AMD says 65W, triple-core A6-3500 APU now available

A press release from AMD just dropped into our inbox announcing the imminent availability of a new model of desktop CPU based on Llano silicon, the A6-3500. Like the four other desktop variants of Llano, this processor will drop into a Socket FM1-style motherboard and provide both CPU and graphics processing power to the system.

At first blush, the specs for the A6-3500 read almost exactly like the specs for the next model up, the A6-3600. Both run their CPU cores at 2.1GHz and can raise them to 2.4GHz via Turbo Core, both have 444MHz Radeon IGPs with 320 shader ALUs, and both will fit into a 65W TDP envelope. The difference? The A6-3500 lost a limb in a horrid tree-chipper accident, so it has only three cores and 3MB of L2 cache, versus the A6-3600’s quad cores and 4MB of L2.

The A6-3500 is priced at $95, about $40 below the fastest desktop Llano, the A8-3850.

AMD tells us the A6-3500 "is now available for purchase through system builders and online retailers." However, we think perhaps the press release has gotten ahead of the horse. Quick searches at Amazon, Newegg, and at our PriceGrabber search engine currently reveal zero hits for the A6-3500 processor. AMD may be supplying these parts to the channel, but they haven’t quite made it to the virtual shelves of online retailers.

If the A6-3500 were available for purchase today, it would be the first 65W desktop Llano variant to hit retail. The other two, the A6-3600 and the A8-3800, aren’t yet at the major online retailers yet, either. Those parts were announced in late June, alongside some 100W desktop models that are, to date, the only desktop Llano derivatives firmly in the sales channel.

We’d be more worried about a triple-core Llano being the first and only one to reach 65W on the desktop if we hadn’t caught wind of the imminent availability of quad-core 65W parts. Our sense is that the A6-3500’s release is simply AMD filling out its Llano-based lineup, not a tacit admission of problems reaching 65W with a similarly clocked quad-core chip.

The $95 price tag on the A6-3500 does tell us something interesting, though. Those 65W quad-core parts, the A6-3600 and A8-3800, are likely to be priced between $95 and $135, which is a fairly narrow range. That raises some difficult competitive questions, because the 65W parts have substantially lower clock speeds than their 100W siblings, yet those 100W parts struggle to keep pace with the Intel Core i3-2100 in CPU performance tests (though the situation is reversed in GPU tests). The Core i3-2100 is priced at $117 and has a 65W TDP rating.

Comments closed
    • Arclight
    • 8 years ago

    I think that the integration, in time, will bring benefits as the APU (or however you like to call it) will have the best from both worlds. But as it is right now the IGP for people like me, represent only additional heat and in a way a handicap to the CPU since it could have used that space/transistors for more cache or what have you.

    I recognize truely the need for an APU, low power users who only do web browsing and very lite gaming (as in helium lite 🙂 ) and i see why people want them, but for now i’d love to see a SKU of Core i5 2500 for example without an IGP and a 65 W TDP instead of 95W (i’m pulling the numbers out of my ass since i don’t know what’s the TDP of the IGP but you get what i mean).

      • OneArmedScissor
      • 8 years ago

      Ok…except the GPU is power gated when you’re using a graphics card, and TDP is a ball park figure of what to expect when pushing [i<]everything[/i<] in the chip as far as it will go, at the same time. They're not going to give you more cache, as it's likely useless in that particular configuration. You can already see that there's no real difference between the i5 and i7 when it's not benefitting from hyperthreading. The GPU in Sandy Bridge is already a small part of the chip and it will just shrink more with Ivy Bridge, so it's not like you're losing out on something. You're basically saying to Intel that you'd like to spend hundreds of dollars on snake oil. Not a good idea.

        • Arclight
        • 8 years ago

        [quote<]The GPU in Sandy Bridge is already a small part of the chip and it will just shrink more with Ivy Bridge, so it's not like you're losing out on something. You're basically saying to Intel that you'd like to spend hundreds of dollars on snake oil. Not a good idea.[/quote<] So your saying that if i ask Intel to make a chip with 1 less part (IGP) they will charge me more? Also are you saying that there is positively no way they could use the extra space to improve performance? I am at loss....

    • d0g_p00p
    • 8 years ago

    Brazos, Zacate, Llano, C-350, Zambezi, A8-3800, Bulldozer, Stars, FX-8100, Bobcat, Fusion…….

    No wonder the consumer is confused on just what the hell they are buying. I don’t even know what CPU TR is talking about anymore without a quick trip to wikipedia to refresh my memory.

      • Joe Miller
      • 8 years ago

      … about the same from Intel …

      • thefumigator
      • 8 years ago

      To make it easier for you:

      AMD Llano = general purpose.
      Intel Core i3 = facebook.

      AMD Zacate = general purpose on low power battery friendly laptop/netbook.
      Intel Atom = facebook on low power battery friendly netbook only.

      AMD Bulldozer = Promises delayed.
      Intel Core i7 = the thing that “should be” squashed by Bulldozer in price/performance terms at least.

      AMD Fusion = future on a budget
      Intel Bridges = future for a price.

      Is that simple.

    • Krogoth
    • 8 years ago

    I find it funny that enthusiast are arguing over a product that is clearly geared towards the OEM/mainstream market.

    CPU performance doesn’t really much that much anymore for mainstream. Intergrated GPU performance matters sightly more, as long as it can effortlessly handle HTML5, H.246 decoding/encoding.

    • ronch
    • 8 years ago

    I’m not sure anyone here at TR or its readers will find this chip terribly exciting, unless he/she is building a PC for his/her grandmother. I know AMD needs to stoop low at the low end, but $95 seems a tad high considering how crippled this thing is. Compared to a PDC with an IGP or a same-performance discrete graphics card, however, this A6 may be your ticket.

    • flip-mode
    • 8 years ago

    price and power too high. performance too low.

      • TheEmrys
      • 8 years ago

      If it were a 45w part, I’d be all over it. As it is…. I’d rather go Zacate or 2100t.

        • Farting Bob
        • 8 years ago

        Its likely a noticablely less power hungry chip than the 4 cores, TDP doesnt mean much anyway, they are always very conservative estimates. In reality it probably uses as much power as a chip labelled 45w with all its cores intact.

      • Anonymous Coward
      • 8 years ago

      Yeah 2.1ghz, three K10 cores and 65W is not inspirational. Pretty bottom of the market stuff.

    • jensend
    • 8 years ago

    I call BS. To say the 3850 “struggles to keep pace with the Intel Core i3-2100 in CPU performance tests (though the situation is reversed in GPU tests)” is disingenuous and misleading at best. Your own testing showed the 3850 an average of 2.2% behind the i3-2100 in CPU tests- 3.9% behind if you take a geometric mean rather than an arithmetic mean. That’s hardly a “struggle” – it’s not even noticeable.

    Even if you totally ignore all the cases where the 3850 won (perhaps you’re certain everything you’re interested in is single-threaded), when it lost, it lost by an average of 15%- noticeable but just barely. In the mean time, in IGP testing the i3-2100 was anywhere from half to 1/4 the 3850’s performance i.e. 50-75% behind. That’s hardly just the same situation reversed.

      • Damage
      • 8 years ago

      You may not like the phrasing, but I am not unfamiliar with the performance results I generated and reported to the world. No need to quote ’em at me. 😉 I was simply trying to provide a quick summary of a complex issue. Note the hyperlinks included to the complete reviews I’ve done on Llano desktop and mobile.

      My main intended point was that the 65W Llano parts look to be pretty slow, CPU-wise: 500MHz less on base clock for the A8-3800 vs. 3850, and 200MHz less with Turbo Core doing its best when possible–and Turbo Core has so far proven to be not consistently helpful/measurably better in Llano. The A8-3800 is likely to be quite a bit slower than the i3-2100 and friends in multithreaded tests and much, much slower in single-threaded ones.

      Yes, the IGP is superior in Llano, but it looks possible the A8-3800 may sell for a price closer to the Core i3-2105 (with a bit better IGP) or the Core i3-2120 ($138, 3.3GHz, 65W), neither of which I mentioned in the post, but if I were going for maximum fairness, I suppose I should have spent time on them. All of which makes the CPU situation a bit uncomfortable.

      Yes, some folks may choose to prioritize the better IGP over the better CPU, but this class of product has little history of caring about gaming graphics performance. Those who do care tend to want something more robust than *any* extant IGP. So… there is difficulty in the 65W Llano products being priced so close to Intel’s 65W Sandy Bridge parts, if in fact that’s what happens. My sense is that AMD should be building in more of a discount, all things considered.

        • jensend
        • 8 years ago

        I wasn’t saying you didn’t know what the results were, I was calling you out on the irrational spin you’ve given those results, which misleads your audience. A lot of people will clearly remember that you’ve consistently panned Llano as being a total disappointment of a product; [i<]not so many will remember the benchmark figures that show otherwise[/i<], especially given that you've presented those results in a way that tricks people into thinking they justify your remarks. (The clearest example of this was focusing on the matchup between the A8-3500M and desktop Sandy Bridge models with 60W higher TDPs, overwhelming readers with benchmark figures which are much better than the 3500M's but totally irrelevant to its success-- and not including any comparisons to its direct competition, the Core i3 SB mobile parts.) Rather than simply spreading gloomy speculation about how awful the performance of the 65W parts will be, you could have looked at extant reviews ([url<]http://www.xbitlabs.com/articles/cpu/display/amd-a8-3800.html[/url<]) and seen that the A8-3800 is not that slow (an average of 11% behind the i3-2100 and the identically-clocked 2105 in xbitlab's CPU tests, 14% if you look at geometric mean) and still blows the i3-2105 out of the water in integrated gaming tests (still >2x as fast on average even though the 2105's IGP is much better than the 2100's). It will be competitive at its price point. I think there's a simple explanation for why AMD isn't selling the 65W Turbo-enabled versions for much less than their 100W counterparts: supply and demand. For this first while as they work on optimizing their yields, I don't think they could supply the number of 65W parts people would be demanding if these parts were $30 cheaper than their 100W counterparts and their Intel competition. The OEM market may not care quite as much but I'd bet a lot of the retail interest in Llano is related to inexpensive SFF and other situations where the power is going to make a big difference (and where a 65W APU's power advantage over a discrete graphics setup would make it enticing).

          • Damage
          • 8 years ago

          [quote<]The clearest example of this was focusing on the matchup between the A8-3500M and desktop Sandy Bridge models with 60W higher TDPs, overwhelming readers with benchmark figures which are much better than the 3500M's but totally irrelevant to its success-- and not including any comparisons to its direct competition, the Core i3 SB mobile parts.[/quote<] Wow, strange claim. Truth is, we talked with AMD when putting together our mobile Llano review expressly about the question of product positioning and the competition. And we took their advice that the Core i5-2410M was the closest competitor to the A8-3500M. We then ordered a comparable laptop based on it, as noted here: [url<]https://techreport.com/articles.x/21099/6[/url<] [quote<]For comparison to the Llano review unit, we ordered up the closest analog we could find in stock at Newegg, the HP ProBook 6460b pictured above. Like the Llano system, the ProBook has a 14" 1366x768 display with a matte coating, a Hitachi 7K500 mobile hard drive, 4GB of RAM, an optical drive, and Windows 7. The processor in the ProBook is a Core i5-2410M, which we believe to be the closest competitor to the A8-3500M APU. Like the A8, the Core i5-2410M has a 35W TDP rating, but the i5-2410M has only two cores to Llano's four.[/quote<] Choosing the i5-2410M required some trust of AMD on our part, since we didn't have exact pricing or a clear read on how retail laptop prices would shake out in the long run prior to the product's launch. Even if A8-3500M laptops end up selling at more of a discount than expected vs. Sandy Bridge over time, we did our best to make the right comparison while working directly with AMD. Moreover, that i5-2410M-based laptop was *the* focus of our comparison in the mobile Llano review, not desktop Sandy Bridge parts or anything else. The rest was just context, as we plainly noted. Your claims to the contrary indicate a misreading of our intentions and efforts. That seems to be a consistent issue for you, though: [quote<]A lot of people will clearly remember that you've consistently panned Llano as being a total disappointment of a product[/quote<] Our overall assessment of the mobile Llano was quite positive, which you seem not to recall. What we wrote: [quote<]Meanwhile, AMD has Intel utterly outclassed in integrated graphics. You've seen our discussion of texture filtering quality and our performance results. Where the Llano IGP delivers playable frame rates in some of the latest games, Intel's HD 3000 treads on the edge of uselessness. Add in a host of other considerations, including the vastly superior hardware feature set of the Radeon IGP and its ability to partake of the sweet, sweet stream of Catalyst driver updates, and this is a difference between the products that truly matters.[/quote<] [quote<]If the big PC makers can translate what we've seen of the A-Series APUs into systems that are as sleek, cool-running, and endurance-endowed as their Sandy Bridge counterparts, AMD should have a hit on its hands.[/quote<] Our review of the 100W desktop parts was less so, since the CPU portion of the part failed to outperform even AMD's own Phenom II X4 840 yet costs ~$35 more--and, again, we think IGPs are less attractive in 100W (i.e. large) desktops. Our biggest complaint there was that 100W parts seemed not to play to Llano's strong suit, and we noted that the 65W parts may make more sense in certain markets. We still think that's true, but pricing is an issue. I'll agree yield problems may be affecting the supply of 65W Llanos, but if those problems are leading to inflated prices, well, those parts have a competitiveness problem--and consumers should know about it.

            • jensend
            • 8 years ago

            They aren’t “inflated” prices unless they’re not competitive, and as I pointed out there’s no reason to think they will be uncompetitive at their price points.

            Sure, you talked about the i5-2140M comparison (and I have no beef with you making that comparison), but your first order of business was two pages of comparing it to 95W TDP desktop i5s and lots of other desktop parts. Only then did you throw in one page where you removed irrelevant desktop parts from the picture. You can claim that’s “just context,” but placing data in an irrelevant context makes it misleading- often just as misleading as totally falsified data. I would hope that, as a full-time technical journalist, you’ve read “The Visual Display of Quantitative Information” by Tufte and seen him break down this and other ways people mislead by the way they deal with data.

            • Damage
            • 8 years ago

            So when we said this:

            [quote<]Our first round of performance tests will compare our two mobile systems against a range of desktop processors in many of the components of our CPU test suite. We think these comparisons can be a nice backdrop for our A8-3500M-versus-Core i5-2410M contest, but remember the mobile processors have to work within much smaller power envelopes.[/quote<] ..and when we colored all of the desktop parts in light gray, to delineate them from the mobile processors (such as the i5-2410M, which was included in bright blue all of those results)... ...and then provided eight more pages of mobile-only results.... ...and produced a nuanced and generally positive conclusion.... You decided that our visual presentation needs work, that we've used "irrational spin" that "misleads your audience", and that "A lot of people will clearly remember that you've consistently panned Llano as being a total disappointment of a product". Hmm.

            • flip-mode
            • 8 years ago

            Damage, you forgot about the intangible benefits of AMD’s products though. I can’t think of them off the top of my head, but that shouldn’t stop you from factoring them in and concluding that Llano is fierce and the competition is just a hot tranny mess.

            • derFunkenstein
            • 8 years ago

            I think this was meant to be humorous rather than serious, but you got dinged anyway. I helped to correct that oversight.

            • jensend
            • 8 years ago

            It doesn’t make a “nice backdrop” any more than charting the performance of mobile processors against the salaries of NBA stars makes a “nice backdrop.” It’s irrelevant information, and you put it front and center. You did have a brief disclaimer blurb, but that makes little difference to the cognitive impact of the charts on most people reading those pages; spend a couple lines of print saying one thing and then spend twelve charts and ten paragraphs saying something diametrically opposed, and which point are people going to come away with?

            When I wrote my first reply to you in which I said you’ve “consistently panned Llano” and cited the 3500M review as an example of dodgy ways of presenting data, I was writing from memory. I admit I didn’t remember how you phrased your conclusion in the 3500M review at that point- that wasn’t the core of what I (or most readers, I’ll wager) took away from that article.

            If you wanted to have some kind of appendix (or another article) where you explore the question of how laptop processors compare to their desktop counterparts, that would have been just dandy. But those two pages were the first 2/3 of what you had to say about the 3500M’s general performance (sure, you had more pages devoted to mobile-only, but only one page on the subject of general performance).

            Anyway this whole discussion about the example about benchmark presentation doesn’t address the primary concern here: your claims that the 3850 “struggles to keep pace with the Intel Core i3-2100” and that the 3800 will be so much worse that it won’t be able to compete at a >$100 price level, neither of which are borne out by the published benchmarks.

            • Damage
            • 8 years ago

            You seem like a passionate guy who’s pretty on top of things. Sorry to hear you’ve focused on a couple of our choices to the exclusion of everything else.

            -Our choice to include desktop CPUs for comparison was a small part of a much larger enterprise, and I’ll admit I was trusting the reader to interpret the data properly when I included so much. We have a history of asking more from our readers than the average publication, and we will probably continue to make those demands and provide that extra info where we can. Many folks like that, even if you don’t.

            -My wording “struggles to keep pace” seems straightforwardly accurate to me. The A8-3850 is clearly slower than the Core i3-2100 in CPU tests. It doesn’t “fail to keep pace” in that it’s not in an entirely different performance class, but it doesn’t breeze through at parity or above, either. And it needs 100W to do it. Shades of meaning there, but you’ve misread my intentions–and really, really fixated.

            Those things seem to have rubbed you the wrong way. I’d encourage you to look past them and realize we’re not out to mislead anybody about Llano’s relative merits and competitiveness–even if our view of it is less rosy than your own.

            • SPOOFE
            • 8 years ago

            I, for one, like that you show mobile performance in context with desktop performance. The more perspective the better.

            • jensend
            • 8 years ago

            I certainly do like the fact that you ask more from your readers than the average publication; I wouldn’t be spending any time here complaining (and, from my point of view, trying to spur you to do even better) if I didn’t think that on the whole you’ve done a better job with tech journalism than just about any site out there (AnandTech is the only generalist site that really competes IMHO). But I don’t think that expecting a lot from your readers exempts you from two of the cardinal rules of good technical writing: the most important information should generally come first, and the size of passages and related visuals should normally be roughly proportional to their importance.

            Though the way I said several things in this thread probably gave a wrong impression, I didn’t mean to be accusing you of having some kind of conscious conspiracy to mislead everyone about Llano. I have felt that you’ve had a bit of a bias against it – probably because you don’t fall into the class of people who would seriously consider an IGP-only setup for your own machines (unless it miraculously provided quality, framerates, and resolutions to match lower-high end discrete cards like the 68xx)- and that this comes out quite clearly in your writing and presentation.

            • JumpingJack
            • 8 years ago

            [quote<]You seem like a passionate guy who's pretty on top of things. Sorry to hear you've focused on a couple of our choices to the exclusion of everything else.[/quote<] Perhaps it is a good time to revisit Cyril's article: [url<]https://techreport.com/discussions.x/21294[/url<] He may be passionate but he probably can't help it. EDIT: Or maybe it's a self-image preservation issue, [url<]http://arstechnica.com/science/news/2011/08/users-treat-criticism-of-favorite-brands-as-threat-to-self-image.ars[/url<] [url<]http://www.sciencedirect.com/science/article/pii/S105774081100057X[/url<]

            • jensend
            • 8 years ago

            Ad hominem anybody?

            I’m not an AMD fanboy- I haven’t actually ever bought an AMD part (though that’ll probably change with the a8-3800), and you’ll often see me defending nV against the AMD crowd in discussions about discrete cards.

            If I’m a fanboy for anything or anyone, I’m a fanboy for competition.

            When one company starts to get a near-monopoly in a market (Intel’s 80% market share definitely qualifies) you’d better believe I’m going to be rooting for the underdog to some extent. Having a competitive market is the only way to keep consumers from getting totally screwed. Whenever the monopolist bothers developing any new technology, they use it as a price-discrimination tactic (where non-crippled versions of the product command a huge premium), making it so all the [url=http://en.wikipedia.org/wiki/Economic_surplus<]economic surplus[/url<] involved in transactions goes to them. When the market becomes competitive again, they're forced to include their "secret sauce" product-differentiation features in their products across the board; the price-discrimination tactics lose power, consumers get a fair share of the economic surplus involved in transactions, and meaningful technological progress actually starts to reach the majority of the market rather than just the wealthy enthusiasts. You see this pattern again and again in every area of tech; as a couple of examples in the CPU space, hyperthreading would never have filtered down to the consumer market and speedstep etc would never have made it anywhere outside of $2000 business laptops if it weren't for bursts of time when AMD broke into the market enough to provide a sufficient threat. But just developing a competitive product isn't enough to break into a near-monopolistic market; it takes a lot of press and a marketing budget beyond most upstarts' capacity to overcome the "information inertia"- people have been used to the idea that Brand X is the brand to buy, and "nobody ever got fired for buying IBM." So though what I feel or say about the matter will have little impact on the bigger picture, I feel like putting a word in when a product that's quite competitive with the monopolist's seems to be summarily dismissed with no justification. (2.2% performance difference? Give me a break!)

            • JumpingJack
            • 8 years ago

            Ok, ok… your not a fanboy, but it’s fair to say you are extremely verbose.

            You do seem very passionate.

            • sweatshopking
            • 8 years ago

            “As it turns out, AMD’s Llano processors aren’t all they were cracked up to be.” from the system guide [i<] is [/i<] a negative spin on the processor. justified or not. i do agree you guys have failed to see the value of this processor, for moms, and sisters, but i also agree that intel does make better cpu's. I don't think the value of this chip has been honestly shown, but i can't really fault you when intel's cpu performance is really that much greater. i'm not complaining, you know we're soulmates, scott.

          • SPOOFE
          • 8 years ago

          [quote<] and still blows the i3-2105 out of the water in integrated gaming tests[/quote<] Which isn't saying much: [i<]Integrated graphics for gaming suck, whether it's the super-sucky Intel or the slightly less sucky AMD[/i<].

            • jensend
            • 8 years ago

            Sure, Llano’s integrated “suck” by comparison if you were planning on spending $100 or more on a graphics card. But the people who do that are a minority among those playing 3d games on their computers (not to speak of the general public). With Llano, just about all modern games are playable at 1400×1050/1600 x 900 or lower resolutions at high quality and 1680×1050 or lower with less aggressive quality settings. A lot of people will be happy to take that as good enough and save $100.

            • SPOOFE
            • 8 years ago

            I just re-checked the reviews, and the numbers do not support your assertions.

            This is all splitting hairs, anyway. The fact remains that AMD’s chip is not as good a CPU as Intel’s. The passage you take umbrage with reflects this fact, and even acknowledges the GPU superiority. You are, in essence, complaining that an accurate statement was accurate, and that’s just plain silly.

            • dragosmp
            • 8 years ago

            Seen from a different perspective TR’s Llano-related recommendations make sense. From your discussion with Damage (btw thanks to both) I see that TR base their recommendation on how fast a CPU is within a certain power envelope. By this rationale any 100W CPU would not be recommended against a 65W CPU if it’s only “as fast” or “same price” – the 100W CPU should be either faster or cheaper. Llano has “features” like fast IGP, OCing headroom, mobos with more features for a lower price, but TR when makes their recommendation by placing as much importance on these as on Nvidia’s Physics or CUDA. The gripe I had is that TR doesn’t make this obvious in reviews, one must read between the lines to see it (for marketing reasons?) and one may just as easy miss this point and think TR is slamming Llano for no obvious reason. In the end it’s not that the phrasing in reviews is unlikable, it’s just interpretable both ways. For example this is clear and probably unprintable way of phrasing “as seen on these scatter plots A8 3850 is @220%, i3 2100 @230% which makes the i3 4% faster than Llano on average for the same price. If we discard Llano’s additional features like fast IGP, OC, mobo features/$ and also discard i3’s QuickSync & AES-NI, then we can only conclude the i3 is a better recommendation as it consumes less to achieve slightly better performance.”

            Guys, an Econobox-like rig based on Phenom II / Athlon II x4, a 70$ AMD870 board w/ USB3 and a 70$ HD5670 is faster in all metrics and about the same price as an A8-3850 + FM1 board. The K10 + discrete matches all the features of Llano, consumes more, but it’s measurably faster for the same cost. I seem to hear an i3 screaming from somewhere that it should be the better choice for enthusiasts and it is upset with Intel that it’s locked at stock speed when it could have easily reached 4GHz if only a 1$ PLL had been built in the SouthBridge.

            • sweatshopking
            • 8 years ago

            to be fair though, the 100w this chip consumes is for a gpu [i<] AND [/i<] a cpu. to reach this performance level, at least, with a 2100, you'll need a discrete gpu, which IS going to put you over 100w. for THIS level of performance, 100w isn't bad.

        • ronch
        • 8 years ago

        He’s just trolling. 🙂

    • abw
    • 8 years ago

    The CPU/CPU difference is less than the GPU/GPU one , so no,
    the i2100 , doesnt have to struggle to keep pace in GPU perfs ,
    it is simply out of the equation…

      • SPOOFE
      • 8 years ago

      But like it or not, GPU performance is less important than CPU performance.

      EDIT: With regards to this kind of product; you can always add in a graphics card, but there’s no add-on to increase processor cores.

        • abw
        • 8 years ago

        If you cant match it , denigrate it by downplaying its strong point..

        Intel gfx sucks far more compared to radeon than Llano CPU compared to SB2C CPU.

        Like it or not, the A8 has far better value..

          • SPOOFE
          • 8 years ago

          Who’s “you”? Techreport is not a competitor to AMD. And I’m pretty sure I’m not, either.

          The A8 is a better value if and ONLY if you fall in the tiny subgroup of people that doesn’t care about CPU perf AND is perfectly happy with a merely-adequate gaming processor. Non-gamers will be perfectly happy with any current IGP, and gamers are less likely to be happy with anything that’s not a discrete solution. All this, and higher TDP’s to boot.

          A8 is an interesting indicator of things to come, and will most certainly rouse Intel out of their torpor, but it is still a feature that appeals to a small niche, all things considered.

        • BobbinThreadbare
        • 8 years ago

        I disagree for this product. People buying this will not be buying addon cards. Also, both CPU and GPU simply have to hit a baseline, any extra power is not not necessary.

          • SPOOFE
          • 8 years ago

          [quote<]People buying this will not be buying addon cards.[/quote<] I agree, at least with the sentiment; I'm just explaining why I think the number of people that consider it perfect for them will be small.

      • NeelyCam
      • 8 years ago

      As usual, you missed the point. Good thing SPOOFE explained it to you.

        • abw
        • 8 years ago

        You re right, Neely , for sure , i missed the wrong point….

        These CPGPUs will mainly find their way in all in one PCs with no possible GFX upgrade…..

          • SPOOFE
          • 8 years ago

          And in which case, modern graphics-heavy gaming will suck on either platform. You could stick with Torchlight, but then, that game runs amazingly well on Intel’s IGP, too.

            • abw
            • 8 years ago

            I prefer 30% slower CPU with 150%+ faster GPU.

            • SPOOFE
            • 8 years ago

            And that’s a valid preference. But it doesn’t change the fact that AMD’s IGP is “good” only relative to Intel’s, and touting it for “gaming” is still very unimpressive. Its downsides – lower CPU performance and higher wattage – are also significant, and probably major considerations for a much larger number of people than those that care about games; and of those that do care about games, they are much more likely to forego any IGP in favor of a discrete solution, anyway.

            In short, it is a VERY tiny niche that A: won’t care at all about the lower CPU perf and higher heat/’lectricity and B: will be satisfied with the gaming performance of the IGP. Don’t get me wrong, I’m very happy that integrated is improving, but let’s not kid ourselves.

            • abw
            • 8 years ago

            Graphics matter as much if not more than CPU…
            You just dont know how much a GPU can offload a plateform…

            Ask all the guys that have a few years laptop , they will all tell you the same :
            The CPU is still good enough , it s the IGP that sucks and ruin the expected life cycle of the LT.

            Fortunately , this will come to an end for Llano users , but surely not for SB users
            that dont even have DX11 on a so called up to date processor…

            • SPOOFE
            • 8 years ago

            [quote<]Graphics matter as much if not more than CPU...[/quote<] And I completely disagree, and I feel I've adequately outlined why. I feel it is a significant tradeoff that appeals to a relatively small number of people, and that the slower CPU and greater heat are detriments that really need consideration.

            • srg86
            • 8 years ago

            I have an Arrandale based Pentium Dual Core laptop. The graphics run fine for my uses, I just want Aero and Kwin compositing to work. It feels very responsive.

            I wonder if you’ve been drinking the Jen-Hsun Huang Koolaid.

            • shank15217
            • 8 years ago

            Yea the whole point of something “good” is that its relative to something else. You’re saying CPU performance matters and abw is saying GPU performance is more important relatively. Truthfully, AMDs cpu can run all the software Intel’s can but Intel’s GPU is missing several features of a modern GPU making it not only slower but far more limited. I am tired of hearing people complain about AMDs CPU. This horse has been beaten to death, the architecture is now two generations behind Intel’s.

            • srg86
            • 8 years ago

            That’s up to you, I would personally prefer 30% slower GPU with 150% (or whatever it is) faster CPU. My uses aren’t GPU intensive, outside of just wanting desktop graphics to run smoothly. Especially on a laptop

            • abw
            • 8 years ago

            The gpu is 60% slower while the cpu is 20% faster…
            Do you stll prefer it?……..

            • thefumigator
            • 8 years ago

            (almost) all games run on intel GPU, the only problem I see is that graphics quality is really bad, half of them are unstable, or have graphics glitches.

            On the Radeon, games runs much better, not only in performance terms, but overall image quality is greater too.

    • burntham77
    • 8 years ago

    If this translates into some solid AMD-based gaming laptop options, then I am all in.

    • Goty
    • 8 years ago

    The only Llano variant I’m remotely interested in it the A8-3800, and that’s only if AMD has made some improvements in their Llano UVD implementation since launch.

    • Joe Miller
    • 8 years ago

    This is all good, but I still wait for Llano laptop availability where I live. There are none available yet, and I want to buy one.
    Desktops are not very interesting for me anymore.

      • riviera74
      • 8 years ago

      Drop by [url=http://www.newegg.com/Product/ProductList.aspx?Submit=ENE&N=100006740%20600165139&IsNodeId=1&name=AMD%20A-Series<]here[/url<].

      • NeelyCam
      • 8 years ago

      [b<]PAPER LAUNCH![/b<] GloFo 32nm is so solid it makes me wanna cry... 2 years behind Intel, and the gap is widening.

Pin It on Pinterest

Share This