Comments Locked

170 Comments

Back to Article

  • bigboxes - Tuesday, December 13, 2016 - link

    Still lacking a lot of details. It would be nice if this was competitive with the i7-6900K, albeit at a lower price point. Wonder why AMD didn't stick with Zen instead of Ryzen. I think Zen had a better name for marketing, if AMD would ever do such a thing.
  • bananaforscale - Tuesday, December 13, 2016 - link

    Based on the rendering demos it's as fast if not faster, at a lower TDP.
  • Meteor2 - Tuesday, December 13, 2016 - link

    95W, to be precise, for 8 cores 16 threads and 'at least' 3.4 GHz.
  • Meteor2 - Tuesday, December 13, 2016 - link

    ...while the i7-6900K is 140W.
  • Sttm - Tuesday, December 13, 2016 - link

    As a 6900K owner I am not sure that I care about the wattage. This is a part for my home PC, not servers. Performance, and the cost of that performance, is where Zen needs to shine to compete with the 6900k, or lets be realistic the 7900k. Which so far looks good.

    I hope they manage to beat the 6900k and do so at a price that is about say $600. Put their high end Ryzen up against Intel's mainstream -E offering.
  • Railander - Tuesday, December 13, 2016 - link

    While TDP does relate to power consumption, it is not a unit of power consumption but rather of power dissipation as heat.
    Having a lower TDP directly affects what sort of cooling solution you need, which obviously translates to smaller/quieter coolers and how much you can push the chip while overclocking.
  • Tewt - Tuesday, December 13, 2016 - link

    I'm sure they mentioned it because anytime AMD gets close in performance, some fan of the competition always throws in a "BUT they have higher TDP so it is not as good a product because I have to spend more on cooling, blah blah blah."
  • bill4 - Thursday, December 15, 2016 - link

    oh lord...in rx 480 vs 1060 ti debates i had on reddit, you wouldn't BELIEVE the amount of people who would act like the 1060's slightly lower power draw (efficiency!) was the end all be all! Just funny to see the arguments reversed here.

    People would hold their nose and exaggerate "oh, the 1060 Ti is 500% more efficient than the 480, what is wrong with AMD!" I'd be like no, it's a few watts, who cares? And what matters (within reason, you dont want a space heater like the FX 9590 chips) is price and performance!

    I mean literally people would act like "efficiency" was FAR more important than raw power or price. Just because Nvida happened to have a current, small, advantage there!

    Nvidia/Intel fanboys are funny...
  • inmytaxi - Friday, January 6, 2017 - link

    As are the people who are engaging them for some strange reason.
  • bigboxes - Tuesday, December 13, 2016 - link

    I run my file server 24/7. I run my HTPC (i7 920) 24/7. I run my main rig (i7 4790K oc)... drumrolll... 24/7. I am not concerned about the cost of running my computers, but less is always better. Less heat, less noise, higher threshold for oc. If the TDP is 95W, while the chip it is being compared to is 140W, that is a huge Δ.
  • Samus - Tuesday, December 13, 2016 - link

    That's a good point on TDP. Even your i7-920 being 130w TDP generally idles around 20w and wont use much more doing basic media playback. The platform (x58) however, is a power hog no matter what the situation, a key factor people generally ignore. AMD platforms based on FM sockets have been power whores since their introduction (AM moved most of the platform functions onto the CPU where things are more efficient) and regarding AM, up until AM3, there were a lot of auxiliary chips (non-native USB 3.0, for instance) that used power.

    I'm excited for Zen. Mostly because it's been a decade since AMD had anything really competitive in performance per watt. I know it doesn't matter to some people but it matters to me, because as you said, less heat, less noise, higher threshold for oc.
  • Nagorak - Wednesday, December 14, 2016 - link

    Yeah, there's this strange concept called shutting down when not in use. It'd probably safe you big on your power bill, increase the longevity of your components and pollute the environment less too. A modern OS with an SSD can boot up in less than 30 seconds soon there's not much downside.
  • bigboxes - Wednesday, December 14, 2016 - link

    Shutting down your computer and then starting it back up does not increase the longevity of your computer. Now, if you only use your pc to game once in a while then sure turn the thing off. I work from home and my main rig is doing something all the time. It has an SSD, but also 4xHDDs that I don't want to constantly spin up and spin down. Haswell is fairly energy efficient and there are all lower power states that it goes into during the rare times I'm not crunching something.

    The file server... well, that needs to be up all the time. I have multiple devices that hit that all the time. 3 TVs, 3 desktops, 3 tablets, 3 smartphones and remotely from when I'm on the go. Not to mention anyone who hits my FTP server. It's got an SSD for the OS drive and 8 spinners to store all those gee bees.

    My HTPC gets put into hibernation when not in use. It pretty much sips on the juice and then wakes up in less than 30 seconds as it loads everything to a state I last left it in. But not turned off. SSD only.

    My wife's desktop gets turned off whenever she's not using it or we go to bed (in bedroom). SSD only.

    For all the holier-than-thou posters, my electrical bill was $83 last month. If you'd actually read my post instead of had a knee-jerk reaction you'd see where I said that energy savings was important. I just don't come into your home and tell you what to do, not knowing your usage.
  • Laststop311 - Friday, December 23, 2016 - link

    It actually lowers the longevity to turn it on and off everyday. All the silicon chips do better at a fixed temp instead of being cold then hot over and over. HDD's have to park the reading heads and unpark. The PC i'm using now hasn't been off for more than a couple hours in at least 4 years and it's 7 years old i7-980x + fiji pro and not a single hardware failure at all. I did a single GPU upgrade from a radeon 5870 to a fiji pro and on the fiji pro i was able to unlock some of the stream processors. It's within a few fps of a full fiji x. If you use your PC everyday just leave it on.
  • Demiurge - Monday, December 19, 2016 - link

    I agree with you in theory, but in practice, considering that you are essentially talking about the power of 2 versus 3 60W light bulbs -- you probably are already wasting the energy elsewhere. Besides, ever heard of sleep modes if you are that concerned about energy... sheesh!
  • TheJian - Tuesday, December 13, 2016 - link

    If this chip can really match a 6900k at 45w less they should charge $1000 for it period. They are not in business to be your friend, and it's high time they make some CASH while they can. Intel won't take forever to respond. Intel shareholders won't want to see a price war today, they want profits high after losing 4.1B a year on mobile for ages. They won't want those gains that come from quitting that crap to go away again due to a price war. Intel can't afford to take a huge stock price hit at this time either. They are about to lose the fab wars and need all the cash they can get. That is why this is the perfect time to attack for AMD.

    Pricing your chips stupid will get them killed. There are people already paying for Intel chips at current prices and they're actually selling a LOT of them. No point in low-balling if you have an equal performing chip at 2/3 watts. I'd probably sell it at price parity at all levels of Intel chips if this is the case. Just make sure you don't screw the low chip with less PCIE lanes and you're golden. People can still get in cheap then and upgrade later to top end stuff or rev2 if AMD makes AM4 last through 10nm starts. Win with stuff like that, not pricing. Price what the market will take and no less. Sure I want a cheap chip, but I also want AMD to make a few billion, get debt free and have a few billion left in the bank cash for the future.

    If they price like you're saying they're idiots and management should be fired immediately. If you sell out the first batch, raise the price another $100. Clearly people would pay an extra $100 when Intel charges double that for the 40 lanes and that's on top of the board costs. So AMD would still be doing small favors then. If you're crunching all day on this (like I will be in handbrake etc) then that 50w bulb burning all day will add up over 5yrs of this things life also. The TCO is what I care about.

    If Intel's next chip comes Q2, you have 3 months to make as much money as possible then cut prices if needed. Of course, if all Intel manages to do is drop watts to AMD's watts with their next rev, I still wouldn't drop prices until I had units stuck on shelves...LOL. AMD has lost billions (and everything they owned) in the last ~15-17yrs. Time to make some bank. It looks like Intel quads will be 95w, so no low pricing please AMD. Make money.
  • MobiusPizza - Tuesday, December 13, 2016 - link

    Yes and no. AMD is not a charity, they need to make money. But, they also need to gain market share, and to do so they need to price low and sacrifice profitability.
  • Manch - Wednesday, December 14, 2016 - link

    Exactly. AMD needs to turn a profit thru volume, not per chip. AMD has earned themselves a stigma that they can only partially erase by bringing out a performing chip.They need to win mind share by offer the same for less. This will bring customers back into the fold. Then if they remain competitive, they can increase prices. When this chip does launch, if its good, hits parity with the competing Intel procs at a lower price, off feature parity, then they will win back customers. If its compete with a big ole * next to it....down teh drain they go.
  • NesuD - Wednesday, December 14, 2016 - link

    There will be a price war. Intel has not lowered their prices through 2 new process nodes in any significant way. They have lots of margin built into their pricing because of higher yields so they can drop prices to compete with amd pricing at almost any level and stay profitable.
  • Manch - Thursday, December 15, 2016 - link

    PCWorld has a good point regarding price.
    http://www.pcworld.com/article/3149101/components-...
  • svan1971 - Tuesday, December 13, 2016 - link

    At the same price AMD will never sell as well as intel, even when they had a better chip 10 years ago intel under cut them until core 2 arrived.
  • artk2219 - Tuesday, December 13, 2016 - link

    yes, because they cheated. they made it so that if oem's used amd chips instead of Intel they would have to pay a higher price for the Intel chips. Dell specifically was paid millions by Intel so that they wouldn't produce any amd systems. Intel was appropriately sued for these action. They basically received a 1.4 billion dollar slap on the wrist which I believe they still haven't paid. That's nothing compared to the billions in opportunity and market loss that amd faced.
  • fanofanand - Wednesday, December 14, 2016 - link

    I believe Intel either just made their final payment or is about to make their final payment, but the $1.4 billion for non-competitive practices has been paid. That said, the damage done was FAR worse than 1.4 Billion
  • Tewt - Thursday, December 15, 2016 - link

    I'm glad some people realize this. Looking at Intel's profit over the past years, they gladly paid off AMD with that pittance.
  • just4U - Wednesday, December 14, 2016 - link

    The sad thing svan1971 is Intel didn't undercut them. Intel was selling their chips at a higher cost but had sweet partner deals keeping Amd out. Not that it mattered I suppose since Amd couldn't keep up with demand anyway.. STILL..

    I remember that time quite well and I still remember the turds out there recommending the P4s over the Athlon 64s and X2s. Was bs. Amd did raise prices though.. Briefly and kept their top end chip at close to cost parity with Intels.. It didn't last though. Had it there for maybe 4 months "IF" that.
  • fanofanand - Wednesday, December 14, 2016 - link

    You obviously get it, it's frustrating to see so few do so.
  • deeps6x - Friday, December 16, 2016 - link

    Keep in mind that AMD is going to be considered as a platform of questionable reliability for a while until enough time goes by without any problems popping up. They can't price their chips the same as Intel's chips running at approximately the same speed. They have to under cut them by a significant margin to actually start to steal market share from them. It will be interesting to see what they have as a full line up in a few months. I wish I could wait that long before upgrading, but I can't. I really hope you are finally competitive with Intel again AMD. (and with Nvidia when Vega is released).
  • tamalero - Wednesday, December 14, 2016 - link

    You need to understand that AMD as always focused first on server market then to customers. They did the same with AMD64 and OPTERON.
  • viciouzex - Thursday, December 15, 2016 - link

    Knowing AMD it's going to cost $99.
  • smilingcrow - Thursday, December 15, 2016 - link

    The TDP is for the whole range of chips and the 8 core is not the fastest as Intel have a 10 core at silly money.
    So actual power consumption for the 8 core will be less and as mentioned in the article they actually showed the wattage for both systems and they differed by a few watts.
    Can't read too much into that as I imagine they were recording system power consumption so the choice of motherboard can have a big influence here.
  • ddriver - Tuesday, December 13, 2016 - link

    As expected, they demoed performance with open source software, which they could properly optimize. That's not a bad thing, consider that most proprietary benchmarking software is actually optimized exclusively for intel. Both rendering and encoding are perfect CPU performance benchmark workloads, and I for one would not complain about not using cinebench LOL...

    So performance is very much competitive, and given the TDP (with a grain of salt) the uarch seems to be efficient. So hopefully amd has rysen from the dead when it comes to CPU performance.

    One thing I wanted to see is announcing the price as well. All in all, the even was kinda amateurish, full of forced out enthusiasm, lots of repetition, lots of irrelevant stuff, too much "how cool is that?".
  • 0ldman79 - Tuesday, December 13, 2016 - link

    Agreed. She may be good at running AMD but she isn't a master at working the crowd at events.

    I found it a little distracting and I imagine I'm not the only one. I'm paying attention to what is going on, watching the demo, she's talking trying to get a reaction. Dead silence while watching is my reaction... lol

    Looks very interesting. Price?
  • ddriver - Tuesday, December 13, 2016 - link

    Slides are pretty lame too, WTF is "artificial network"? The love child of "artificial intelligence" and "neural network"?
  • MobiusPizza - Tuesday, December 13, 2016 - link

    It should have ben artificial neural network
  • Meteor2 - Wednesday, December 14, 2016 - link

    I actual like that Su comes across as a pretty normal person.
  • Railander - Wednesday, December 14, 2016 - link

    she comes off to me as overly trying to keep appearances.

    want a normal and honest sounding person? jeff kaplan.
  • 0ldman79 - Thursday, December 15, 2016 - link

    I'll give you that.

    I'd be like a deer in the headlights... lol

    Not my thing.
  • mikato - Friday, December 16, 2016 - link

    Su comes across as an engineer. Good! Lame public appearances? Who cares. I surely would find it distracting too, she's probably forcing it, but I don't watch events like that. The chips can speak for themselves hopefully. And so far her work has spoken for her as well.
  • bananaforscale - Wednesday, December 14, 2016 - link

    Then again, Handbrake isn't open source and Intel lost by about 10%. That's not optimization, that's simply a faster CPU.
  • bananaforscale - Wednesday, December 14, 2016 - link

    (Well feck, I've been using HB for years and haven't noticed it *is* open source. Ignore. :D)
  • m-p-3 - Tuesday, December 13, 2016 - link

    So far it seems to beat the 6900K by a 10% margin in Handbrake with the same file and transcode profile, according to their live stream on YouTube.
  • Tewt - Tuesday, December 13, 2016 - link

    They did do a Blender test against a 6900k. At stock speeds, Ryzen rendered the image at essentially the same time as the Intel part. Lisa Su then made a point that is without the CPU's tuning feature running(if I understood what tech she was referring to correctly.) that increases performance. Not sure how many people think Blender as a good benchmark.
  • Tewt - Tuesday, December 13, 2016 - link

    Oh yeah and as mentioned already, it did it at a lower TDP than Intel. Nice. :)
  • casperes1996 - Tuesday, December 13, 2016 - link

    I'm sure anyone who uses Blender would fine it a pretty good benchmark of what it's worth to them
  • xype - Wednesday, December 14, 2016 - link

    "Not sure how many people think Blender as a good benchmark." As long as it’s open source and people can actually check the code to see what’s going on, I’d say they’ll be happy with it being used for benchmarking.
  • Tewt - Thursday, December 15, 2016 - link

    The reason I said that is because I've seen others mention "oh, they chose Blender because they can tweak the code to favor AMD chips." They made it sound like using open source software invalidates any benchmarks shown.
  • brumsky - Thursday, December 15, 2016 - link

    Don't forget the Blender test uses AVX2. That means the 6900k actually down clocks to, correct me if i'm wrong, 3.1GHz and doesn't turbo up at all.
  • Meteor2 - Tuesday, December 13, 2016 - link

    I think Zen is the architecture, why Ryzen will be the brand name -- for both desktop and notebook, interestingly.

    Hopefully this means goodbye to the A12 FX875350380 nonsense.
  • AS118 - Tuesday, December 13, 2016 - link

    I agree. "Ryzen" is a terrible name, and "Phenom Z", regular "Zen" or even "FX Next" or just anything other than the trashy 90's "extreme" sounding "Ryzen" would've been better. Sometimes I wonder if AMD even has a real marketing department, or if they just have someone really tasteless in charge.

    Ask an average consumer if they want a "Core i5" chip or a "Ryzen" chip, and they'd just laugh and ask you if you were serious with that second name. It's so bad, I think many people would literally think it's a joke.
  • 3ogdy - Tuesday, December 13, 2016 - link

    No. Ryzen is just fine and makes sense too.
  • bji - Tuesday, December 13, 2016 - link

    It's just a name and I have no problem with it. Maybe it sounds weird at first but trust me, Athlon sounded a lot weirder back in the day.
  • tarqsharq - Monday, December 19, 2016 - link

    Remember Duron?

    Oh boy, what a dull sounding name.
  • doggface - Wednesday, December 14, 2016 - link

    I assumed Ryzen was analogous to Broadwell or Skylake.
    Codenames for the processor generation.

    So where as previously we had bulldozer based arch we now will have zen based arch's that will be called *zen.

    As to consumer branding, I would also assume they will keep with their APU series and Fx series branding.
  • NesuD - Wednesday, December 14, 2016 - link

    Actually the Question would more appropriately be to ask a consumer if they want a Core i5 or a Rhyzen s5 chip. Since that is how it will actually be marketed.
  • lordken - Wednesday, December 14, 2016 - link

    SR5 you meant
    SR7 SR5 SR3 are expected names, as far as I heard
  • mikato - Friday, December 16, 2016 - link

    It sounds like an anime character. It's he from Dragonball-Z? I do like Zen better, but don't really care. Very true that Ryzen is much better for name recognition.
  • eek2121 - Tuesday, December 13, 2016 - link

    I like the name, though I'm surprised they didn't decide to pronounce it 'risen' instead. Can't wait for the benchmarks. The stuff that has leaked so far is very promising...slightly edging out the IPC of all of Intel's offerings. Kaby Lake may still end up being a couple percent faster, but only due to it's 4.2 GHz clock speed vs Ryzen's 3.4 GHz...but AMD hasn't detailed it's quad core chips yet either. I wouldn't be surprised to see a 4.2-4.5 GHz quad core chip.
  • wow&wow - Tuesday, December 13, 2016 - link

    "Wonder why AMD didn't stick with Zen instead of Ryzen."

    Zen is too generic to be a product branding?

    Ryzen (pronounced "risen") sounds better and means better?
  • phoenix_rizzen - Wednesday, December 14, 2016 - link

    Except that it's pronounced "rye-zen", not "rih-zen". Makes you wonder if the next iteration will be "wheat-zen", then "barley-zen", then "other-grains-zen", etc. ;)
  • Lolimaster - Thursday, December 15, 2016 - link

    Probably Goxen.
  • Lolimaster - Thursday, December 15, 2016 - link

    Gozen* No edit in 2016 GG
  • zeos - Tuesday, December 13, 2016 - link

    because zen is too close to xen. They didn't want to be associated with the taint of half-life's worst level.
  • shing3232 - Wednesday, December 14, 2016 - link

    Well, Ryzen is the name for the processor, and Zen is for the core
  • Shadowmaster625 - Wednesday, December 14, 2016 - link

    Googling anything Zen related produces a lot of garbage results. Ryzen, on the other hand... It is good to have a unique name.
  • Ratman6161 - Wednesday, December 14, 2016 - link

    "I think Zen had a better name for marketing". Perhaps because Asus already has so many products out there with "Zen" in the name??? Zen Pad, Zen Book, Zen Phone, etc...
  • wishgranter - Tuesday, December 13, 2016 - link

    Yup looking on the stream now..
    look very interesting !!!!!
  • Meteor2 - Tuesday, December 13, 2016 - link

    Live stream was very good. They showed a Ryzen+Vega computer playing Rogue One in 4K at better than 60fps. Looked really rather good!
  • Meteor2 - Tuesday, December 13, 2016 - link

    God, if AMD can deliver on these promises, and price the chips competitively, Zen could be quite something.

    XFR sounds amazing, in particular. I see no reason why a human could tune a CPU better than the CPU could do itself. Clearly it'll be limited by temperature.
  • darkfalz - Tuesday, December 13, 2016 - link

    Probably would never buy an AMD platform myself but if it stops Intel releasing tiny 5% IPC improvements every 2 years then it's a win for everyone. 10 years old Intel and NVIDIA only being held back by their own products hasn't exactly been great times (although a longer upgrade cycle is kinder on the wallet).
  • Tewt - Tuesday, December 13, 2016 - link

    It's not a win if AMD cannot sell enough and have decent margins for more than one quarter. Unfortunately, your reaction is something I fear where AMD cannot shake the bargain-bin CPU and even if they match or surpass Intel performance, they are not allowed to charge a premium like their competition.
  • Threska - Tuesday, December 13, 2016 - link

    Depends upon the market. Consumer? Yeah. Servers? Premium is the norm. It's also a very big market for AMD, in which Intel has a near monopoly.
  • Tewt - Thursday, December 15, 2016 - link

    "...in which Intel has a near monopoly." That's the problem and ignores which market you are talking about because Intel dominates them all. I want AMD to enjoy multiple quarters of profit, IF Zen is actually competitive just to clarify, because of Intel's past shenanigans. Intel has too much influence due to their cash they can throw at the problem such as engaging in a price war to keep their marketshare to which AMD has to respond thus again giving them slim margins or Intel can again give 'incentives' to retailers or OEMs to showcase Intel products, relegating AMD to sub-par parts like providing only single channel memory in laptops when dual-channel is supported, consumers having to actively search for any AMD product on the site, etc.

    AMD has had competitive products in the past but only in the recent past did we find out why their quarterly profits did not rise in proportion.
  • bigboxes - Tuesday, December 13, 2016 - link

    Well, I'd gladly go back to AMD if they put out a competitive chip. I was with them ten years and they were great machines. On my second i7. Would love to have an 8 core/16 thread beast in a mainstream sku.
  • 0ldman79 - Tuesday, December 13, 2016 - link

    My main box is an FX 6300, but my new laptop has a Core i5 6300, 2.3GHz and it is faster than my overclocked FX.

    How this plays out will determine what my TVBox's FX will be replaced with. Comes down to price and availability if this demo was accurate.
  • Azix - Tuesday, December 13, 2016 - link

    Silly thing to say. They made the 6700K look like trash essentially. I do think a lot of people would still buy a quad i7 with 8 threads over a 8 core 16 thread zen beast, stupidly of course
  • barleyguy - Thursday, December 15, 2016 - link

    The test against the 6700k was kind of weird because most streamers with a 6700k would be using Shadowplay (or now AMD's competition) to do the rendering for the stream rather than trying to encode on the CPU. So it was a strangely gimped demo of the 6700k.

    That said, it did show the advantage of having 8 cores, which is that you can game and run something else in the background at the same time.

    I am absolutely buying a Ryzen, assuming they price it reasonably, which I fully expect. Well, I also need a motherboard with a PCI slot. That probably sounds silly, but I have a very expensive studio sound card (RME Audio) that I don't want to have to replace.
  • crazylocha - Tuesday, December 13, 2016 - link

    And if you could buy the Ryzen CPU, Vega GPU, and MB for under the price of Intel's 6900k+MB? And if Ryzen matches or beats Intel's #'s?
  • 3ogdy - Tuesday, December 13, 2016 - link

    What planet do you expect that to happen on, again?
  • Meteor2 - Wednesday, December 14, 2016 - link

    This one. Have you seen the price of a 6900K+MB? Why shouldn't a high-end AMD CPU GPU and motherboard combination cost less than $1,500?
  • TheinsanegamerN - Wednesday, December 14, 2016 - link

    Because AMD constantly being the bargin bin chip will lead to them going bankrupt? Consider that vega chip alone will probably be $550 or so, even more for the big daddy vega 11, the 8 core CPU might be around $500. Hitting $1500 with all the highest end stuff isnt too far outside the realm of possibility.
  • Targon - Wednesday, December 14, 2016 - link

    I am expecting to pay $500 for the CPU, $250ish for the motherboard, then RAM. If it is competitive, then over time, it will save me a fortune. Remember, in two years, we can expect three new sockets from Intel, while AMD will still be on socket AM4 or AM4+, so I won't need to do more than upgrade the CPU. Grab a new motherboard, and I will have two AMD machines for a lot less than if I were dealing with Intel socket insanity.
  • Michael Bay - Wednesday, December 14, 2016 - link

    >zen
    I`ll wait for the actual units to emerge before deciding.
    >vega gpu
    Never ever.
  • leafar12 - Saturday, December 17, 2016 - link

    So those plus together Ryzen + Vega + MB + lowest TDP + lowest price, we could called
    Godzen
  • diglo1 - Tuesday, December 13, 2016 - link

    Well Zen is now matching 6900k when Ryzen is locked at 3,4ghz and 6900k is allowed to boost as much as it wants to. Now I wanna know what kind of clockspeeds 6900k can handle when doing these types of rendering/encoding tasks. If 6900k managed to pull that 3,7ghz then Zen's IPC has surpassed Skylake. Tom's Hardware tested 6900k and they managed to sustain 3,7ghz all across the board when doing stresstesting so by that RYZEN matches 6900k with lower clockspeeds. Now to be fair that 6900k might have only went up to 3,3-3,5ghz so lets say they are even.
    Now of course RYZEN has not enabled all of it's turbo technologies so those things alone suggests that RYZEN is gonna be superior to 6900k.
    Now even if all of that still only matches 6900k when done with more benchmarks, RYZEN still has 95w tdp which pretty much says that Zen uses less then 100w of power. Again when Tom's Hardware did their benchmarking they did hit 140w limit so I can pretty much proclaim Zen being more efficient then Skylake CPUs, though that might not be the case at higher clockspeeds.
    I still think that with Intel's boost clocks of 4,5ghz they still have the singlecore performance crown.
    Anyway nice to know that AMD is making a stand.
  • CiccioB - Tuesday, December 13, 2016 - link

    I think you are running too much ahead of facts.
    We know how good is AMD at making fake comparisons and then let down.
    Fiji and Polaris are not enough? So let's add Bulldozer being in par with i7 only on selected multihreaded tests just to be heavily beaten on anything that used less than 8 threads? Is now enough?

    I too hope this Zen is really good. I was also waiting for Bulldozer to be. But I have now learned to keep quiet and wait for the real world benches.
    I do not care for the chip to be faster than this or that. I just want it to be affordable and most important not to be an oven in my silent case. If it is 20% slower, costing 40% less and using half the power would still be in my wish list while I wait for my PC (i7-3770K based) to become obsolete.
  • sharath.naik - Tuesday, December 13, 2016 - link

    I second that, AMD has been doing these misleading benchmarks for more than a few years in a row. I understand it is to boost their stock for a short time/ sell a few more because of the hype. But they have made it into a habit, simply put if history goes even this sounds likes the same.
    I really hope at the very least they can deliver on a per core performance. I seriously doubt them even getting close to Intel's TDP for multicore, so not expecting much on multithread performance. With NVidia's Pascal not sure about what benefit a fast IGPU will bring to the table, but would provide an option to skip dGPU in a budget laptop I guess.
  • rhysiam - Tuesday, December 13, 2016 - link

    I agree absolutely with the history of over hype, and you may well turn out to be right. However, at least this time AMD is giving us quite a bit more information and basing their performance claims on well known open source tools that are both useful in the real world, and reasonably indicative of overall multi-threaded performance.

    Past over-hype was often based on hand picked benchmarks. For GPUs we were told about performance at 4K in AotS and Doom, for example, which isn't indicative of overall gaming performance. Other hype fuelling releases came from press statements with vague performance claims, or often even rumours or leaks which care more about grabbing attention with extravagant headlines than balanced analysis.

    At least this time we have an actual AMD CPU running open source and widely tools in Handbrake and Blender, compared to Intel's >$1K 8 Core CPU. OK, maybe they searched for a specific encoding profile which best suits their architecture. But it's not like Intel is somehow weak on these applications. In fact, Handbrake and Blender are exactly the sorts of workloads that motivate people to go out and spend big dollars on an 8 Core Intel workstation.

    You're right to say that we need to wait for release and independent benchmarks. But it's also fair to say that things are looking very positive.

    One quick note of clarification... @diglo1 suggests a 95W tdp and thus vastly better efficiency. But the handbrake test showed an 8 Core Ryzen @ 3.4Ghz running within a couple of watts of the 6900K in the same workload. That's absolutely more than 95W. It's impressive to achieve similar efficiency. But no one's claiming vastly improved efficient over the modern Intel architectures.
  • caqde - Tuesday, December 13, 2016 - link

    Remember where AMD was in Handbrake 0.9.9 (Anand's 2x4K test) where the A10-7870K (3.9ghz) scores 10.78FPS vs the i3 4330 (3.5Ghz) that gets 10.63 fps or the i3 4360 (3.7ghz) that gets 11.21 fps. Zen is doing phenomenal in this regard. AMD did horribly at this workload with the Bulldozer derivatives. (A10 being Steamroller). Even AMD's fastest FX-9590 (4.7Ghz) scored 19.05fps which was slower than Intel's i5 4690 (3.5ghz) that got 19.53fps. This means that AMD's new architecture is over twice as fast as it's old one in handbrake. And to note based on some math the i3 core was ~10% faster than each Kaveri (Steamroller) Module. The i5 core on the otherhand was 37.7% faster than the FX-9590 Vishera (Piledriver) Module. Again it seems AMD's new Zen core is impressively faster than their old Steamroller Module.
  • Meteor2 - Wednesday, December 14, 2016 - link

    What was fake about the Polaris demo? It was against a 960 iirc, and was not misleading.
  • CiccioB - Wednesday, December 14, 2016 - link

    No, it was again a castrated chip (gtx950) a most important with vsync on.
    The 85W vs 140W meant nothing, as we saw 950 in the end was faster when not capped. So was apple vs orange comparison resulting in a fake win for AMD
  • Meteor2 - Friday, December 16, 2016 - link

    I see what you mean. But when I watched the demo, I took it to mean AMD had reached performance/power parity with 9xx Nvidia. I don't think that's wrong.
  • thisguy365 - Wednesday, December 14, 2016 - link

    The i7-6900k is actually Broadwell-E not Skylake. Intel does a weird thing where their HEDT processors are a generation behind the Mainstream line. So Ryzen has better IPC then last gen intel. I would have been interested to see a true comparison of a quad core Zen against the 6700k(Skylake mainstream). The one they did with Gamin+encoding is unfair as both the 6900k and Ryzen have 4 cores over the 6700k to deal with the encoding.
  • tarqsharq - Monday, December 19, 2016 - link

    It would only be unfair if they were priced apart from each other. If the 8c/16t is in the same price bracket as the 6700k I would consider that fair.

    But since we don't have a clue on pricing yet, that is TBD.
  • jjj - Tuesday, December 13, 2016 - link

    The tone is dismissive so well beyond negative.
    The numbers don't quite match the live demos either but lets say they got 10% extra perf in Handbrake in the last minute.
    Claiming that the Intel with base clocks at 3.2GHz and turbo at 3,7GHz is similarly clocked to the 3.4GHz Zen is quite bold too.
    If you are too bored to make an effort,have a beer or something.
  • iranterres - Tuesday, December 13, 2016 - link

    There's one thing that they finally showed though, they're going to have some VERY competitive products, and i'm not talking about just performance, but in the entire segment.
  • Targon - Wednesday, December 14, 2016 - link

    The Ryzen processors used in the demo had Turbo disabled, just so show it wasn't a blip while the processor was cool but would be lower in the real world. With turbo enabled, performance should be quite a bit higher.

    Keep in mind the price point of the Intel that was used. Intel with base of 3.2 with Turbo, vs. 3.4GHz without Turbo. If Turbo was enabled on the Ryzen processor and it was hitting 4.5GHz, people would say it isn't fair because both base clocks and Turbo are higher than the $1100 Intel processor it is aiming to beat.
  • Pinn - Tuesday, December 13, 2016 - link

    boooooring, especially with 2 separate HBM monsters on the horizon in the video card space.
  • sorten - Tuesday, December 13, 2016 - link

    Looking good. I hope this holds up in independent testing and makes AMD competitive again. I won't be switching back any time soon because gaming hasn't been CPU bound for a long time and I would also need to replace the motherboard. But if they get some design wins in the 2-in-1 market (Surface Pro 5?) then I'd be a customer.
  • Death666Angel - Tuesday, December 13, 2016 - link

    "It is pronounced ‘Rye-zen’, not ‘Riz-zen’, to clarify."
    You realize there is a Phonetic Alphabet, which is internationally understood, right? Please use that.
  • iranterres - Tuesday, December 13, 2016 - link

    You forgot to mention that the phonetic alphabet you say is very complex and needs specific linguistic knowledge. So try nitpicking something else next time.
  • 3ogdy - Tuesday, December 13, 2016 - link

    You do realize nobody gives a damn about that and people reading this article, including you, have understood exactly how to pronounce the word without the phonetic bullshit, don't you?
    If you don't, then please do. Cut us some slack.
  • bji - Tuesday, December 13, 2016 - link

    Yeah I haven't used the phonetic alphabet since they used to make us write out word pronunciations with it back in the 80's in grade school. Always thought it would be cool to replace English hodge-podge spelling with the phonetic alphabet everywhere though. Of course then you'd have to deal with regional differences in pronunciation causing different spellings of the same word ... but at least we'd all know how to say them (until pronunciation deviated from the phonetic spellings over time of course).
  • MajGenRelativity - Tuesday, December 13, 2016 - link

    The Hype Train certainly passes by my house. Looking forward to seeing the release and reviews! If AMD can compete, I may go red team. If they can compete.
  • bill.rookard - Tuesday, December 13, 2016 - link

    Agreed. My main rig is certainly powerful enough right now with a OC'd Xeon X3470, but it could use an update. If Zen (I refuse to call it Ryzen at the moment) is competitive, and price appropriate, I may switch. It doesn't even really have to beat Intel, it just has to be -close enough-.

    Give them a run for their money, something their previous architecture couldn't accomplish.
  • HisDivineOrder - Tuesday, December 13, 2016 - link

    I hope AMD does well. That said, I think there's a reason AMD isn't announcing price. If they wanted to make a big splash, they'd have said, "Look at our performance and AT THIS PRICE!"

    I think people are going to be disappointed.
  • rhysiam - Tuesday, December 13, 2016 - link

    They can't really lock in pricing until they've finalised clock speeds. So I'm not surprised they're holding back on pricing announcements for now. Also, they're only talking about competing with the incumbent Intel, not beating them, so their only real point of competition is going to be on price. As the underdog in the market, they have to undercut Intel on price if they want to offer a compelling alternative to the status quo. Having said that, if their 8 Core 3.4Ghz CPU turns out to be something like $800, which is a sizeable ~$250 less than the 6900K... I agree that it would still be a huge disappointment.
  • firerod1 - Tuesday, December 13, 2016 - link

    IF it beats or matches the 6900k across the board in most tests/benchmarks, they will price it around $500-600. Which is how much a 6800/6850K costs, plus if they make a slightly lower clocked RZ-8/16, they could charge $400-450 and have their 6C/12T($200-350) do battle with i7-7700/k, i5-7600/k and the quad core APUs handle regular i5s/i3-7350K.
  • TheJian - Tuesday, December 13, 2016 - link

    Management should be fired if they follow your advice. Price like Intel if you match or beat them, just like the old days when they won for 3yrs. No need to do you any favors until they're stuck on a shelf not moving. Both Intel and Nvidia raised prices of top end stuff and sold even more, so clearly there is still room to move higher and still sell. More people are paying for top end stuff these days. The only thing that stopped them from selling more back then was the inability to produce more. That should not be a problem today, so price right and pay off your debt. They won't get another shot at catching intel off guard.
  • Meteor2 - Wednesday, December 14, 2016 - link

    AMD would die if they followed your advice, Jian. AMD also have to compete against Intel's brand. With equal performance and equal price, would you buy the newcomer or the established, reliable brand?

    No, to get any kind of market penetration, AMD will have to be cheaper.
  • TheinsanegamerN - Wednesday, December 14, 2016 - link

    AMD isnt a newcomer, they've been in the CPU game for a LONG time.

    And who cares about the brand? It's a CPU, it's about the performance, and there is pent up demand for something to beat the intel chip back a bit.

    AMD's "enter the market cheaper" strategy has spectacularly backfired more then once, and has left them with a tenth the money of nvidia. they need cash to survive, not positive thoughts.

    If they can take the 6900k to task, they should absolutely charge as much as possible for it, to recoup all the money theyve been loosing. AMD fans will buy it, anyone who cant afford it will wait for the cheaper 6 and 4 core chips.
  • Targon - Wednesday, December 14, 2016 - link

    People tend to buy products based on name recognition. You don't think Intel pays for advertising for nothing, do you? AMD is an unknown to most of the general public, and to compete with Intel, they need to go with lower prices at this point.
  • lordken - Wednesday, December 14, 2016 - link

    And who cares about the brand? well isheep does and similar kind of ppl...sadly while I agree with TheJian about pricing - though I would not price exactly as intel, a few bucks lower - I'm afraid that average potato Joe that will be buying new PC (who doesn't know what CPU is) would probably opt for Intel because of brand name.
    But if Zen will perform flawlessly and will gain positive reviews and article news around the web, maybe with time even potato Joe would ask for AMD...

    Personally I would love to get SR7 (6900k perf) at 400-500$ but as it was said, we need AMD to get profit, if they dies we are screwed and future is bright with i3 at 500$ :)
  • barleyguy - Thursday, December 15, 2016 - link

    Yeah, it's one of my pet peeves when people call AMD a newcomer, as they've been in business since 1969 and been making general purpose microprocessors since the 8-bit days. Intel was one generation ahead, because they had the 4004, which Gordon Moore drew on a piece of graph paper.
  • barleyguy - Thursday, December 15, 2016 - link

    Scratch that last part, sort of. Wikipedia says the 4004 was designed by Federico Faggin. But it was done at Intel, and it was done by hand on paper.
  • Meteor2 - Friday, December 16, 2016 - link

    'Newcomer' to the average consumer. Not people who read Anandtech who are familiar with the Fairchildren :)
  • haukionkannel - Wednesday, December 14, 2016 - link

    Yep. The have too se how Many chips They can get that clocks up to 3.4 and how Many only manage to get 2.8 GHz or similar. The 3.4 GHz is the flagship the best binned cpu.
    So the price will depend a lot of how Many of those cpu can achieve the higher speed.

    Also the have to see what Intel do in the meantime. But all in all, getting back the competition in the CPU Front is a big deal, even if And when Intel keeps the ip lead, the perf/dollar can improve for a long time.
  • Targon - Wednesday, December 14, 2016 - link

    It seems you didn't watch the event. 3.4GHz is the low end, with 3.4+ being displayed, indicating that at launch there will be faster versions. You are right that price will be based in part on the yields, but I expect we will be seeing 3.8GHz at launch, with 4.2GHz being easy to hit.
  • 3ogdy - Tuesday, December 13, 2016 - link

    Thinking about it now...Ryzen R9 , top of the line CPU from AMD, great match for the R9 cards.
    I'm curious about how AMD will end up naming the chips, since it will most likely segregate the performance tiers to match Intel with its three main DTCPU lines.
  • firerod1 - Tuesday, December 13, 2016 - link

    Ryzen-SR8, Ryzen-SR6, Ryzen-SR4. SR= Summit Ridge, Numerical value= The number of cores.
    Core i7-7700k, Core i5-7600k, Core i3-7350k, you tell me what those numbers mean!
  • Meteor2 - Wednesday, December 14, 2016 - link

    But everyone *knows* what those numbers mean. Intel's branding follows a consistent pattern, and a higher number is better performance.
  • fanofanand - Wednesday, December 14, 2016 - link

    This is a factually inaccurate statement. For Intel's processors a higher number only means a higher performance tier within a given tdp. For instance, the 6700T is slower than a 6600k. This gets more pronounced as you get down to the Y processor family. So no, Intel's naming scheme is not logical unless you know the ins and outs of Intel's naming.
  • Meteor2 - Friday, December 16, 2016 - link

    Heat rejection (and thus what TDP class chip is used) is set by the chassis, so assuming you're comparing devices in the same form-factor, higher numbers do mean higher performance. I don't think many people go shopping comparing gaming desktops to ultrabooks.
  • thisguy365 - Wednesday, December 14, 2016 - link

    (class)-(gen)(segment)(spot in series generally definedby frequency)0(k skus are unlocked and overclockable)
    i7-7700K i7-7th gen, top line mainstream, only one in the seiries, unlocked for the entusiasts.
  • Kaihekoa - Tuesday, December 13, 2016 - link

    No benches? Wtf
  • Makaveli - Tuesday, December 13, 2016 - link

    This is not a product review

    Wtf!
  • mkaibear - Wednesday, December 14, 2016 - link

    It's not headlined as a review, at no point does anyone say it's a review, and the hardware in question isn't even out yet.

    Why on earth were you expecting a review?
  • Tchamber - Tuesday, December 13, 2016 - link

    I'm pretty much going to get the AMD offering. I've had Intel since the debut of the i7 920, and my current rig is just its big brother, the i7 970. It's still 'fast enough', so anything with DDR4, PCIe, and SATA 6gig, and or m.2 is going to be a nice improvement. I game only very casually, and most of what I do is photo-related. I'm actually putting off building so I can support the under-dog...wish me well!
  • fanofanand - Wednesday, December 14, 2016 - link

    I'm with you. I've been waiting 2 years to build a new rig just so I can make sure to support AMD. I'm not a fanboy, my current CPU is a Q6600 (so yes I'm WAY overdue). I don't need top of the line, I just don't want to make a stupid purchase. If AMD is even close to Intel I will be buying.
  • webdoctors - Wednesday, December 14, 2016 - link

    No one preorders a CPU, just like a video card, I don't trust any of the companies or benchmarks until done by a third party. Until I see tests posted by Anandtech or equivalent company, AMD can continue to "publish" whatever results they want. Just like Nvidia/Intel/IBM/ARM, the real proof is what actually shows up on shelves. Just like how car prototypes look amazing but the finished product is a heavily neutered version.
  • epobirs - Wednesday, December 14, 2016 - link

    They actually mentioned SATA Express? Or do they mean the U.2 replacement for SATA for the never never supported with actual drives SATA Express?
  • asmian - Wednesday, December 14, 2016 - link

    Yeah, I noticed SATA Express too. Like WTF? That was in the revamped Broadwell-E boards too, despite no-one wanting it and not a single product for it years after it was specified. I cannot understand the engineering mindset behind that, unless it's just a stupid tickbox marketing thing.

    All the rest, though, looks very interesting.
  • HollyDOL - Wednesday, December 14, 2016 - link

    I'll stay calm and wait for thorough non AMD made reviews. There has been way too much promiseware coming from AMD last years for my liking.
  • HollyDOL - Wednesday, December 14, 2016 - link

    ** mean reviews with benchmarks
  • kevlarkian - Wednesday, December 14, 2016 - link

    so... I just ran their blender test on my 2010 Mac Pro ([email protected], 28GB ram, GTX 980), and it came back in 39.72 seconds. I'm suspecting the test wasn't particularly taxing, and cherry-picked for its near-identical returns across both CPUs.
  • RdVi - Wednesday, December 14, 2016 - link

    x5670's can turbo up to 3.33GHz. Also 12 cores at ~3ghz, you're pretty much looking at a 50% advantage that can only be eroded by poorer IPC. We all know how little Intel's IPC has increased since then, it's certainly hasn't been by a factor of 50% total.

    Also the fact that you are comparing $2600-3000 worth of CPU's in 2010/11 money to one that will surely be under $1000 is telling.
  • fanofanand - Wednesday, December 14, 2016 - link

    Seems disingenuous to point out it's a 2010 when your hardware is substantially newer.....Just sayin'
  • T1beriu - Wednesday, December 14, 2016 - link

    "We’re still waiting for detailed information on (...) power consumption (...)"

    I guess you missed when Lisa said the TDP is 95W.
  • alphajoza - Wednesday, December 14, 2016 - link

    What about this pricing?
    http://www.3dnews.ru/assets/external/illustrations...
  • iranterres - Wednesday, December 14, 2016 - link

    I dunno... $150 more for 200MHz on the same silicon... Sounds weird, unless we're talking about less cache juice to come. By the other end, the last 2 options are very promising (assuming this price list is meant to be real)
  • iranterres - Wednesday, December 14, 2016 - link

    It's absolutely undeniable that AMD's new CPUs are going to be competitive, but not sure about the true performance envelope....
  • jamyryals - Wednesday, December 14, 2016 - link

    Unlike quite of few of the people here, I like the name Ryzen. XFR sounds like one of the best features for regular folks since Turbo Boost was introduced. Hopefully AMD can get good yields so they can put these chips out at a good price.
  • brumsky - Wednesday, December 14, 2016 - link

    I think the Blender test revealed some interesting information regarding Zen's implementation of AVX2. I'm paraphrasing, so bare with me. According to Anandtech's own article it splits a 256 bit instruction so it can run across 2 128bit units.

    I ran the test on my single E5-2683 v4 16c/32t and it completed in 44.02 seconds. However, I had HWiNFO open because I wanted to see how it would handle turbo during the test. I discovered it dropped down to 2.0GHz from 2.1GHz which means it was using AVX2. Otherwise, the all turbo speed is 2.3GHz.

    Since the 6900k is a Broadwell chip, this means it does the same drop in frequency as my E5. Someone with a Broadwell please confirm the exact speed but I'm going to guess it drops the same 100 MHz when AVX2 is in use. I do know Broadwell can independently change the speed on a per core basis, compared to Haswell, when running AVX2. Why does this matter?

    This means the 6900k wasn't running 3.2GHz with boost up to 3.7GHz but at a reduced speed of 3.1GHz without boost.

    It was said that the Zen chips boost wasn't enabled. However, they didn't mention anything about how it handles running AVX2 instructions. Specifically if it has to drop it's speed the same way Intel currently does.

    Also, it may appear that the current splitting of a 256 bit instruction might not have as great of a performance impact. I'm even more interested to find out what Zen can do!
  • Meteor2 - Friday, December 16, 2016 - link

    Presumably these frequency drops are because AVX2 really fills the silicon so the system is clocking down to stay within design TDP. An interesting test of whole-chip efficiency then. How widely used is AVX2 though?
  • redraider89 - Wednesday, December 14, 2016 - link

    "today AMD is lifting the lid on some new features in order to whet the appetite (and appease the hype-train, perhaps)"

    C'mon Anandtech, stop the nonsense. Intel NEVER has a hype-train does it. No, it never said the Pentium 4 could scale to 10 GHz, nah!

    "The Pentium 4 "Prescott" is, despite its innovative architecture, a failure. Intel expected to scale this Pentium 4 architecture to 5 GHz, and derivatives of this architecture were supposed to come close to 10 GHz. Instead, the Prescott was only able to reach 3.8 GHz after numerous revisions. And even then, the 3.8 GHz is losing up to 115 Watt, and about 35-50% (depending on the source) is lost to leakage power."

    http://www.anandtech.com/show/1611/6
  • Meteor2 - Friday, December 16, 2016 - link

    It's been eleven years. Let it go.
  • kitfox - Wednesday, December 14, 2016 - link

    Hmm...i'm actually impressed. I was expecting Zen to be another mediocre attempt by AMD at competing with Intel's low/mid range processors. But they're actually going after Intel's top dog right out of the gate.

    I haven't considered AMD to be a realistic option since the Athlon 64 days...this is going to be VERY interesting.
  • jsntech - Thursday, December 15, 2016 - link

    All of this. I'm very much looking forward to getting back to AMD giving Intel a real run for its money.
  • iranterres - Thursday, December 15, 2016 - link

    Looks like they're back in business!
  • Meteor2 - Friday, December 16, 2016 - link

    I'm guessing AMD felt top-line Intel was an easier target than top-line Nvidia (though I am looking forward to Vega). Given Intel's stagnation, they're undoubtedly right.
  • KarnageS - Thursday, December 15, 2016 - link

    Nice amd finally a worthy product
    But something i dont understand. They claim 95w tpd yet in the article i read ryzen comsumed a little less W than 6900k, which is far away from 95w tdp
  • fanofanand - Thursday, December 15, 2016 - link

    It had already been said a few times, but tdp is the heat dissipation not the wattage used.
  • KarnageS - Friday, December 16, 2016 - link

    AFAIK most of the power goes to heat. So that means something is wrong with Either Intel or amd number of tdp.
    Or am I thinking wrong?
  • juggernatxtr - Friday, December 16, 2016 - link

    your thinking is wrong, 95 watts means less heat dissipation required to keep the chip with in thermal design.less cooling needed.
    a better way to say it is that the intel chip is creating more heat to do the same amount of work.
    or saying that it is less efficient at the work involved. less heat generated by the chips to be dissipated.
  • juggernatxtr - Friday, December 16, 2016 - link

    TDP is not power consumption, it the power required to keep the chip cool.
  • Gigaplex - Friday, December 23, 2016 - link

    No, it's the typical heat output from the chip. Which, due to conservation of energy, is essentially power consumed. The fans don't draw 95+W.
  • SloppySlim - Saturday, December 17, 2016 - link

    amd might be setting pretty atm, there's alot of pent up cpu/mb upgrade demand due to intels higher combined cpu/mb pricing.
    i just upgraded my net machine from an amd 555x2 to a 6300fx, and i'm sitting on a 2600 i7 because the bang for buck sweet spot is a 6700k at the next price cut.

    amd might swoop that
  • orl1574 - Saturday, December 17, 2016 - link

    I hope to upgrade to Ryzen from Haswell i5 4690k. All depends on price and performance.
  • hyno111 - Sunday, December 18, 2016 - link

    "With the 100 MHz base clock on all modern CPUs"
    Some CPU of the Intel Silvermont architecture line have different base clocks, (Z8750 have 80Mhz) but not sure if they count as "modern"..
  • abufrejoval - Monday, December 19, 2016 - link

    A chip which handles pretty much everything dynamically is a nightmare to benchmark!

    And then there is the binning, of course. I remember Penryns, some of which would falter beyond 2.87 GHz and draw around 80 Watts, other would race to 3.4 GHz and draw 50 when their TDP was officially 150 Watts (X5492). So no two CPUs will be alike, benchmark wars to no end ;-)

    Neural networks: You run 20 benchmarks it will just wind up saying 42 and shut down bored!

    However, what I find most troublesome is this rumor that AMD is licensing their GPUs to Intel to replace Intel's iGPU. With Intel able to produce true HSA APUs on their process nodes, perhaps with EDRAM or HBM2, what's left for AMD?

    At the one point in time, when AMD has something that neither Intel nor NVidia can deliver, a unified HSA CGPU architecture capable of handling everything from gaming to machine learning in way that's both much more comfortable to program than CUDA and works much better on a standard Linux *and* competitive on both CPU and GPU fronts,.... they are giving their ace away.

    With AMD APU tech under their belt, Intel can throw so much OpenCL at NVidia it actually makes a dent into CUDA while they never leave enough revenue or market share to AMD to allow them grow dangerous.
  • LordSojar - Wednesday, December 21, 2016 - link

    Looking more and more likely that I finally MIGHT have a reason to upgrade... Intel sure hasn't given me a reason to ditch my 2500k @ 5.0GHz.

    Maybe 2017 will be the year I upgrade to a "many core" CPU and finally move to 4K with either a Vega or Volta GPU.

    Weeee, progress!
  • Milocinia - Monday, January 2, 2017 - link

    I'm very much interested in this, not because I want to buy one but because there might finally be some competition pressure put on Intel. I don't have a great deal of money to be building/upgrading my home PC so when my motherboard failed a couple of months ago, I had to keep the cost extremely low.

    1155 motherboards had become almost impossible to find so I looked into a motherboard with an up to date socket and a new processor. After looking at benchmarks, I couldn't believe how much money I'd have to spend on a current gen processor as good as or better than an i5-3470! 3-4 "generations" down the road and it didn't look like there had been any tangible improvements to performance - to a user of average intelligence like myself, not a knowledgeable enthusiast. Certainly not enough of a performance improvement to justify the cost of a new processor.

    Thankfully I managed to source an 1155 motherboard at a reasonable price.

    So however good AMD's latest products are or aren't, the performance competition pressure on Intel is most welcome.
  • IntoGraphics - Tuesday, January 3, 2017 - link

    Yes, on paper everything is possible.

    "You make me ..... Promises! Promises!
    Knowing I'd believe ..... Promises! Promises!
    You knew you'd never keep."
    https://www.youtube.com/watch?v=WBupia9oidU&t=...
  • IntoGraphics - Wednesday, January 4, 2017 - link

    AMD is a 4 letter word. And it is not good or best.
  • Cooe - Thursday, May 6, 2021 - link

    Hahahahaha. You were saying? ;)

Log in

Don't have an account? Sign up now