Good to see that power consumption is down (would like to get more data on it). I've been wanting to my my Phenom x3 that I have an office computer with an 8-core, and I might just do it with this one. Black Friday!!!!
Hey, be nice to AMD. These new processors are totally competitive with Intel...'s Sandy Bridge processors from 3 years ago.
Seriously AMD, Intel has almost been standing still for 3 years and you're still behind. Come up with something better or us enthusiasts are not going to have a reason to even look at your chips. Your GPUs are a good option, with competitive performance and better prices (I recommend Radeons in almost every price bracket right now because of this). If only AMD could put out some decent CPUs.
Has anyone been looking at their chips seriously?? I stopped using AMD a long while ago. Intel blew them away and it'll take some competitive chips from AMD to get me to think otherwise.
Don't get me wrong, I sympathize with underdogs, and especially considering in the case of AMD, it was decades of Intel's anti-competitive practices that have pretty much doomed AMD to be a perpetual runner up, but I can't deny the obvious. AMD is where Intel always wanted it to be - a handicapped "competitor", barely enough for Intel to say "we are not a monopoly". I've been a long time AMD consumer, but today it is not worth it, maybe on the low end where power consumption is not that much of an issue and the lower price may make some sense, at least in the short term. But ever since I became a prosumer, AMD CPUs simply have nothing to offer. Radeons are pretty neat though, since they offer tremendously better DP performance for the buck.
Seriously, how hard is to make a decent CPU... Intel surely has a manufacturing lead as well, but the poor performance of AMD is mostly because of the inferior design, process might give Intel 20-30% advantage, but the rest is all the design.
On a side note, the CPU performance stalling is actually a good thing, even if AMD was competitive, there are physical limits which CPUs will hit soon, not to mention Intel has already reached the point the chips are too small to displace that much hear efficiently. I remember I used to upgrade CPUs every 18 months or so, but now I have a 2 year old CPU that is only a notch slower than the latest and greatest.
Moore's "law" is a dud, it is obvious that performance cannot exponentially grow, it is more of a half of bell shaped curve, as we reach physical limits, performance gains will slowly go stale. And quite frankly - that's OK, enough is enough. Did we really make much good use of extra performance? Nope, we got sloppy technologies, slow programming languages, bloated runtimes which kind of "equal out" the experience. I have 10 tabs in my browser open, with total content of all websites not exceeding 100 MB, and yet it takes 1200 MB of ram... Lousy skype, a puny messenger - 100 MB of ram usage? WTF?
"not to mention Intel has already reached the point the chips are too small to displace that much hear efficiently." So why is my i7-4770k @ 4.5GHz and 1.35V cooler (60°C core average while running Prime95 27.9, delidded) than my old i7-860 @ 3.8GHz and 1.34V (80°C core average with same load)? It is pretty much just Intel cheaping out on manufacturing of the consumer line CPUs, with weak TIM and too much glue for the IHS.
"Moore's "law" is a dud, it is obvious that performance" Way to show that you don't know jack. Moore's Law just says that the number of transistors on the same area will double every 2 years. Nothing about the performance of said transistors.
You forgot the economics portion of Moore's law, which ties number of transistors per dollar. If there wasn't a financial incentive, we would have never had the law in the first place ;) Shrinking didn't just improve die sizes (and almost always performance), but it also made economic sense. If you take into account the economic aspect, ddriver isn't far off. We're not far off heretofore hypothetical scenario of 'It's just too expensive and doesn't provide enough benefit anymore'. In fact, for some semicos we're already there.
Even so, that is impossible just as well. This would require that transistors can get infinitely small, which is simply not the case. There are limits to how small a transistor can be, and even in the case of insanely slow and expensive atomic assembly transistors will still be several atoms big, and that won't even make it in mass production.
Moore's law IS A DUD and only applies to a short interval in time, Moore apparently didn't look far enough into the future, and made a foolish assumption transistors will keep shrinking perpetually, and even considering how thick chips were back in his days is no excuse for him making such a statement, nor is there any excuse to people who like religious zealots believe it will hold forever.
If you bother to actually map transistor size vs time you will not get a straight line but the half of an inverted bell shaped curve, indicating initially slow transistor shrinking, gradually speeding up towards the middle and gradually slowing down to a complete halt the decade or two.
Something like that: www.futuretimeline.net/subject/images/transistor-size-timeline.gif
I wish I could look more than 40 years into the future...
But on a serious note, Moore's law still stands true, and will for the foreseeable future. You are right, we can't shrink things forever (we're already approaching the sub-atomic level) but there are creative ways around this; 3D transistors, quantum computing (currently vaporware IMHO) etc.
Well, even though I readily agree that Moore's Law will eventually come to an end, regarding the present technologies, I don't think this will happen until some years(may be 5? 10?).
I bought my i7 920 at the end of 2008. So ~ 6 years later or about 3 Moore's cycles I bought a laptop with an intel cpu that consumes officially 1/3 of the energy of the old beast. Taking into consideration the on-chip gpu, this is probably considerably less for the cpu part. Yet it performs, as a cpu, about 10% faster on single threaded and moderately threaded applications(3-4 threaded apps). So, for any practical purpose, it's 10% faster for less than one third of the energy requirements. If you do the multiplications and the divisions you will probably find that moore's law still holds true, not only from a manufacturing perspective but from a more substantial one;Flops per Watt.
The only problem is that performance per dollar went backwards for various reasons that are artificial and not based on the physical feasibility of things. Thus while you were purchasing a 130 watt cpu back in 2008 for ~300 dollars, now you cash out on average 500. But this is nothing. Think about it. Many people happily waste hundreds of dollars per year for the¨"latest mobile tech", ie they spend for chips that are from an order of magnitude to possibly 2 orders of magnitude less capable, and it seems normal to almost everyone. Now go support the mobile involution in order to pay for an ordinary cpu not 1000 but 10000 dollars and then be nostalgic for the good old days when you could buy a haswell extreme for 1000 bucks.
On one hand, AMD's stagnation is pissing me off because Intel hasn't had any competitive pressure to produce anything faster since Nehalem (Haswell is only about 20% faster in IPC, yet 5 years newer)
On the other hand, Intel would still be assfucking us with Netburst if it wasn't for AMD's Athlon.
Intel strongarmed AMD out of the game. The lost revenue destroyed their R&D budget and they haven't rebounded since.
My biggest issue with intel is the decision to cram IGP inside high end products, which makes about ZERO sense. Integrated graphics are OK for low and lower midrange products, but in a high end product it is just waste of die space. Intel wastes 2/5 of the die on graphics which ends up never being used, and for what? To get better "video card" market share, even at the price of selling stuff that ends up never being used? Throw in 50% more cores and cache please, or just make the chip smaller and cheaper...
While I agree with you from an engineering perspective, you are free to purchase XEON or -E class CPU's if you don't like 2/5 of your die wasted.
It does seem the Bloomfield Nehalem (Socket 1366) CPU's made the most efficient use of die for performance, being 45nm and all, and as soon as Intel integrated the graphics onchip it all the sudden lost its triple-channel memory controller (because there was no room left for it) so the die area can be better used, but in practice, the integrated graphics helps Intel's bottom line more than a triple channel controller or more cache, both of which 90% of consumers will benefit little from, while also increasing power consumption.
It is not just Intel's fault though. AMD's marketing and timing is notoriously bad and has been for a long time. And they still keep fucking it up. Remember Mantle? Any other company would have been cashing in on that performance advantage, bigtime, even if it was only a couple percent. The only things that help AMD sales are factors beyond their control such as the sales bump on their GPU's when Bitcoin mining became popular.
I doubt will see anything worthwhile from AMD let alone something that touches i7 until K12 ships. I can see them refreshing the FX chips next year using steamroller or a follow on core namely one that hopefully involves mostly 65W processors with 95w being the tdp
AMD isn't going to be able to get even with Intel(much less get ahead of them) until they move away form this two alu and 1 fpu per module garbage. The first gen Athlon has Intel running scared. All AMD can do now is compete on price..and they aren't doing that very well either. AMD is now facing what Intel was facing during the P-4 days. Hotter cpus that don't perform many things as well as the competition. AMD is a solid 2 generations behind in nearly all areas right now and with this current design they won't catch up.
searching on the net it seems the 8370 and the 8370e both OC the same. without the power limits it seems 4.7-5.0ghz for 1.45v is average for this cpu. so expect about a 20-25% OCability if your motherboard can supply enough power
if power/temps are not an issue one site even suggested the 8350 the better value buy. though from the data it seems at the same voltage and same MHz the 8370e is much lower in temps compared to the 8350. could just be good chip lotto.
my FM2 A10-7850 is twice as fast as my old Phenom x3 and it didn't cost much to upgrade even if it meant a whole new mobo. I'd no doubt go this route over an incremental upgrade using the old vishera design. My old Phenom x3 was AM2+ though, might want to double check your motherboard, if it IS just AM2+ you may have no choice! But have no fear, it is quite a bit snappier if you just go with the AMD APU. In fact if your uninterested in Mantle I'd just go with an old Trinity/ Richland design!
I know this is an old discussion but recent prices make this situation different. An Intel I5-4690k will almost always beat out an AMD 8370E, for example. But the AMD will be about 80% as powerful as the I5 at 60% of the price. If you need every last bit of processing power then go for the Intel CPU's, but for me the AMD is the clear winner for everyday tasks and gaming.
AMD doesn't even belong in the same sentence/paragraph/write-up/world/etc with that processor line. No, to answer your question. AMD is improving the power consumption of CPUs they already have out (3 years old now). Timing is just an coincidence with Intel's new Haswell-E launch.
Maybe.. but AMD gets a lot of criticism for it's power hungry processors so it's nice to see this get bumped down to 95watts. Intel's new line has 125watt cpu's which pretty much got a free pass in reviews. I found that odd.. especially with my 4790K coming in at 88watts.
I just bought and installed 4790K in my own Z87 desktop. The thing is that these 88 W in case of Intel are a typical, but not a very peak CPU wattage; e.g., my 4790K at 4.4 GHz on all the cores at 1.225 V consumes around 115 W in CPU-heavy tasks like Prime95 AVX 8 threads. So, take these 88 W "with a grain of salt".
"e.g., my 4790K at 4.4 GHz on all the cores at 1.225 V consumes around 115 W" So you are running the CPU out of spec but expect it to be in spec in other areas? If you OC, you increase the TDP! Don't OC and it will stay very much within its TDP limit.
The Turbo for all the cores loaded is officially 4.2 for 4790K. However, even at 4.2 it won't stay in its 88 W, unless you are lucky to get a very good CPU instance in order to undervolt it enough and still maintain stability - just look all the reviews of 4790K - it consumes quite a bit more power than 4770(K) under heavy CPU load, despite the formal 4 W (88 W vs 84 W) TDP difference. I haven't finished my tuning yet, but anyway, obviously, going from 4.2 to 4.4 at the same voltage won't increase TDP from 88 W to 115 W.
Thermal Design Power is the amount of power that the chip is supposed to generate, not power used. Also, how are measuring power consumption? There are a lot of factors involved and its difficult to isolate just the CPU.
They're Bulldozer done better, though not quite right, so you can perhaps forgive the "3 years old now" comment; still, I'd really like to see detailed power consumption figures especially when compared to the FX-8150, just to see how far they've actually come in that area.
Yes, if you add up initial Bulldozer FX-8150, OR-B2 ("Orochi B2") and the likes, then it happens to be 3 years, I agree, because these Piledriver FXes are just OR-C0 ("Orochi C0)".
This is perhaps the single most unimportant x86 cpu ever. This is just feeding a complex Mubadala GF agreement leading to stagnation and brand erosion.
The strong data we have is that is a fat, expensive, power hungry cpu, only fit for a minimal minority. And it shows in the financial results for AMD. Its burning money faster than electricity.
Let this old tech die as soon as possible.
At 1998 or so it cost me aprox 50 usd, and 10 minutes, to upgrade the 200MHz K6 to a 400MHz K6-2 using the same MB. Thats more like it. This is not remotely the same value. Burn baby burn.
It's not really a good comparison.. back then ( right up until the Athlon and later the Core2) You could get some pretty noticeable jumps. These days processor power has gone somewhat sideways /w features and efficiency rather than raw noticeable power. One of many reasons why people are holding onto their systems longer.. I still got lots of people on 5+ year old platforms running SSDs newer power supplies and new video cards but saying no.. their cpu is good enough.
For a CPU review, I actually think the baseline is somewhat distracting. Baseline at idle is largely dependent on motherboard, for the CPU review, the delta is the important data point.
So, and pardon me if I'm reading too much or too little into this, but AMD has released these 'new' processors in order to not find themselves in yet another Llano/FM1 situation where there's little uptake but big inventory of motherboards so as to avoid yet another massive write-off but this time on AM3+?
Isn't the alternative of not embarrassing themselves and their partners by stamping AM3+ with a long overdue 'EOL' branding the better option here? Who on earth is even buying these things?It seems that AMD has erroneously conflated and misinterpreted the outcry from enthusiasts for a much needed, competitive AMD, and the sentiment of "I'd like to buy an AMD processor, but..." with the harsh reality of their currently not-at-all competitive and completely unattractive product line. They've still got two separate chipsets on the board, and those chipsets are essentially the same chipsets we saw back in 2007!!!
Personally, I'd much rather see AMD make a comprehensive and honest effort to open up about their future products and roadmaps and answer some important questions than re-release a product that should never have been released in the first place. If they want to retain the ears of enthusiasts, it would be wise to open their ears to enthusiasts first.
"Personally, I'd much rather see AMD make a comprehensive and honest effort to open up about their future products and roadmaps"
AMD has been fairly clear on where they're going in their investor briefs, but it's probably not what you want to hear. They're executing a "transformation strategy" to get 50% of their revenue from the "Enterprise, Embedded and Semi-Custom segment" as opposed to the "Computing and Graphics segment" which is the traditional AMD/ATI business.
Where do you think this funding comes from? The CPU/APU/GPU research and development they're not doing, like a FX successor. And with that no new chipset, no new socket, no new motherboard validation. If you haven't noticed, they've been on a skeleton crew for some time now. All they're hyping is HSA which is semi-custom tech because normal apps have to serve the 80%+ not running AMD APUs.
For us, the mass market consumers of retail CPUs and GPUs I can give you a pretty accurate but grim forecast, AMD will continue to dodge direct competition. They can't compete with Intel on performance or power efficiency and with Intel so far ahead in processing technology they can't afford to undercut them on price either. Whatever their dual ARM/x86 processor ends up being, I'm pretty sure it's not primarily for us regular consumers.
Did you read the ansver from AMD about future FX prosessors. It seems that power efficiency is the way they are going next. Guite sensible because FX prosessor are fast enough all normal people, and they have no chance against Intel in the high end. They just not have enough money to compete properly. Maybe we see in the future a little bit faster Jaguar mini cores in FX line and normal Jaguar editions in mainstream APU machines. In that segment they have some possibilities to fight ARM and Intel processors.
This is a case of AMD kicking the can down the road. This release is meant to keep socket AM3+ limping along for a little while longer. Without a design refresh (Steamroller) and a new chipset, there hasn't been a reason to recommend the socket AM3+ platform for over year. Sure, it is nice that current owners have an upgrade path to faster parts. Considering the time it has taken to get these faster or more energy efficient parts to market one has to question if the wait has been worth it? Especially when compared to the evolving platforms on the Intel side.
I don't see any substantial improvement in the per/watt ratio for these new SKUs. The performance data still look dismal against Intel's low end chips. As much as I want AMD to compete on the intended space for these chips, but these Bulldozer-based cups are truly pathetic. I'm really fed with these CPUs coming out in reviews for year by now.
So the comments, as predicted, are full of AMD-bashing. But lets think about his a second: do you really expect a company with 1/10th the resources of Intel to even come close to match it in performance and efficiency? Not only that, but AMD has multiple product types to juggle it's resources on (CPU/GPU/APU) whereas Intel has only a single focus.
Not only that, but at least when buying AMD you don't get that dirty feeling you experience when you buy from a company like Intel, with it's lengthy history of anti-competitive business practices.
AMD is working on a new architecture, but it takes time. Patience, grasshopper!
Oh, are they? What TDP? What platform? Desktop/notebook/server? All 3? Socket? Target market segment? What about performance? Will it be better performing or equivalent but lower power? Neither? Does the x86 version outperform the ARM? Or is it vice versa? Is it going to be the same general architecture or are we talking two completely different architectures with different features? SoC or chipset? SoC + chipset? Is it BGA only or are we going to be able to buy socketed versions? Is AMD even going to release another chip for DIY'ers and enthusiasts post-Carrizo at all?
All of the those are unanswered. Every single one. Think about that for a second and digest it. You're asking for patience when there isn't even a general direction or roadmap. So patient for what exactly?
The comments above aren't AMD bashing, but rather product bashing. AMD hasn't been competitive for a long while now and rehashing the same wasn't-competitive-even-then product doesn't help matters any.
So don't mistake peoples' frustrations with AMD as AMD-hating or AMD-bashing, but AMD-disappointment. There's a huge portion of the population that wants to buy an AMD CPU/APU but has been waiting since the Bulldozer debacle to find something even remotely interesting -- and I'm most certainly in that camp.
No, no they haven't. Keller has stated that a new 'high performance architecture' is in the works, and that it will be an x86 variant but nothing outside of that. Bear in mind that AMD still considers Vishera a 'high performance architecture', so that statement has no meaning.
Stop hoping for AMD and buy/build a modern Intel desktop i5/i7 config, e.g. as I did last year. It will be around twice as fast as Deneb - I owned Deneb 940 (used at 3.4 GHz) in the past, so I'm comparing from my own experience.
I have both a Sandy Bridge and a Haswell laptop as my daily drivers. The desktop just sort of sits there soaking up files :P I've long since gotten over my AMD love affair. It's been a long time since the T-bred and Barton days.
twice as fast, yeah right go figure.... So now you just read this article twice as fast???
the issue with review sites is theoretical benchmarks, daily use you don't even notice the difference, HD choice and stupid slow Microsoft OS make the difference in daily tasks. In specific multitasking, sure but then again these AMD parts do actually work well and most of these tasks are done in the background anyhow while doing other stuff, so who cares it would take 1min longer.
So what is the issue? its review benchmark charts and e-penis behaviour.
Now people start complaining that Intel is deliberatly reducing renew cycles while its consumers own fault choosing the famous jingle and brand. No reason to buy a atom - celeron - pentium or i3 while APU series offer better overall added value.
I read your comments above. All you did was declare everyone who didn't fawn over the new chips to be a "troll", dodged a question someone asked then suggested that everyone who buys an Intel chip should "feel dirty" (which is trolling)...
These sorts of chips do rather well in multithreaded workloads. Most games really don't fit this category yet.
I'd get the 8320E over this any day though - there's a rather uncomfortable price difference despite what should be very similar performance. Still, an eight-core Steamroller clocked at even 8320 (125W edition) speeds would outperform pretty much all of AMD's FX line and use less power, but I imagine not by quite enough to warrant its release.
"These sorts of chips do rather well in multithreaded workloads. Most games really don't fit this category yet."
That's the problem word - "yet". AMD has spent the last 5 years endlessly waiting for "the day after tomorrow's games" instead of giving people what they want today : +50-70% faster cores (which runs all software faster multi-threaded or not) just to close the gap with Intel...
did you also care to look at multithreaded benchmarks where the 54W get trashed everywhere..
multithreading is still the future and you just put yourself back to 2009 and earlier, but hey you do run a few fps faster which you can't even notice within the game :).
Alright, AMD made it clear they don't think it's worth to upgrade their CPUs to their current architecture (they already have the cores) and at least 28 nm manufacturing (cheaper per transistor). And they could be right in that people just wouldn't buy enough of these to make up for the development costs.
So they try to make the best use of what they have. Yet.. do they? I could imagine them doing something much better with 4-module Piledriver dies: put 2 of them onto the same package, like they do for Opterons! Sell them at decent prices and anyone wanting cheap multithread number crunchers might be interested. They might hurt 1 and 2 socket Opteron sales a bit.. but without any updates, how competitive are they anyway? And you can't put much RAM into desktop boards anyway, so this wouldn't be that much of a thread.
Two dies won't fit in AM3+; so what you suggest is rebranding the socket G34 for a desktop and selling it cheaper, I suppose. Still a questionable idea IMHO to me.
Purposeful apples to oranges comparison on that chart. If they put the same GPU with the Intel CPU, the Intel CPU price drops $50 and is flat out in the middle between the 8370e and 8320e. What a joke. I know they're trying to drum up hype for their product, but FFS at least do an even comparison when possible.
And for the comment about how AMD is 1/10th the size of Intel, give me a break. AMD's CPU division is floundering, but they've been flat out abusing NVIDIA on price/performance for the last several years now - another company that is probably larger than AMD. The inferred excuse that "they can't compete because they're smaller" is a joke.
I love my R9's but I happily put them in an i7 setup because AMD CPU's are still not up to snuff and are still too power hungry (by comparison).
Multi threaded performance is decent.. single is meh.. price $200 putting it in i5 category.. it needs to be sitting around $150 to be competitive... than it becomes a interesting buy. People worried about the numbers I am on a 4790K and also have systems based around the new Pentium and Amd's A10 and Vishera 6300.. I am on and off those systems quite extensively and you know what?
I don't go why so slow.. omg ... similarly configured their all pretty fast.
You want to know the really nice thing about AMD's AM3+ I can replace aging motherboards that have died and still keep the cpu. It's a good platform overall just people want a real update I guess. I build/upgrade 20+ systems a year and do use the AM3+ platform when the right deals come along. I am fine with the update... although I do think the price on that cpu should be around $160 to make it viable.
I'd really like to see an article (blog whatever..) about a baseline system. What you feel is still viable for todays computing needs. Occasionally I still have to do work on X2's and P4's and have come to the conclusion that they should have been retired long ago.. but Phenom 2 setups and Core2's (8x series not 6..) still seem to be trucking along perfectly fine with new hardware surrounding them (SSD's video etc.)
Basically something you could refresh once a year or so.. you know? be real cool to see that and since it would be going thru the battery of tests put thru on new setups it can be included in new cpu reviews as well as part of the comparisons.
So AMD's 6 core CPU is MARGINALLY faster than Intel's quad core i5 on a test written specifically to maximize the advantage of many threads and HALF as fast in the single threaded test? Come on...
You're better off with a Sandy Bridge chip from 3-4 years ago than you are with a brand new AMD CPU. This is sad indeed.
I feel like Intel, at this point, might have the next breakthrough, like Conroe or Sandy Bridge, but they have no reason to release it because they've essentially stood still for 4 years and AMD still can't do more than "achieve" HALF the performance of an Intel counterpart.
The fact that Amazon had the FX-8350 on sale several weeks ago for a mere $159 and that I was able to throw that CPU into an Asus Crosshair IV Formula from 2010 (while selling my X6 1090T for $175 on eBay), and overclock that baby to 4.7GHz, and OUTPERFORM a $350 Intel i7-3770K AND an i7-4770K in Passmark, scoring 10,700 points proves that AMD's chips are extremely competitive. Made mincemeat of Intel's lineup. LOL
so you're comparing an overclocked AMD chip to a stock Intel chip. Doesn't seem fair.
with the 6 cores out for just $60 more than a quad core, that price to performance ratio rises dramatically.
a 4790k+mobo is cheaper and faster in the long run than an FX 8 core. The power costs alone will make up the price difference over a few years. Then there's the undeniable single threaded performance, which means 4 threaded applications get 100% of the cpu, while in AMD lesser than 8 threads means all that horsepower is under utilized.
You just don't realize the performance you're missing.
And don't get me started on things like min fps in gaming, which AMD can't match even with all cores in use.
It's a good upgrade deal for sure; kind of dumb to tout the "thrashing" when I can simply go into the bios on my Asus Maximus Formula and simply click on a single button to OC to at least 4.2.
I'll run Passmark OC'd and see what I get on my 3770k. I bet it's not taking such a bad thrashing then.
Absolutely stock settings with nothing turned off win 8.1 x64. Clicked Auto OC to 4.6; So 10k is really good, what about other apps ? Why does it do so badly in gaming then ?
Not my fault your a braindead consumer willing to pay more than twice as much for a CPU that ultimately delivers the same performance. I have to hand it to you though, you are 100% correct, Intel IS STANDING STILL when their 2014 processors match that of AMD's 2012 offerings. http://imgur.com/J6RoBXO
it is interesting you are comparing your OC'd 8350(4.7g) vs stock 3770k(3.5g) and a stock 4770k(3.5g). since if you oc'd the i7s they would make "mincemeat" of your scores too. now if your comparing pure performance per dollar then yes your amd chip is better ratio on that one particular test than intel's offerings. BUT an intel system doesn't produce as much heat as an AMD system. for the same noise level, almost silent, I can run a 4770k(4.5g) @ 1.4v with the stock cooler. Stable enough to run all my games and stay <80C. how loud must the 8350(4.7g) be to to get a score of 11k on passmark?
i'm not a intel fan boy, just saying passmark isn't the only facet of a cpu that you should look for unless benching passmark is all you do. because then anyone can say a $70(not on sale) G3258(4.7g) single threaded cinebench scoring 173 makes "mincemeat" of a $159(on sale) 8350(stock) scoring 88
Except for those who don't know the significance of this combination. Many 760G motherboards will support ECC RAM, and all the current AMD processors support full virtualization. Using this for a cheap, high thread count home server, where I would have to spend several hundred dollars more to get similar server performance from an Intel product.
Interesting option for home server/virt labbing. 8 cores for much cheaper than the intel equivalent. I do hope its ECC and VT-D compatible on the mobos.
A M ... what? I think this will be the reaction of the next gen of gamers to the once legendary AMD name. What has gotten into them ? I am an Intel and Nvidia fanboy but the problem is, the lack of competition from AMD (with their much hyped but pseudo multi core processors) to Intel has made sure Intel can trump them with just dual core fixed clock CPUs. I think Intel would have been forced to bring unlocked quad cores to the lowest end processors with proper competition.
One thing that has become obvious during the past 14 years or so is that EVERY TIME a new benchmark is introduced, AMD processors are ALWAYS competitive, just like the new x265 4K benchmark. Even the APUs are competitive.
Then Intel updates its free compiler too and while its CPUs maintain their performance, the AMD chips are suddenly less capable in the new version of the program.
Intel is free to do whatever it want with its compiler and the software is actually good and useful, BUT the developers using Intel's compiler should respect AMD users and patch their software accordingly.
It is publicly known that if you use Intel's compiler on your software, it will identify the CPU and if it's an Intel one, it will run the most efficient routine, but if it's an AMD chip, it will run the "most compatible" routine that's considerably slower.
If the developer respects his customers and wants to offer a correctly and completely optimized piece of software, he should patch his product to run the most efficient routine on AMD CPUs too.
The reason AMD is making this part has less to do with data showing users who want to upgrade and more to do with users who are looking for new upgrades. They go and they can either buy CPU's that show as released over two years ago or they can buy new ones with new names that show newer releases dates. Even if the reviews state these CPU's are essentially the same as the older parts, there is a large group of users who won't care so long as the CPU's are "new."
This is the same reason you saw the R9 280/280X, etc, get their flashy new name last year in spite of the cards being the exact same design as the older one.
GUYS ENOUGH WITH SUPERBENCHMARKING AND SUPER INTEL TECHNOLOGY. THIS TIME WE DONT NEED THE EXTRA PERFORMANCE. INTEL DID THE SAME THING -PUMB THE SPEED AND VOUALA YOUR I7 4790K. THEY HAVE PROBLEMS WITH 14NM PROCESS. SO WHEN THE SANDY BECOMES THE REAL WALL POINT AND THE 22NM DIDNOT OFFER A PUMP OF 35% REAL PUMB ON PERFORMANCE IS LOGICAL TO SEE TINY AMD STAYS AT 32NM WITH LOW PRICES. AS FOR ENERGY EFFICIENCE ARE JUST BUBU 40$ OR 50$ LESS WITH INTEL IN 5 YEARS. ITS SAD THAT SANDY AND PILLDRIVER IS STILL HERE IN PERFORMANCE WAY. THE OLD DAYS OF 50% PUMB FROM GENERATION TO GENERATION ARE OVER. AS FOR INTEL WILL GIVE YOU 6 CORES WITH 12 THREADS I7 5790K BROADWELL AND I5 4 CORES WITH 8 THREADS I7 5820K SHOWED THAT< IF THE 14NM WILL SCALE GOOD ENOUGH IN EARLY 2015 OR SUMMER. SO AMD WITH 14NM OR 16NM WILL STRIKE WITH NEW DESIGNED 8C0RES WITH 16 THREADS IN 2015 Q4 OR 2016 Q1 IF THIS WILL BE FAILDOZER OR BRUTAL ATHLON WE WILL SEE THAT. SORRY FOR MY BAD ENGLISH.
So... That's for people that are still using Phenon ll's /Athlon ll's (maybe even low end fx's) Those low TDP chips are for people that have 95w max power boards. Simple as that. Yes, they do quite fine but i would NOT recommend them to any new builders. They are 'viable' upgrade options for AMD users. That said, USD 200.00 on that chip? No, and no. Buy an I5, period. If you have a AM3 board and you fit in that description (Phenonll/Athlonll)... go ahead, better performance with less power/heat.
I wonder how many times I have to hear the FX series isn't as good as the i5 or i7. I think I get it, most didn't plan on buying just wanted to sound cool! News flash....you don't.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
107 Comments
Back to Article
boot318 - Tuesday, September 2, 2014 - link
AMD, you sure the CPU speed is right? :/Good to see that power consumption is down (would like to get more data on it). I've been wanting to my my Phenom x3 that I have an office computer with an 8-core, and I might just do it with this one. Black Friday!!!!
ddriver - Tuesday, September 2, 2014 - link
Yeah, they fixed power consumption, all they need to do now is double the performance and they are back in the game LOLFlunk - Tuesday, September 2, 2014 - link
Hey, be nice to AMD. These new processors are totally competitive with Intel...'s Sandy Bridge processors from 3 years ago.Seriously AMD, Intel has almost been standing still for 3 years and you're still behind. Come up with something better or us enthusiasts are not going to have a reason to even look at your chips. Your GPUs are a good option, with competitive performance and better prices (I recommend Radeons in almost every price bracket right now because of this). If only AMD could put out some decent CPUs.
SpaceRanger - Tuesday, September 2, 2014 - link
Has anyone been looking at their chips seriously?? I stopped using AMD a long while ago. Intel blew them away and it'll take some competitive chips from AMD to get me to think otherwise.ddriver - Tuesday, September 2, 2014 - link
Don't get me wrong, I sympathize with underdogs, and especially considering in the case of AMD, it was decades of Intel's anti-competitive practices that have pretty much doomed AMD to be a perpetual runner up, but I can't deny the obvious. AMD is where Intel always wanted it to be - a handicapped "competitor", barely enough for Intel to say "we are not a monopoly". I've been a long time AMD consumer, but today it is not worth it, maybe on the low end where power consumption is not that much of an issue and the lower price may make some sense, at least in the short term. But ever since I became a prosumer, AMD CPUs simply have nothing to offer. Radeons are pretty neat though, since they offer tremendously better DP performance for the buck.Seriously, how hard is to make a decent CPU... Intel surely has a manufacturing lead as well, but the poor performance of AMD is mostly because of the inferior design, process might give Intel 20-30% advantage, but the rest is all the design.
On a side note, the CPU performance stalling is actually a good thing, even if AMD was competitive, there are physical limits which CPUs will hit soon, not to mention Intel has already reached the point the chips are too small to displace that much hear efficiently. I remember I used to upgrade CPUs every 18 months or so, but now I have a 2 year old CPU that is only a notch slower than the latest and greatest.
Moore's "law" is a dud, it is obvious that performance cannot exponentially grow, it is more of a half of bell shaped curve, as we reach physical limits, performance gains will slowly go stale. And quite frankly - that's OK, enough is enough. Did we really make much good use of extra performance? Nope, we got sloppy technologies, slow programming languages, bloated runtimes which kind of "equal out" the experience. I have 10 tabs in my browser open, with total content of all websites not exceeding 100 MB, and yet it takes 1200 MB of ram... Lousy skype, a puny messenger - 100 MB of ram usage? WTF?
Death666Angel - Tuesday, September 2, 2014 - link
"not to mention Intel has already reached the point the chips are too small to displace that much hear efficiently."So why is my i7-4770k @ 4.5GHz and 1.35V cooler (60°C core average while running Prime95 27.9, delidded) than my old i7-860 @ 3.8GHz and 1.34V (80°C core average with same load)? It is pretty much just Intel cheaping out on manufacturing of the consumer line CPUs, with weak TIM and too much glue for the IHS.
Death666Angel - Tuesday, September 2, 2014 - link
"Moore's "law" is a dud, it is obvious that performance"Way to show that you don't know jack. Moore's Law just says that the number of transistors on the same area will double every 2 years. Nothing about the performance of said transistors.
mrdude - Tuesday, September 2, 2014 - link
You forgot the economics portion of Moore's law, which ties number of transistors per dollar. If there wasn't a financial incentive, we would have never had the law in the first place ;) Shrinking didn't just improve die sizes (and almost always performance), but it also made economic sense. If you take into account the economic aspect, ddriver isn't far off. We're not far off heretofore hypothetical scenario of 'It's just too expensive and doesn't provide enough benefit anymore'. In fact, for some semicos we're already there.Samus - Wednesday, September 3, 2014 - link
Seriously, to imply that an IPC should double in performance every 24 months is ridiculously impossible.Moore's law is a law of production, not design.
ddriver - Wednesday, September 3, 2014 - link
Even so, that is impossible just as well. This would require that transistors can get infinitely small, which is simply not the case. There are limits to how small a transistor can be, and even in the case of insanely slow and expensive atomic assembly transistors will still be several atoms big, and that won't even make it in mass production.Moore's law IS A DUD and only applies to a short interval in time, Moore apparently didn't look far enough into the future, and made a foolish assumption transistors will keep shrinking perpetually, and even considering how thick chips were back in his days is no excuse for him making such a statement, nor is there any excuse to people who like religious zealots believe it will hold forever.
If you bother to actually map transistor size vs time you will not get a straight line but the half of an inverted bell shaped curve, indicating initially slow transistor shrinking, gradually speeding up towards the middle and gradually slowing down to a complete halt the decade or two.
Something like that: www.futuretimeline.net/subject/images/transistor-size-timeline.gif
however this one starts at the middle.
fluxtatic - Wednesday, September 3, 2014 - link
Yeah, that 40 years it held true was a total goof. That Moore, what an idiot! /sSamus - Wednesday, September 3, 2014 - link
I wish I could look more than 40 years into the future...But on a serious note, Moore's law still stands true, and will for the foreseeable future. You are right, we can't shrink things forever (we're already approaching the sub-atomic level) but there are creative ways around this; 3D transistors, quantum computing (currently vaporware IMHO) etc.
ironargonaut - Wednesday, September 3, 2014 - link
You are the fool Moore made no such assumption. Nor did he make any such statement. Others took what he said and condensed it into "Moore's law".Gigaplex - Tuesday, September 2, 2014 - link
"Seriously, how hard is to make a decent CPU"Let's see you do a better job if you think it's easy.
imin83 - Wednesday, September 3, 2014 - link
"Seriously, how hard is to make a decent CPU"-the correct reply to that statement would be, "Watch out, we got a badass over here"
IUU - Friday, September 5, 2014 - link
Well, even though I readily agree that Moore's Law will eventually come to an end, regarding the present technologies, I don't think this will happen until some years(may be 5? 10?).I bought my i7 920 at the end of 2008. So ~ 6 years later or about 3 Moore's cycles I bought a laptop
with an intel cpu that consumes officially 1/3 of the energy of the old beast. Taking into consideration the on-chip gpu, this is probably considerably less for the cpu part. Yet it performs, as a cpu,
about 10% faster on single threaded and moderately threaded applications(3-4 threaded apps). So, for any practical purpose, it's 10% faster for less than one third of the energy requirements.
If you do the multiplications and the divisions you will probably find that moore's law still holds true, not only from a manufacturing perspective but from a more substantial one;Flops per Watt.
The only problem is that performance per dollar went backwards for various reasons that are artificial and not based on the physical feasibility of things. Thus while you were purchasing a 130 watt cpu back in 2008 for ~300 dollars, now you cash out on average 500. But this is nothing.
Think about it. Many people happily waste hundreds of dollars per year for the¨"latest mobile tech", ie they spend for chips that are from an order of magnitude to possibly 2 orders of magnitude less capable, and it seems normal to almost everyone. Now go support the mobile involution in order to pay for an ordinary cpu not 1000 but 10000 dollars and then be nostalgic for the good old days when you could buy a haswell extreme for 1000 bucks.
Samus - Wednesday, September 3, 2014 - link
On one hand, AMD's stagnation is pissing me off because Intel hasn't had any competitive pressure to produce anything faster since Nehalem (Haswell is only about 20% faster in IPC, yet 5 years newer)On the other hand, Intel would still be assfucking us with Netburst if it wasn't for AMD's Athlon.
Intel strongarmed AMD out of the game. The lost revenue destroyed their R&D budget and they haven't rebounded since.
ddriver - Wednesday, September 3, 2014 - link
My biggest issue with intel is the decision to cram IGP inside high end products, which makes about ZERO sense. Integrated graphics are OK for low and lower midrange products, but in a high end product it is just waste of die space. Intel wastes 2/5 of the die on graphics which ends up never being used, and for what? To get better "video card" market share, even at the price of selling stuff that ends up never being used? Throw in 50% more cores and cache please, or just make the chip smaller and cheaper...Samus - Wednesday, September 3, 2014 - link
While I agree with you from an engineering perspective, you are free to purchase XEON or -E class CPU's if you don't like 2/5 of your die wasted.It does seem the Bloomfield Nehalem (Socket 1366) CPU's made the most efficient use of die for performance, being 45nm and all, and as soon as Intel integrated the graphics onchip it all the sudden lost its triple-channel memory controller (because there was no room left for it) so the die area can be better used, but in practice, the integrated graphics helps Intel's bottom line more than a triple channel controller or more cache, both of which 90% of consumers will benefit little from, while also increasing power consumption.
silverblue - Wednesday, September 3, 2014 - link
QuickSync could be useful...?Vayra - Saturday, September 13, 2014 - link
It is not just Intel's fault though. AMD's marketing and timing is notoriously bad and has been for a long time. And they still keep fucking it up. Remember Mantle? Any other company would have been cashing in on that performance advantage, bigtime, even if it was only a couple percent. The only things that help AMD sales are factors beyond their control such as the sales bump on their GPU's when Bitcoin mining became popular.SunLord - Wednesday, September 3, 2014 - link
I doubt will see anything worthwhile from AMD let alone something that touches i7 until K12 ships. I can see them refreshing the FX chips next year using steamroller or a follow on core namely one that hopefully involves mostly 65W processors with 95w being the tdphescominsoon - Friday, October 17, 2014 - link
AMD isn't going to be able to get even with Intel(much less get ahead of them) until they move away form this two alu and 1 fpu per module garbage. The first gen Athlon has Intel running scared. All AMD can do now is compete on price..and they aren't doing that very well either. AMD is now facing what Intel was facing during the P-4 days. Hotter cpus that don't perform many things as well as the competition. AMD is a solid 2 generations behind in nearly all areas right now and with this current design they won't catch up.bebimbap - Tuesday, September 2, 2014 - link
searching on the net it seems the 8370 and the 8370e both OC the same. without the power limits it seems 4.7-5.0ghz for 1.45v is average for this cpu. so expect about a 20-25% OCability if your motherboard can supply enough powerif power/temps are not an issue one site even suggested the 8350 the better value buy.
though from the data it seems at the same voltage and same MHz the 8370e is much lower in temps compared to the 8350. could just be good chip lotto.
Sigh, i miss the old thoroughbred days.
Knowname - Monday, September 15, 2014 - link
my FM2 A10-7850 is twice as fast as my old Phenom x3 and it didn't cost much to upgrade even if it meant a whole new mobo. I'd no doubt go this route over an incremental upgrade using the old vishera design. My old Phenom x3 was AM2+ though, might want to double check your motherboard, if it IS just AM2+ you may have no choice! But have no fear, it is quite a bit snappier if you just go with the AMD APU. In fact if your uninterested in Mantle I'd just go with an old Trinity/ Richland design!Scionero - Saturday, December 27, 2014 - link
I know this is an old discussion but recent prices make this situation different. An Intel I5-4690k will almost always beat out an AMD 8370E, for example. But the AMD will be about 80% as powerful as the I5 at 60% of the price. If you need every last bit of processing power then go for the Intel CPU's, but for me the AMD is the clear winner for everyday tasks and gaming.blackmagnum - Tuesday, September 2, 2014 - link
Does this AMD chip have anything to do with this?: August 29th, The Haswell-E Launch.boot318 - Tuesday, September 2, 2014 - link
AMD doesn't even belong in the same sentence/paragraph/write-up/world/etc with that processor line. No, to answer your question. AMD is improving the power consumption of CPUs they already have out (3 years old now). Timing is just an coincidence with Intel's new Haswell-E launch.just4U - Tuesday, September 2, 2014 - link
Maybe.. but AMD gets a lot of criticism for it's power hungry processors so it's nice to see this get bumped down to 95watts. Intel's new line has 125watt cpu's which pretty much got a free pass in reviews. I found that odd.. especially with my 4790K coming in at 88watts.TiGr1982 - Tuesday, September 2, 2014 - link
I just bought and installed 4790K in my own Z87 desktop. The thing is that these 88 W in case of Intel are a typical, but not a very peak CPU wattage; e.g., my 4790K at 4.4 GHz on all the cores at 1.225 V consumes around 115 W in CPU-heavy tasks like Prime95 AVX 8 threads.So, take these 88 W "with a grain of salt".
just4U - Tuesday, September 2, 2014 - link
I noticed with cpu-z that power is all over the place. so yeah.. but still..Death666Angel - Tuesday, September 2, 2014 - link
"e.g., my 4790K at 4.4 GHz on all the cores at 1.225 V consumes around 115 W"So you are running the CPU out of spec but expect it to be in spec in other areas? If you OC, you increase the TDP! Don't OC and it will stay very much within its TDP limit.
just4U - Tuesday, September 2, 2014 - link
4.4GHZ is turbo boost on the 4790K. It runs at that speed out of the box.
Akrovah - Tuesday, September 2, 2014 - link
Only on a single core. He specified 4.4 Ghz on all cores. hence the assumption of overclocking.TiGr1982 - Tuesday, September 2, 2014 - link
The Turbo for all the cores loaded is officially 4.2 for 4790K. However, even at 4.2 it won't stay in its 88 W, unless you are lucky to get a very good CPU instance in order to undervolt it enough and still maintain stability - just look all the reviews of 4790K - it consumes quite a bit more power than 4770(K) under heavy CPU load, despite the formal 4 W (88 W vs 84 W) TDP difference. I haven't finished my tuning yet, but anyway, obviously, going from 4.2 to 4.4 at the same voltage won't increase TDP from 88 W to 115 W.Flunk - Tuesday, September 2, 2014 - link
Thermal Design Power is the amount of power that the chip is supposed to generate, not power used. Also, how are measuring power consumption? There are a lot of factors involved and its difficult to isolate just the CPU.techguyz - Thursday, September 4, 2014 - link
And do you think a 95w AMD CPU uses just 95w? Nope.TiGr1982 - Tuesday, September 2, 2014 - link
No, these Piledriver-based FXes are just 2 years old, not 3 :)These were first released back in October 2012.
silverblue - Tuesday, September 2, 2014 - link
They're Bulldozer done better, though not quite right, so you can perhaps forgive the "3 years old now" comment; still, I'd really like to see detailed power consumption figures especially when compared to the FX-8150, just to see how far they've actually come in that area.TiGr1982 - Tuesday, September 2, 2014 - link
Yes, if you add up initial Bulldozer FX-8150, OR-B2 ("Orochi B2") and the likes, then it happens to be 3 years, I agree, because these Piledriver FXes are just OR-C0 ("Orochi C0)".silverblue - Tuesday, September 2, 2014 - link
I remember a B3 stepping that never surfaced; perhaps the process improvements were baked into C0 along with the architectural tweaks.TiGr1982 - Tuesday, September 2, 2014 - link
That is what actually happened. The sad thing that since then AMD never really did something really new about FXes - not a single stepping, AFAIK.krumme - Tuesday, September 2, 2014 - link
"Strong data"?This is perhaps the single most unimportant x86 cpu ever. This is just feeding a complex Mubadala GF agreement leading to stagnation and brand erosion.
The strong data we have is that is a fat, expensive, power hungry cpu, only fit for a minimal minority. And it shows in the financial results for AMD. Its burning money faster than electricity.
Let this old tech die as soon as possible.
At 1998 or so it cost me aprox 50 usd, and 10 minutes, to upgrade the 200MHz K6 to a 400MHz K6-2 using the same MB. Thats more like it. This is not remotely the same value. Burn baby burn.
just4U - Tuesday, September 2, 2014 - link
It's not really a good comparison.. back then ( right up until the Athlon and later the Core2) You could get some pretty noticeable jumps. These days processor power has gone somewhat sideways /w features and efficiency rather than raw noticeable power. One of many reasons why people are holding onto their systems longer.. I still got lots of people on 5+ year old platforms running SSDs newer power supplies and new video cards but saying no.. their cpu is good enough.zebrax2 - Tuesday, September 2, 2014 - link
Can we get some idle power consumption numbers? The delta isn't really useful without a baselineEssence_of_War - Tuesday, September 2, 2014 - link
For a CPU review, I actually think the baseline is somewhat distracting. Baseline at idle is largely dependent on motherboard, for the CPU review, the delta is the important data point.mrdude - Tuesday, September 2, 2014 - link
So, and pardon me if I'm reading too much or too little into this, but AMD has released these 'new' processors in order to not find themselves in yet another Llano/FM1 situation where there's little uptake but big inventory of motherboards so as to avoid yet another massive write-off but this time on AM3+?Isn't the alternative of not embarrassing themselves and their partners by stamping AM3+ with a long overdue 'EOL' branding the better option here? Who on earth is even buying these things?It seems that AMD has erroneously conflated and misinterpreted the outcry from enthusiasts for a much needed, competitive AMD, and the sentiment of "I'd like to buy an AMD processor, but..." with the harsh reality of their currently not-at-all competitive and completely unattractive product line. They've still got two separate chipsets on the board, and those chipsets are essentially the same chipsets we saw back in 2007!!!
Personally, I'd much rather see AMD make a comprehensive and honest effort to open up about their future products and roadmaps and answer some important questions than re-release a product that should never have been released in the first place. If they want to retain the ears of enthusiasts, it would be wise to open their ears to enthusiasts first.
jann5s - Tuesday, September 2, 2014 - link
+1asimov1979 - Tuesday, September 2, 2014 - link
I totally agreeasimov1979 - Tuesday, September 2, 2014 - link
These would have been a good option back in 2012Kjella - Tuesday, September 2, 2014 - link
"Personally, I'd much rather see AMD make a comprehensive and honest effort to open up about their future products and roadmaps"AMD has been fairly clear on where they're going in their investor briefs, but it's probably not what you want to hear. They're executing a "transformation strategy" to get 50% of their revenue from the "Enterprise, Embedded and Semi-Custom segment" as opposed to the "Computing and Graphics segment" which is the traditional AMD/ATI business.
Where do you think this funding comes from? The CPU/APU/GPU research and development they're not doing, like a FX successor. And with that no new chipset, no new socket, no new motherboard validation. If you haven't noticed, they've been on a skeleton crew for some time now. All they're hyping is HSA which is semi-custom tech because normal apps have to serve the 80%+ not running AMD APUs.
For us, the mass market consumers of retail CPUs and GPUs I can give you a pretty accurate but grim forecast, AMD will continue to dodge direct competition. They can't compete with Intel on performance or power efficiency and with Intel so far ahead in processing technology they can't afford to undercut them on price either. Whatever their dual ARM/x86 processor ends up being, I'm pretty sure it's not primarily for us regular consumers.
haukionkannel - Tuesday, September 2, 2014 - link
Did you read the ansver from AMD about future FX prosessors. It seems that power efficiency is the way they are going next. Guite sensible because FX prosessor are fast enough all normal people, and they have no chance against Intel in the high end. They just not have enough money to compete properly.Maybe we see in the future a little bit faster Jaguar mini cores in FX line and normal Jaguar editions in mainstream APU machines. In that segment they have some possibilities to fight ARM and Intel processors.
Spoony - Tuesday, September 2, 2014 - link
I so much want to be able to shop between Intel and AMD when I purchase a CPU. Make it happen for me, AMD!zmeul - Tuesday, September 2, 2014 - link
when your FX CPU performs worse than an i3 CPU (in quite a lot of tests), it's time to change or get out of the businessKevin G - Tuesday, September 2, 2014 - link
This is a case of AMD kicking the can down the road. This release is meant to keep socket AM3+ limping along for a little while longer. Without a design refresh (Steamroller) and a new chipset, there hasn't been a reason to recommend the socket AM3+ platform for over year. Sure, it is nice that current owners have an upgrade path to faster parts. Considering the time it has taken to get these faster or more energy efficient parts to market one has to question if the wait has been worth it? Especially when compared to the evolving platforms on the Intel side.landerf - Tuesday, September 2, 2014 - link
Jesus those charts were brutaltexasti89 - Tuesday, September 2, 2014 - link
I don't see any substantial improvement in the per/watt ratio for these new SKUs. The performance data still look dismal against Intel's low end chips. As much as I want AMD to compete on the intended space for these chips, but these Bulldozer-based cups are truly pathetic. I'm really fed with these CPUs coming out in reviews for year by now.Germanicus - Tuesday, September 2, 2014 - link
So the comments, as predicted, are full of AMD-bashing. But lets think about his a second: do you really expect a company with 1/10th the resources of Intel to even come close to match it in performance and efficiency? Not only that, but AMD has multiple product types to juggle it's resources on (CPU/GPU/APU) whereas Intel has only a single focus.Not only that, but at least when buying AMD you don't get that dirty feeling you experience when you buy from a company like Intel, with it's lengthy history of anti-competitive business practices.
AMD is working on a new architecture, but it takes time. Patience, grasshopper!
mrdude - Tuesday, September 2, 2014 - link
Oh, are they? What TDP? What platform? Desktop/notebook/server? All 3? Socket? Target market segment? What about performance? Will it be better performing or equivalent but lower power? Neither? Does the x86 version outperform the ARM? Or is it vice versa? Is it going to be the same general architecture or are we talking two completely different architectures with different features? SoC or chipset? SoC + chipset? Is it BGA only or are we going to be able to buy socketed versions? Is AMD even going to release another chip for DIY'ers and enthusiasts post-Carrizo at all?All of the those are unanswered. Every single one. Think about that for a second and digest it. You're asking for patience when there isn't even a general direction or roadmap. So patient for what exactly?
The comments above aren't AMD bashing, but rather product bashing. AMD hasn't been competitive for a long while now and rehashing the same wasn't-competitive-even-then product doesn't help matters any.
So don't mistake peoples' frustrations with AMD as AMD-hating or AMD-bashing, but AMD-disappointment. There's a huge portion of the population that wants to buy an AMD CPU/APU but has been waiting since the Bulldozer debacle to find something even remotely interesting -- and I'm most certainly in that camp.
Germanicus - Tuesday, September 2, 2014 - link
They've already answered some of your questions, but I'm not going to Google for you.mrdude - Tuesday, September 2, 2014 - link
No, no they haven't. Keller has stated that a new 'high performance architecture' is in the works, and that it will be an x86 variant but nothing outside of that. Bear in mind that AMD still considers Vishera a 'high performance architecture', so that statement has no meaning.Germanicus - Tuesday, September 2, 2014 - link
There is no question your bashing and trolling.mrdude - Tuesday, September 2, 2014 - link
Yep. Me with my 5870 and 955 Deneb that I bought several years ago, still waiting, bashing and trolling for an upgrade.Here's the link to Jim Keller's interview:
https://www.youtube.com/watch?v=SOTFE7sJY-Q
Germanicus - Tuesday, September 2, 2014 - link
Watched it already.TiGr1982 - Tuesday, September 2, 2014 - link
Stop hoping for AMD and buy/build a modern Intel desktop i5/i7 config, e.g. as I did last year.It will be around twice as fast as Deneb - I owned Deneb 940 (used at 3.4 GHz) in the past, so I'm comparing from my own experience.
mrdude - Tuesday, September 2, 2014 - link
I have both a Sandy Bridge and a Haswell laptop as my daily drivers. The desktop just sort of sits there soaking up files :P I've long since gotten over my AMD love affair. It's been a long time since the T-bred and Barton days.duploxxx - Wednesday, September 3, 2014 - link
twice as fast, yeah right go figure.... So now you just read this article twice as fast???the issue with review sites is theoretical benchmarks, daily use you don't even notice the difference, HD choice and stupid slow Microsoft OS make the difference in daily tasks. In specific multitasking, sure but then again these AMD parts do actually work well and most of these tasks are done in the background anyhow while doing other stuff, so who cares it would take 1min longer.
So what is the issue? its review benchmark charts and e-penis behaviour.
Now people start complaining that Intel is deliberatly reducing renew cycles while its consumers own fault choosing the famous jingle and brand. No reason to buy a atom - celeron - pentium or i3 while APU series offer better overall added value.
bsim500 - Tuesday, September 2, 2014 - link
So the "95w" 8-core chip actually runs at 3.3GHz and gets thrashed in 100% of tested games by even a 2-core 54w i3-4360?......And the faster "125w" 4GHz version pulls an eye-watering (and consistent across multiple tests) +233w under load?
2009 just called and want their CPU's back...
Germanicus - Tuesday, September 2, 2014 - link
See my comment above.bsim500 - Tuesday, September 2, 2014 - link
I read your comments above. All you did was declare everyone who didn't fawn over the new chips to be a "troll", dodged a question someone asked then suggested that everyone who buys an Intel chip should "feel dirty" (which is trolling)...Germanicus - Tuesday, September 2, 2014 - link
Everyone has different ethics. Not trolling.Most everyone here is parroting the same points. This isn't adding anything to the discussion. It is a bash-fest.
Germanicus - Tuesday, September 2, 2014 - link
Fawn? Whatever. Never implied that. I wish they were more competitive.silverblue - Tuesday, September 2, 2014 - link
These sorts of chips do rather well in multithreaded workloads. Most games really don't fit this category yet.I'd get the 8320E over this any day though - there's a rather uncomfortable price difference despite what should be very similar performance. Still, an eight-core Steamroller clocked at even 8320 (125W edition) speeds would outperform pretty much all of AMD's FX line and use less power, but I imagine not by quite enough to warrant its release.
bsim500 - Tuesday, September 2, 2014 - link
"These sorts of chips do rather well in multithreaded workloads. Most games really don't fit this category yet."That's the problem word - "yet". AMD has spent the last 5 years endlessly waiting for "the day after tomorrow's games" instead of giving people what they want today : +50-70% faster cores (which runs all software faster multi-threaded or not) just to close the gap with Intel...
duploxxx - Wednesday, September 3, 2014 - link
did you also care to look at multithreaded benchmarks where the 54W get trashed everywhere..multithreading is still the future and you just put yourself back to 2009 and earlier, but hey you do run a few fps faster which you can't even notice within the game :).
beginner99 - Tuesday, September 2, 2014 - link
My 5 year old 45 nm Lynnfield is faster single-threaded than this thing. OMG.just4U - Tuesday, September 2, 2014 - link
According to bench.. it's somewhat above Intel's core2 8x and just below their first i5/3s.TiGr1982 - Tuesday, September 2, 2014 - link
Which is not surprising, because Piledriver is slower at the same clocks than Core 2, let alone any Core i7.MrSpadge - Tuesday, September 2, 2014 - link
Alright, AMD made it clear they don't think it's worth to upgrade their CPUs to their current architecture (they already have the cores) and at least 28 nm manufacturing (cheaper per transistor). And they could be right in that people just wouldn't buy enough of these to make up for the development costs.So they try to make the best use of what they have. Yet.. do they? I could imagine them doing something much better with 4-module Piledriver dies: put 2 of them onto the same package, like they do for Opterons! Sell them at decent prices and anyone wanting cheap multithread number crunchers might be interested. They might hurt 1 and 2 socket Opteron sales a bit.. but without any updates, how competitive are they anyway? And you can't put much RAM into desktop boards anyway, so this wouldn't be that much of a thread.
TiGr1982 - Tuesday, September 2, 2014 - link
Two dies won't fit in AM3+; so what you suggest is rebranding the socket G34 for a desktop and selling it cheaper, I suppose. Still a questionable idea IMHO to me.zero2dash - Tuesday, September 2, 2014 - link
Purposeful apples to oranges comparison on that chart.If they put the same GPU with the Intel CPU, the Intel CPU price drops $50 and is flat out in the middle between the 8370e and 8320e.
What a joke. I know they're trying to drum up hype for their product, but FFS at least do an even comparison when possible.
And for the comment about how AMD is 1/10th the size of Intel, give me a break.
AMD's CPU division is floundering, but they've been flat out abusing NVIDIA on price/performance for the last several years now - another company that is probably larger than AMD. The inferred excuse that "they can't compete because they're smaller" is a joke.
I love my R9's but I happily put them in an i7 setup because AMD CPU's are still not up to snuff and are still too power hungry (by comparison).
just4U - Tuesday, September 2, 2014 - link
Multi threaded performance is decent.. single is meh.. price $200 putting it in i5 category.. it needs to be sitting around $150 to be competitive... than it becomes a interesting buy. People worried about the numbers I am on a 4790K and also have systems based around the new Pentium and Amd's A10 and Vishera 6300.. I am on and off those systems quite extensively and you know what?I don't go why so slow.. omg ... similarly configured their all pretty fast.
Germanicus - Tuesday, September 2, 2014 - link
Exactly. Thank you for your refreshing dose of reality.just4U - Tuesday, September 2, 2014 - link
You want to know the really nice thing about AMD's AM3+ I can replace aging motherboards that have died and still keep the cpu. It's a good platform overall just people want a real update I guess. I build/upgrade 20+ systems a year and do use the AM3+ platform when the right deals come along. I am fine with the update... although I do think the price on that cpu should be around $160 to make it viable.dj christian - Wednesday, September 3, 2014 - link
"I don't go why so slow.. omg ..."What?
just4U - Tuesday, September 2, 2014 - link
Ian,I'd really like to see an article (blog whatever..) about a baseline system. What you feel is still viable for todays computing needs. Occasionally I still have to do work on X2's and P4's and have come to the conclusion that they should have been retired long ago.. but Phenom 2 setups and Core2's (8x series not 6..) still seem to be trucking along perfectly fine with new hardware surrounding them (SSD's video etc.)
Basically something you could refresh once a year or so.. you know? be real cool to see that and since it would be going thru the battery of tests put thru on new setups it can be included in new cpu reviews as well as part of the comparisons.
Hrel - Tuesday, September 2, 2014 - link
So AMD's 6 core CPU is MARGINALLY faster than Intel's quad core i5 on a test written specifically to maximize the advantage of many threads and HALF as fast in the single threaded test? Come on...You're better off with a Sandy Bridge chip from 3-4 years ago than you are with a brand new AMD CPU. This is sad indeed.
I feel like Intel, at this point, might have the next breakthrough, like Conroe or Sandy Bridge, but they have no reason to release it because they've essentially stood still for 4 years and AMD still can't do more than "achieve" HALF the performance of an Intel counterpart.
Come on AMD, introduce some damn competition!
TauxiC - Tuesday, September 2, 2014 - link
The fact that Amazon had the FX-8350 on sale several weeks ago for a mere $159 and that I was able to throw that CPU into an Asus Crosshair IV Formula from 2010 (while selling my X6 1090T for $175 on eBay), and overclock that baby to 4.7GHz, and OUTPERFORM a $350 Intel i7-3770K AND an i7-4770K in Passmark, scoring 10,700 points proves that AMD's chips are extremely competitive. Made mincemeat of Intel's lineup. LOLtechguyz - Thursday, September 4, 2014 - link
so you're comparing an overclocked AMD chip to a stock Intel chip. Doesn't seem fair.with the 6 cores out for just $60 more than a quad core, that price to performance ratio rises dramatically.
a 4790k+mobo is cheaper and faster in the long run than an FX 8 core. The power costs alone will make up the price difference over a few years. Then there's the undeniable single threaded performance, which means 4 threaded applications get 100% of the cpu, while in AMD lesser than 8 threads means all that horsepower is under utilized.
You just don't realize the performance you're missing.
And don't get me started on things like min fps in gaming, which AMD can't match even with all cores in use.
royalcrown - Tuesday, September 16, 2014 - link
It's a good upgrade deal for sure; kind of dumb to tout the "thrashing" when I can simply go into the bios on my Asus Maximus Formula and simply click on a single button to OC to at least 4.2.I'll run Passmark OC'd and see what I get on my 3770k. I bet it's not taking such a bad thrashing then.
royalcrown - Tuesday, September 16, 2014 - link
http://s1191.photobucket.com/user/drinkoldcoke/lib...Absolutely stock settings with nothing turned off win 8.1 x64. Clicked Auto OC to 4.6; So 10k is really good, what about other apps ? Why does it do so badly in gaming then ?
TauxiC - Tuesday, September 2, 2014 - link
Not my fault your a braindead consumer willing to pay more than twice as much for a CPU that ultimately delivers the same performance. I have to hand it to you though, you are 100% correct, Intel IS STANDING STILL when their 2014 processors match that of AMD's 2012 offerings.http://imgur.com/J6RoBXO
bebimbap - Tuesday, September 2, 2014 - link
it is interesting you are comparing your OC'd 8350(4.7g) vs stock 3770k(3.5g) and a stock 4770k(3.5g). since if you oc'd the i7s they would make "mincemeat" of your scores too. now if your comparing pure performance per dollar then yes your amd chip is better ratio on that one particular test than intel's offerings. BUT an intel system doesn't produce as much heat as an AMD system. for the same noise level, almost silent, I can run a 4770k(4.5g) @ 1.4v with the stock cooler. Stable enough to run all my games and stay <80C. how loud must the 8350(4.7g) be to to get a score of 11k on passmark?i'm not a intel fan boy, just saying passmark isn't the only facet of a cpu that you should look for unless benching passmark is all you do. because then anyone can say a $70(not on sale) G3258(4.7g) single threaded cinebench scoring 173 makes "mincemeat" of a $159(on sale) 8350(stock) scoring 88
yannigr2 - Tuesday, September 2, 2014 - link
Piledriver cpus and motherboards with 760G chipset. Nothing more to say.jardows2 - Thursday, September 4, 2014 - link
Except for those who don't know the significance of this combination. Many 760G motherboards will support ECC RAM, and all the current AMD processors support full virtualization. Using this for a cheap, high thread count home server, where I would have to spend several hundred dollars more to get similar server performance from an Intel product.bebimbap - Tuesday, September 2, 2014 - link
Ian, I LOVE that i7 990x in the tests I'm still running a i7 920 on my x58 system as my secondary it's still snappy with an SSD of course.Burticus - Tuesday, September 2, 2014 - link
Hello, ASUS? Can you PLEASE update the Bios for my AM3 socket M4N75TD motherboard to support this new 95w CPU? Pretty please?wintermute000 - Tuesday, September 2, 2014 - link
Interesting option for home server/virt labbing. 8 cores for much cheaper than the intel equivalent. I do hope its ECC and VT-D compatible on the mobos.Laststop311 - Wednesday, September 3, 2014 - link
didnt realize the pcie lanes were not directly connected to the fx cpu's wow what trash.TiGr1982 - Thursday, September 4, 2014 - link
It not a trash, it's just an old design back from 2009, essentially.Atakelane - Wednesday, September 3, 2014 - link
A M ... what? I think this will be the reaction of the next gen of gamers to the once legendary AMD name. What has gotten into them ? I am an Intel and Nvidia fanboy but the problem is, the lack of competition from AMD (with their much hyped but pseudo multi core processors) to Intel has made sure Intel can trump them with just dual core fixed clock CPUs. I think Intel would have been forced to bring unlocked quad cores to the lowest end processors with proper competition.East17 - Wednesday, September 3, 2014 - link
One thing that has become obvious during the past 14 years or so is that EVERY TIME a new benchmark is introduced, AMD processors are ALWAYS competitive, just like the new x265 4K benchmark. Even the APUs are competitive.Then Intel updates its free compiler too and while its CPUs maintain their performance, the AMD chips are suddenly less capable in the new version of the program.
Intel is free to do whatever it want with its compiler and the software is actually good and useful, BUT the developers using Intel's compiler should respect AMD users and patch their software accordingly.
It is publicly known that if you use Intel's compiler on your software, it will identify the CPU and if it's an Intel one, it will run the most efficient routine, but if it's an AMD chip, it will run the "most compatible" routine that's considerably slower.
If the developer respects his customers and wants to offer a correctly and completely optimized piece of software, he should patch his product to run the most efficient routine on AMD CPUs too.
HisDivineOrder - Wednesday, September 3, 2014 - link
The reason AMD is making this part has less to do with data showing users who want to upgrade and more to do with users who are looking for new upgrades. They go and they can either buy CPU's that show as released over two years ago or they can buy new ones with new names that show newer releases dates. Even if the reviews state these CPU's are essentially the same as the older parts, there is a large group of users who won't care so long as the CPU's are "new."This is the same reason you saw the R9 280/280X, etc, get their flashy new name last year in spite of the cards being the exact same design as the older one.
Rebadges are more about perception than reality.
Ehetlaios - Thursday, September 4, 2014 - link
GUYS ENOUGH WITH SUPERBENCHMARKING AND SUPER INTEL TECHNOLOGY. THIS TIME WE DONT NEED THE EXTRA PERFORMANCE. INTEL DID THE SAME THING -PUMB THE SPEED AND VOUALA YOUR I7 4790K. THEY HAVE PROBLEMS WITH 14NM PROCESS. SO WHEN THE SANDY BECOMES THE REAL WALL POINT AND THE 22NM DIDNOT OFFER A PUMP OF 35% REAL PUMB ON PERFORMANCE IS LOGICAL TO SEE TINY AMD STAYS AT 32NM WITH LOW PRICES. AS FOR ENERGY EFFICIENCE ARE JUST BUBU 40$ OR 50$ LESS WITH INTEL IN 5 YEARS. ITS SAD THAT SANDY AND PILLDRIVER IS STILL HERE IN PERFORMANCE WAY. THE OLD DAYS OF 50% PUMB FROM GENERATION TO GENERATION ARE OVER. AS FOR INTEL WILL GIVE YOU 6 CORES WITH 12 THREADS I7 5790K BROADWELL AND I5 4 CORES WITH 8 THREADS I7 5820K SHOWED THAT< IF THE 14NM WILL SCALE GOOD ENOUGH IN EARLY 2015 OR SUMMER. SO AMD WITH 14NM OR 16NM WILL STRIKE WITH NEW DESIGNED 8C0RES WITH 16 THREADS IN 2015 Q4 OR 2016 Q1 IF THIS WILL BE FAILDOZER OR BRUTAL ATHLON WE WILL SEE THAT. SORRY FOR MY BAD ENGLISH.BiggieShady - Thursday, September 4, 2014 - link
How come we have AMD processors with lower TDP and no thermal tests?GrigioR - Thursday, September 4, 2014 - link
So... That's for people that are still using Phenon ll's /Athlon ll's (maybe even low end fx's) Those low TDP chips are for people that have 95w max power boards. Simple as that.Yes, they do quite fine but i would NOT recommend them to any new builders. They are 'viable' upgrade options for AMD users.
That said, USD 200.00 on that chip? No, and no. Buy an I5, period. If you have a AM3 board and you fit in that description (Phenonll/Athlonll)... go ahead, better performance with less power/heat.
darknite414 - Tuesday, September 16, 2014 - link
I wonder how many times I have to hear the FX series isn't as good as the i5 or i7. I think I get it, most didn't plan on buying just wanted to sound cool! News flash....you don't.