Is it really? Intel's branding is made to mislead. i7 low power gets you nothing good but the average consumer thinks it's a good performing part just because of its label.
Has i5 2 or 4 cores? What does the X in i7-6950X mean? Intel's naming scheme is worse in my opinion. It can be awkward and confusing, especially if you compare different market segments.
Well, Intel is much more confusing. Bcos they dont state anything. For example. When you buy at the market. You can see an i7. You thought having an i7 will be great. But you need some education on intel's money-tech man. Bcoz Intel have this U - UltraLow versions, which are slow and what you buy is just the Intel and i7 name. The Base version with no U or K version and the K - Unlocked version and for money sake. Celeron and Pentiums have the same CPU stickers. Core i3U, i3, i5U, i5, i5K, i7U, i7, i7K have all the same background logo sticker design. The difference is just the numbers 3, 5 and 7. Of course you can include the i7 premium with having gray colored sticker but its only that. Unlike AMD you can even differ what is better by simply looking at their stickers. AMD E1 and E2 have plain white background. AMD A4, A6 have plain red backgrounds. A8, A10 have this red abstract backgrounds and FX have more complex abstract background. Even just by looking at their stickers. You can tell what AMD series are Higher.
And of course for AMD 7th Gens, AMD E2 and A6 have plain white backgrouds, AMD A9 and A10 have this Solid red background and lastly A12 and FX with this Complex red backgroud.
Dont you have any brain? E - just simply means Everyday, A - All-Around and FX - For eXtremes. Its just in any brand having X means it have eXtreme performance, eXtreme technology, eXtreme computing.
Meanwhile, for power users having laymen parrot "muh i7" is one of the most annoying things I ever see.
WHICH ONE? There are dozens upon dozens of i7 series CPUs ranging 6 generations. For example, I know a coworker that bought a "slightly" used "i7" laptop (specifically a Dell Latitude E6410) for what he believed to be a great deal on Craigslist, for $500, stating that it was half the price of getting a new gaming laptop and that it'd be great for his LoL gameplay.
1) It was an i7-620M, apparently and those are pretty terrible. 2) It ran a 3100M gpu, which is terrible. 3) The fan sounded absurdly loud. 4) He could've gotten the same thing used on eBay or Amazon for ~$300 if he wanted that particular model in a very clean condition.
There's a huge difference between an i7 made 6 years ago and an i7 made today. Having FX9830P may not make sense to the average consumer, but it's great for the actual enthusiasts who know their stuff.
They seem to be milking the heck out of 28nm with lots of optimizations. Though they should be doing a new architecture instead of doing baby steps at a time.
It's a stopgap measure. Even if Zen was 100% ready now they couldn't produce it in quantity. Glofo hasn't put anything 14nm on the market yet as far as I'm aware.
I am a bit baffled that AM4/Bristol Ridge is completely MIA on desktop. Like not even a hint,.... Because not ready for ZEN? The entire point of AM4 was to have one socket design supporting all CPU's but that means also ZEN?
Yeah, where is AM4? And what's the deal with GCN 1.2 in these APUs? I understand they cannot go to GCN 4.0 at 28nm because it would involve a ton of rework for little benefit but GCN 1.3 was produced on 28nm ...
I think the sensible path for AMD is now to concentrate on Zen for desktop and get that right, and not use limited resources bringing Bristol Ridge to desktop.
AM4 Bristol Ridge was on the roadmap, and if released it will likely make money for AMD, so unless serious technical problems crept up, junking it now doesn't make sense. Even with Zen, APU's weren't expected until 2017, so unless that's brought up to 2016, AMD will be left for a long time with an aging APU platform.
I would almost wonder if Zen is the one with the serious technical problems and they're now concentrating effort away from Bristol desktop APUs and towards getting Zen back on track.
It is odd, right up until reading articles this morning, I was under the impression that today was going to be the desktop AM4 launch with Bristol Ridge APUs.
There is only one reason to not even attempt to do this - Zen has been brought forward (to October) as per rumours and it will now be the launch product for AM4.
I bet AMD wish they had done an Excavator based FX 8-core chip last year now.
@psychobriggsy: "There is only one OPTIMISTIC reason to not even attempt to do this - Zen has been brought forward (to October) as per rumours and it will now be the launch product for AM4."
FTFY
There are a couple of pessimistic reasons not to do it. 1) Performance may not have reached target levels such that it is more cost effective to continue to sell Piledriver based chips rather than incur the expense of launching a chip that merely parallels it. Keep in mind that the original excavator had better performance per power than steamroller at low power levels, but quickly lost that advantage as power went up. 2) Technical problems may have occurred such that it is no longer cost effective to bring it to market. The projected Return On Investment needs to be higher than the cost of continuing the project. 3) Technical problems may have occurred such that even if projected ROI is higher than continuing the project, AMD doesn't currently have the necessary cash flow to complete the project. Procuring the needed cash may be difficult for what some may look at as a DOA platform. 4) Some combination of the above. 5) Some business reason that we aren't aware of.
That all said, I hope it is the optimistic reason.
Beyond ~23W Steamroller gets better frequency per watt scaling. Architectural efficiencies can only take you so far if the frequencies don't scale. On the desktop, at 65W to 125W, it is hard to say that excavator would beat Piledriver, much less Steamroller. I'm fairly certain AMD wasn't just holding out on us.
And here I was hoping that AMD could do a decent launch for a change. Don't know why I'm such an optimist.
Seems like AMD keeps saying: we have these chips, we've worked on them to improve the last gen, and now we're not going to give samples to reviewers, because we don't really want anyone to get the wrong impression that we're actually trying to sell them. No, they're just technical exercises we do because we like to develop stuff and make presentations about it.
And of course I'm very disappointed that Bristol Ridge AM4 is a no show.
A pity that you treat Stoney Ridge as an afterthought, that's the interesting platform here. It lands somewhere between Atom and Core M. At higher power ofc but if Apollo Lake isn't a substantial jump in single threaded perf, Stoney Ridge becomes interesting, allowing us to escape the very low single threaded perf Atom offers.
What 28nm brings for Stoney Ridge is cheapness. Which is the only way AMD can compete against bucket-bottom priced Atoms. The single-thread performance increase/leap is going to make low-end devices feel a lot nicer.
OK, so if Excavator doesn't scale well, and it is designed for situations under 35W TDP... put out some Excavator SKUs for AM1. Make an AM1+ socket if you have to, for god's sake. I have sold more PCs with that socket than any other in the past year, oddly enough - it just seems to fit that perfect niche of power/options/price that budget users need.
I think that AM4 would do just as well. Even if Bristol Ridge doesn't reach the top of the like Kaveri, it will likely compete well with its 65W variants. Even the 35W chips have higher GPU clocks than the 95W Kaveri chips, and GCN 1.2 to boot (and DDR4), so my guess is that even if the CPU side is slower, total gaming speed could be higher.
One possible optimistic case is that there's a lot more laptop demand than AMD expected, so they're concentrating on that, which likely means that a lot of 6 CU chips will actually have 8 working CU's.
I wonder if the "design wins" are going to be real this time. Although, Carrizo technically had some design wins from big OEMs these models were practically non-existant in western markets
The same reason that Intel is getting out of that market. It is pretty much impossible to compete with all the ARM players like Mediatek, Qualcomm, and Samsung. Even a competitive product won't make you any money.
What does Intel have to do with this? Heat is a function of chip power and the cooler you use. AMD has had much (much much in the case of the AMD FX-9590) higher power draw than Intel in recent years. Just because your 7 year old undervolted CPU happens to be efficient (or well cooled) doesn't suddenly make it Intel's fault that others aren't, not to mention that your assertion that "everyone assumes" is probably dead wrong even when taken as "most people assume".
But I guess you needed to make an anti-Intel rant (even though it seems you haven't really been an AMD client lately).
And yes intel brought this problem with their finfet tech and specially the pathetic thermal compound on IHS.
In the times of Sandy Bridge the stock 2600K's were topping at 40-45°C (with no undervolt) with coolers similar to mine. Sandy was the last thermal efficiente CPU from intel, also the best overclocker on air 4.8Ghz was basically peace of cake or even a "bad chip" since real air overclocks started at 4.9-5Ghz.
Thanks for the interesting info. I'd still imagine that GPU's got there first, so it's not like Intel broken new ground in heat, it just went where other chips went first.
AMD biggest problem is availability. AMD Sales & Marketing completely underestimate the potential of APU in thin & light ultrabook segments. AMD has been capped in ugly and bulky 15.6"~17.3" segments mostly with 1366x768 panels. I have been looking to get APU thin & light ultrabook in 13"~14.1" segments with anything better than 1366x768 panel. 1440x900, 1600x900 or 1920x1080 and there is none. However, the funny story is OEMs do not seem to be sincere in offering AMD APU products that sells. Some of them even falsely advertised the availability of customization option of FHD panel. Only when one tried to order it, then realized that FHD panel option was only available to Intel models.
Just a minor nitpick. On your "APU generations" chart, Trinity is listed as using Steamroller cores and Kaveri is listed as using Piledriver. I believe these two should be reversed. Good article, and I appreciate the resistance to give into base speculation. While we can assume the lack of information on desktop Bristol ridge parts is due in part to Zen launching at the end of the year, without a statement from AMD, speculation is all it can be.
GCN 1.2, for shame. I know CPU and GPU development are separate teams and all, but I think APU's would be far more viable and far sooner, if they would try to get the latest GCN updates on the latest APU's. I know their naming scheme changed but this needed to have GCN 4.
More deceptive marketing slides from AMD as usual. Disclaimers can be found here http://www.notebookcheck.net/fileadmin/Notebooks/N... Who is AMD trying to fool? That "Kaveri" @15W is an FX-7500 which is a 20W TDP part http://products.amd.com/en-us/search/APU/AMD-FX-Se... thus expect some (aggressive) throttling to run at 15W "cTDP" since its thermal and power management was not designed to run at much lower wattage. Carrizo on the other hand was designed to run as low as 15W. Higher frequency does not always translate to higher performance. Does that A9 compare favourably to Intel's Core i3-6100u? Here's some leaked benchmarks for AMD's latest A9-9410 mobile APUs https://browser.primatelabs.com/geekbench3/search?... Then compare them to Intel's Core i3-6100u https://browser.primatelabs.com/geekbench3/search?... And since AMD's A9-9410 APU is only dual core (2C/2T), then it will have a hard time competing even against Intel's lower tier Pentium 4405u https://browser.primatelabs.com/geekbench3/search?... Seriously it is not a competitor to Intel's Core i3 U-series chips. Additionally AMD's A8-7410 "Carrizo-L" APUs uses much slower (low power) Puma+ cores (in other words, not Excavator cores as found in Carrizo), and that of course skews the IPC comparison further (referring to that Cinebench R11.5 graph).
And who is Intel trying to fool? I've been using AMD and Intel Brands for about more than decades now. And you know what? My AMD powered ones are still Up and Running smoothly. While my Intel ones are dead. Due to Overheating and all sorts of problems. Its just you waste your money for a thing that you can't even show to your grand grandchildrens what computers like 20, 30, 40 years ago. Thats the money digging Intel.
AND FOR MONEY SAKE. ALWAYS REMEMBER DONT COMPARE INTEL PROCESSORS BY NUMBERS. COMPARE THEM WITH VALUE AND PRICE. AMD FX and A12 to Intel i7, AMD A10 to Intel i5, AMD A9 to Intel i3, AMD A6 to Intel Pentium and Celeron, AMD E2 to Intel Atom. Remeber Intel have this U, Base and K version. *Careful for Intel brand scams. You thought having i7, i5 or i3 would make you good to go? Even my AMD A4 - 5000 is much faster than an Intel i3 5th Gen with a U version. Also remember. Numbers. Intel has so many confusing numbers. Example and this is just an example. Core i5 5th gen. Maybe can have this numbers, i5 - 5010U, 5020U, 5100U, 5010, 5020, 5100, 5100K and so on. So count all you want.
PRICEEE!!! So, dont complain why is AMD A10 with Intel i5. Its PRICEEEE!!! Are you blind? Compare with Price. And you'll see with comparing Price and not the crappy brand. AMD will exceed with Performance/Price. And in your head. Really? A10 can overthroned i5? Remember Intel have this U versions which is still expensive. That you thought its best and better than AMD.
You can still buy these today on ITX desktop boards such as the Biostar A10N-9630E Ver. 6.0. I am very happy with mine, not that it comes close to Ryzen, but still good.
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
69 Comments
Back to Article
Eden-K121D - Tuesday, May 31, 2016 - link
We already have articles. niceMeteor2 - Wednesday, June 1, 2016 - link
'FX 9830P' AMD has terrible branding. It's makes no sense.Intel's m3/5/7 and i3/5/7 is so much clearer to a consumer.
mickulty - Wednesday, June 1, 2016 - link
"G3258"ZeDestructor - Wednesday, June 1, 2016 - link
Going by Metero2's example, that would be the Pentium, and nothing more.We just call it the G3258 cause we care what exact chip.
jjj - Wednesday, June 1, 2016 - link
Is it really? Intel's branding is made to mislead. i7 low power gets you nothing good but the average consumer thinks it's a good performing part just because of its label.UtilityMax - Friday, June 3, 2016 - link
Heh, Half of people I know bought high end laptops with i7 U CPUs (low power) for "number crunching".T1beriu - Wednesday, June 1, 2016 - link
If you're comparing with Intel's i3/5/7 remember that AMD has E2/6 A8/10/12/FX branding lines.Meteor2 - Saturday, June 4, 2016 - link
That's my point. Is FX higher or lower than A? What's the significance of X?gruffi - Friday, June 10, 2016 - link
Has i5 2 or 4 cores? What does the X in i7-6950X mean? Intel's naming scheme is worse in my opinion. It can be awkward and confusing, especially if you compare different market segments.FelG - Thursday, September 14, 2017 - link
Well, Intel is much more confusing. Bcos they dont state anything. For example. When you buy at the market. You can see an i7. You thought having an i7 will be great. But you need some education on intel's money-tech man. Bcoz Intel have this U - UltraLow versions, which are slow and what you buy is just the Intel and i7 name. The Base version with no U or K version and the K - Unlocked version and for money sake. Celeron and Pentiums have the same CPU stickers. Core i3U, i3, i5U, i5, i5K, i7U, i7, i7K have all the same background logo sticker design. The difference is just the numbers 3, 5 and 7. Of course you can include the i7 premium with having gray colored sticker but its only that. Unlike AMD you can even differ what is better by simply looking at their stickers. AMD E1 and E2 have plain white background. AMD A4, A6 have plain red backgrounds. A8, A10 have this red abstract backgrounds and FX have more complex abstract background. Even just by looking at their stickers. You can tell what AMD series are Higher.FelG - Thursday, September 14, 2017 - link
And of course for AMD 7th Gens, AMD E2 and A6 have plain white backgrouds, AMD A9 and A10 have this Solid red background and lastly A12 and FX with this Complex red backgroud.FelG - Thursday, September 14, 2017 - link
Dont you have any brain? E - just simply means Everyday, A - All-Around and FX - For eXtremes. Its just in any brand having X means it have eXtreme performance, eXtreme technology, eXtreme computing.JoeyJoJo123 - Wednesday, June 1, 2016 - link
Meanwhile, for power users having laymen parrot "muh i7" is one of the most annoying things I ever see.WHICH ONE?
There are dozens upon dozens of i7 series CPUs ranging 6 generations. For example, I know a coworker that bought a "slightly" used "i7" laptop (specifically a Dell Latitude E6410) for what he believed to be a great deal on Craigslist, for $500, stating that it was half the price of getting a new gaming laptop and that it'd be great for his LoL gameplay.
1) It was an i7-620M, apparently and those are pretty terrible.
2) It ran a 3100M gpu, which is terrible.
3) The fan sounded absurdly loud.
4) He could've gotten the same thing used on eBay or Amazon for ~$300 if he wanted that particular model in a very clean condition.
There's a huge difference between an i7 made 6 years ago and an i7 made today. Having FX9830P may not make sense to the average consumer, but it's great for the actual enthusiasts who know their stuff.
ET - Thursday, June 2, 2016 - link
A Core i5-6200U and Core i7-6500U differ in cache size and clock speed (and not by much). How is that clearer than AMD's names?gsalkin - Thursday, June 2, 2016 - link
I have no idea what makes an A10 different from an A12 or an A8 with Excavator different from an A8 with Beema. Man they have got to clear that up.FelG - Thursday, September 14, 2017 - link
Ehhh? You're just an intel fanboy. Back off.III-V - Tuesday, May 31, 2016 - link
AMD re-used Carrizo on the second slide on this page... good going.nandnandnand - Tuesday, May 31, 2016 - link
That's appropriate, since they've reused the 28nm architecture. Don't waste your money on these.nandnandnand - Tuesday, May 31, 2016 - link
28nm can go straight to the garbage. Zen for mobile plz.pugster - Wednesday, June 1, 2016 - link
They seem to be milking the heck out of 28nm with lots of optimizations. Though they should be doing a new architecture instead of doing baby steps at a time.Kvaern1 - Thursday, June 2, 2016 - link
It's a stopgap measure. Even if Zen was 100% ready now they couldn't produce it in quantity. Glofo hasn't put anything 14nm on the market yet as far as I'm aware.plopke - Wednesday, June 1, 2016 - link
I am a bit baffled that AM4/Bristol Ridge is completely MIA on desktop. Like not even a hint,.... Because not ready for ZEN? The entire point of AM4 was to have one socket design supporting all CPU's but that means also ZEN?plopke - Wednesday, June 1, 2016 - link
Correction I guess they do mention the A8 7410?plopke - Wednesday, June 1, 2016 - link
*A8 7410 comparison with 7thgenArnulf - Wednesday, June 1, 2016 - link
Yeah, where is AM4? And what's the deal with GCN 1.2 in these APUs? I understand they cannot go to GCN 4.0 at 28nm because it would involve a ton of rework for little benefit but GCN 1.3 was produced on 28nm ...ET - Wednesday, June 1, 2016 - link
1.3 is now called 4.0. The latest for 28nm is 1.2.fanofanand - Thursday, June 9, 2016 - link
I wish I had read this comment before my post. D'oh!Meteor2 - Wednesday, June 1, 2016 - link
I think the sensible path for AMD is now to concentrate on Zen for desktop and get that right, and not use limited resources bringing Bristol Ridge to desktop.ET - Wednesday, June 1, 2016 - link
AM4 Bristol Ridge was on the roadmap, and if released it will likely make money for AMD, so unless serious technical problems crept up, junking it now doesn't make sense. Even with Zen, APU's weren't expected until 2017, so unless that's brought up to 2016, AMD will be left for a long time with an aging APU platform.LostWander - Thursday, June 2, 2016 - link
I would almost wonder if Zen is the one with the serious technical problems and they're now concentrating effort away from Bristol desktop APUs and towards getting Zen back on track.psychobriggsy - Wednesday, June 1, 2016 - link
It is odd, right up until reading articles this morning, I was under the impression that today was going to be the desktop AM4 launch with Bristol Ridge APUs.There is only one reason to not even attempt to do this - Zen has been brought forward (to October) as per rumours and it will now be the launch product for AM4.
I bet AMD wish they had done an Excavator based FX 8-core chip last year now.
BurntMyBacon - Wednesday, June 1, 2016 - link
@psychobriggsy: "There is only one OPTIMISTIC reason to not even attempt to do this - Zen has been brought forward (to October) as per rumours and it will now be the launch product for AM4."FTFY
There are a couple of pessimistic reasons not to do it. 1) Performance may not have reached target levels such that it is more cost effective to continue to sell Piledriver based chips rather than incur the expense of launching a chip that merely parallels it. Keep in mind that the original excavator had better performance per power than steamroller at low power levels, but quickly lost that advantage as power went up. 2) Technical problems may have occurred such that it is no longer cost effective to bring it to market. The projected Return On Investment needs to be higher than the cost of continuing the project. 3) Technical problems may have occurred such that even if projected ROI is higher than continuing the project, AMD doesn't currently have the necessary cash flow to complete the project. Procuring the needed cash may be difficult for what some may look at as a DOA platform. 4) Some combination of the above. 5) Some business reason that we aren't aware of.
That all said, I hope it is the optimistic reason.
BurntMyBacon - Wednesday, June 1, 2016 - link
@psychobriggsy: "I bet AMD wish they had done an Excavator based FX 8-core chip last year now."See the power optimized library chart:
http://www.anandtech.com/show/8995/amd-at-isscc-20...
Beyond ~23W Steamroller gets better frequency per watt scaling. Architectural efficiencies can only take you so far if the frequencies don't scale. On the desktop, at 65W to 125W, it is hard to say that excavator would beat Piledriver, much less Steamroller. I'm fairly certain AMD wasn't just holding out on us.
ET - Wednesday, June 1, 2016 - link
And here I was hoping that AMD could do a decent launch for a change. Don't know why I'm such an optimist.Seems like AMD keeps saying: we have these chips, we've worked on them to improve the last gen, and now we're not going to give samples to reviewers, because we don't really want anyone to get the wrong impression that we're actually trying to sell them. No, they're just technical exercises we do because we like to develop stuff and make presentations about it.
And of course I'm very disappointed that Bristol Ridge AM4 is a no show.
yannigr2 - Wednesday, June 1, 2016 - link
1st June. Waiting to see everything from AMD. We get NOTHING.BurntMyBacon - Wednesday, June 1, 2016 - link
Or did we get everything? (0_0)jjj - Wednesday, June 1, 2016 - link
A pity that you treat Stoney Ridge as an afterthought, that's the interesting platform here. It lands somewhere between Atom and Core M. At higher power ofc but if Apollo Lake isn't a substantial jump in single threaded perf, Stoney Ridge becomes interesting, allowing us to escape the very low single threaded perf Atom offers.Any clue on Stoney Ridge die size?
ET - Wednesday, June 1, 2016 - link
The article says 124.5mm2.jjj - Wednesday, June 1, 2016 - link
Thanks,missed that somehow.psychobriggsy - Wednesday, June 1, 2016 - link
What 28nm brings for Stoney Ridge is cheapness. Which is the only way AMD can compete against bucket-bottom priced Atoms. The single-thread performance increase/leap is going to make low-end devices feel a lot nicer.iwod - Wednesday, June 1, 2016 - link
28nm, Non Zen..... Why Bother? Honestly, Why even waste the money to even tape out?milli - Wednesday, June 1, 2016 - link
So you think it would be okay for AMD to make OEMs wait a year for Zen APUs?Meteor2 - Wednesday, June 1, 2016 - link
It is if it's all they can afford to do without compromising Zen.guidryp - Wednesday, June 1, 2016 - link
Please Fix that twitter box (removing it would be nice).http://i.imgur.com/pFry7up.png
This is the second story where I can't read content because the Twittter box is blocking it.
barleyguy - Wednesday, June 1, 2016 - link
If you install a noscript plugin, the twitter box goes away completely.Eschaton - Wednesday, June 1, 2016 - link
OK, so if Excavator doesn't scale well, and it is designed for situations under 35W TDP... put out some Excavator SKUs for AM1. Make an AM1+ socket if you have to, for god's sake. I have sold more PCs with that socket than any other in the past year, oddly enough - it just seems to fit that perfect niche of power/options/price that budget users need.ET - Wednesday, June 1, 2016 - link
I think that AM4 would do just as well. Even if Bristol Ridge doesn't reach the top of the like Kaveri, it will likely compete well with its 65W variants. Even the 35W chips have higher GPU clocks than the 95W Kaveri chips, and GCN 1.2 to boot (and DDR4), so my guess is that even if the CPU side is slower, total gaming speed could be higher.One possible optimistic case is that there's a lot more laptop demand than AMD expected, so they're concentrating on that, which likely means that a lot of 6 CU chips will actually have 8 working CU's.
lefty2 - Wednesday, June 1, 2016 - link
I wonder if the "design wins" are going to be real this time. Although, Carrizo technically had some design wins from big OEMs these models were practically non-existant in western marketsBurntMyBacon - Wednesday, June 1, 2016 - link
Please correct charts:Kaveri used Steamroller cores
Trinity used Piledriver cores
https://en.wikipedia.org/wiki/AMD_Accelerated_Proc...
pugster - Wednesday, June 1, 2016 - link
Not understanding why AMD don't make any APU's about 3-10w for the tablet market.nahubbard - Wednesday, June 1, 2016 - link
The same reason that Intel is getting out of that market. It is pretty much impossible to compete with all the ARM players like Mediatek, Qualcomm, and Samsung. Even a competitive product won't make you any money.Meteor2 - Saturday, June 4, 2016 - link
Intel aren't leaving the tablet market, just the phone market.Lolimaster - Wednesday, June 1, 2016 - link
In non l2 cache reliant applications not even a 4.4Ghz Steamroller APU could touch the stock turbo Athlon 845, easily 15-18%+ IPC.At full load in cpu tasks:
Stock + turbo A10 7850 95w 145w
Stock + turbo A10 7800 65w 125w
Stock + turbo Athlon 845 65w 105w
l2 cache cut in half really hurt the excavator core.
Another thing is that for some reason the Athlon 845 tend to throttle out of nowhere.
It seems to me that AMD focused Excavator too much for mobile, being the 1MB --> 512KB the biggest nerf.
Lolimaster - Wednesday, June 1, 2016 - link
*Whole system powerconsumption.Lolimaster - Wednesday, June 1, 2016 - link
What i hate from intel is that now everyone assumes that having a 65-80°C CPU at full load is fine.My Athlon X4 620 undervolted to 1.1v at full load rarely goes beyond 30°C or now 40°C with a "touched" mobo after a PSU fail.
ET - Thursday, June 2, 2016 - link
What does Intel have to do with this? Heat is a function of chip power and the cooler you use. AMD has had much (much much in the case of the AMD FX-9590) higher power draw than Intel in recent years. Just because your 7 year old undervolted CPU happens to be efficient (or well cooled) doesn't suddenly make it Intel's fault that others aren't, not to mention that your assertion that "everyone assumes" is probably dead wrong even when taken as "most people assume".But I guess you needed to make an anti-Intel rant (even though it seems you haven't really been an AMD client lately).
Lolimaster - Thursday, June 2, 2016 - link
I use a simple Hyper 212 plus cooler.And yes intel brought this problem with their finfet tech and specially the pathetic thermal compound on IHS.
In the times of Sandy Bridge the stock 2600K's were topping at 40-45°C (with no undervolt) with coolers similar to mine. Sandy was the last thermal efficiente CPU from intel, also the best overclocker on air 4.8Ghz was basically peace of cake or even a "bad chip" since real air overclocks started at 4.9-5Ghz.
ET - Sunday, June 5, 2016 - link
Thanks for the interesting info. I'd still imagine that GPU's got there first, so it's not like Intel broken new ground in heat, it just went where other chips went first.Lolimaster - Wednesday, June 1, 2016 - link
Im that close to buying an i7 6700T and undervolt it to reduce max temps further. I can't find the chip except in some eastern europe country.WHERE IS MY UNBENCH THE KENCH, UNRIBBIT THE KERMIT, UNTOAD THE TADPOLE AM4 BRISTOL RIDGE FOR DESKTOP.
kws1cj - Monday, June 6, 2016 - link
AMD biggest problem is availability. AMD Sales & Marketing completely underestimate the potential of APU in thin & light ultrabook segments. AMD has been capped in ugly and bulky 15.6"~17.3" segments mostly with 1366x768 panels. I have been looking to get APU thin & light ultrabook in 13"~14.1" segments with anything better than 1366x768 panel. 1440x900, 1600x900 or 1920x1080 and there is none. However, the funny story is OEMs do not seem to be sincere in offering AMD APU products that sells. Some of them even falsely advertised the availability of customization option of FHD panel. Only when one tried to order it, then realized that FHD panel option was only available to Intel models.ajlueke - Tuesday, June 7, 2016 - link
Just a minor nitpick. On your "APU generations" chart, Trinity is listed as using Steamroller cores and Kaveri is listed as using Piledriver. I believe these two should be reversed.Good article, and I appreciate the resistance to give into base speculation. While we can assume the lack of information on desktop Bristol ridge parts is due in part to Zen launching at the end of the year, without a statement from AMD, speculation is all it can be.
fanofanand - Thursday, June 9, 2016 - link
GCN 1.2, for shame. I know CPU and GPU development are separate teams and all, but I think APU's would be far more viable and far sooner, if they would try to get the latest GCN updates on the latest APU's. I know their naming scheme changed but this needed to have GCN 4.fanofanand - Thursday, June 9, 2016 - link
Wish I could delete this comment.....forgot that GCN 4 is not designed for 28 nm. Get us an edit button Anandtech!rtho782 - Thursday, June 9, 2016 - link
Wow, a processor release more boring than Broadwell-E. I didn't think it could be done!BlueBlazer - Friday, June 17, 2016 - link
More deceptive marketing slides from AMD as usual. Disclaimers can be found here http://www.notebookcheck.net/fileadmin/Notebooks/N... Who is AMD trying to fool? That "Kaveri" @15W is an FX-7500 which is a 20W TDP part http://products.amd.com/en-us/search/APU/AMD-FX-Se... thus expect some (aggressive) throttling to run at 15W "cTDP" since its thermal and power management was not designed to run at much lower wattage. Carrizo on the other hand was designed to run as low as 15W. Higher frequency does not always translate to higher performance. Does that A9 compare favourably to Intel's Core i3-6100u? Here's some leaked benchmarks for AMD's latest A9-9410 mobile APUs https://browser.primatelabs.com/geekbench3/search?... Then compare them to Intel's Core i3-6100u https://browser.primatelabs.com/geekbench3/search?... And since AMD's A9-9410 APU is only dual core (2C/2T), then it will have a hard time competing even against Intel's lower tier Pentium 4405u https://browser.primatelabs.com/geekbench3/search?... Seriously it is not a competitor to Intel's Core i3 U-series chips. Additionally AMD's A8-7410 "Carrizo-L" APUs uses much slower (low power) Puma+ cores (in other words, not Excavator cores as found in Carrizo), and that of course skews the IPC comparison further (referring to that Cinebench R11.5 graph).FelG - Thursday, September 14, 2017 - link
And who is Intel trying to fool? I've been using AMD and Intel Brands for about more than decades now. And you know what? My AMD powered ones are still Up and Running smoothly. While my Intel ones are dead. Due to Overheating and all sorts of problems. Its just you waste your money for a thing that you can't even show to your grand grandchildrens what computers like 20, 30, 40 years ago. Thats the money digging Intel.FelG - Thursday, September 14, 2017 - link
AND FOR MONEY SAKE. ALWAYS REMEMBER DONT COMPARE INTEL PROCESSORS BY NUMBERS. COMPARE THEM WITH VALUE AND PRICE. AMD FX and A12 to Intel i7, AMD A10 to Intel i5, AMD A9 to Intel i3, AMD A6 to Intel Pentium and Celeron, AMD E2 to Intel Atom. Remeber Intel have this U, Base and K version. *Careful for Intel brand scams. You thought having i7, i5 or i3 would make you good to go? Even my AMD A4 - 5000 is much faster than an Intel i3 5th Gen with a U version. Also remember. Numbers. Intel has so many confusing numbers. Example and this is just an example. Core i5 5th gen. Maybe can have this numbers, i5 - 5010U, 5020U, 5100U, 5010, 5020, 5100, 5100K and so on. So count all you want.FelG - Thursday, September 14, 2017 - link
PRICEEE!!! So, dont complain why is AMD A10 with Intel i5. Its PRICEEEE!!! Are you blind? Compare with Price. And you'll see with comparing Price and not the crappy brand. AMD will exceed with Performance/Price. And in your head. Really? A10 can overthroned i5? Remember Intel have this U versions which is still expensive. That you thought its best and better than AMD.Rouxenator - Wednesday, December 16, 2020 - link
You can still buy these today on ITX desktop boards such as the Biostar A10N-9630E Ver. 6.0. I am very happy with mine, not that it comes close to Ryzen, but still good.