 1700x benchmarks.

Well that looks great! I cant wait to see some actual benchmarks of top 8core models. EL1TE Salute to a fellow Sea Hawk user. 🙂 What clock you run it at if i may ask?

Well that looks great! I cant wait to see some actual benchmarks of top 8core models. EL1TE Salute to a fellow Sea Hawk user. 🙂 What clock you run it at if i may ask?

I don't use overclock atm, didn't feel the need for it, i only play MMOs atm. 🙂

FX-8350 @ 4.8GHz http://i.imgur.com/EoaatsO.png

my result FX8320E with 3.5 GHz base frequency and 4.0 turbo (stock is 3.2/4.0) 1049 single thread 6565 multi thread

I don't use overclock atm, didn't feel the need for it, i only play MMOs atm.

I see! Mine runs with slight overclock at 2100mhz, wonder if we will see a Sea Hawk TI version coming out.

Woah...IPC of Ryzen is good..Might switched Our Family PC with 7700K to Ryzen for an All AMD Build..:D For Comparison.. Here's my Main Rig with 6950x @ 4.5ghz: http://imageshack.com/a/img922/324/r3gM1T.jpg Our Family PC 7700K @ 5.0ghz : http://imageshack.com/a/img921/4096/3aDUtG.png

Keep in mind the speed difference Ryzen at 3.7 7600k at 4.2.

And ram speeds too, depends what they used, probably 2400mhz which is slower then ddr3 2400mhz 😀 I get ST 2160, MT ~9350 @ 4.7ghz, ram 2400mhz.

Well actually the correct way should be to calculate the points per MHz te get proper picture on IPC So for Ryzen 1888/3700=0.51027 for @Anarion 3770K 1900/4400 = 0.43181 for @Laci 2500K 1739/4600 = 0.37804 for @justdoge 6600K 2286/4500= 0.508 for @Only Intruder 4690K 1770/3900=0.45384 for @Warrax 2600K 1706/4500 = 0.37911 for @EL1TE 4670K 1909/4200 = 0.45452 which means that if all is scalling in linear fashion with clock on this bench the results will look like this if all the CPUs were runing at 4GHz Ryzen 4000*0.51027 = 2041 6600K 4000*0.508 = 2032 4670K 4000*0.45452 = 1818 4690K 4000*0.45384 = 1815 3770K 4000*0.43181 = 1727 2600K 4000*0.37911 = 1516 2500K 4000*0.37804 = 1512 Can someone of you guys clock to 4.0GHz to test this math ?

Nice math work, yeah that looks logical. Although ram speeds matter a bit, but still overall seems to be spot on. Overall really impressed by its ST, now if all OC to at least 4.5 - 4.7Ghz, then its a win win for AMD. 🤓

But what Intel wanted to do was to put the world on Itanium, and run the world's ram from Rdram. Rdram was many times costlier than DDR SDRAM, and going with Itanium--the only way Intel wanted to do 64-bits at all--would have meant dumping *all* of your x86 hardware & software..! Starting over from ground 0! Intel's vision, of course, was horrible--at least from the consumer's point of view.

True. Though letting go of the current outdated legacy architecture would be a good thing in the long run. PC CPU's and software are so horribly inefficient because of all the legacy support and abstraction layers of the software running on it. Itanium would have rid us of that partly, but had too many caveats.

@Tj - Your maths look OK to me. WOW! AMD is back! Big time. Anyone checked Intels share price? My 4690K @ 4.2GHz scores 1900 (12 points more but it needs 500MHz more than the Ryzen @ 3.7GHz) Estimated then a Ryzen @4.2GHz scores 2143 :banana:

People are arguing that this is not the 1600x but the 1500 because of the 65 W tdp

People are arguing that this is not the 1600x but the 1500 because of the 65 W tdp

They are seeing things, maybe drunk and high fanboys? lmao

They are seeing things, maybe drunk and high fanboys? lmao

Dunno, there is a huge discussion because of it over @ Andantech forum

Those leaks are being released in such an orderly fashion. :3eyes:

Interesting that the 6 core seems to have the full 16MB L3. I wonder if that makes it have a tad better IPC than the 8 core in cache-sensitive operations.