New AMD Demo Shows Llano Outperforming Sandy Bridge - HotHardware
New AMD Demo Shows Llano Outperforming Sandy Bridge

New AMD Demo Shows Llano Outperforming Sandy Bridge

Much of the talk about AMD products has centered around Bulldozer of late, but Llano is on track for launch this year as well. AMD has released a new video pitting Llano against Intel's Sandy Bridge, with results that (un)surprisingly favor AMD's own solution.

According to Godfrey Cheng, AMD's director of Client Technology, Llano was designed on the following hypothesis:  "People are using more modern workloads like 3D graphics, HD video and Internet surfing in a much more prevalent manner...we all dabble with spreadsheets and word processing...but any modern x86 CPU-based PC can handle these workloads with ease.  But with these modern applications, the capacity to multitask, improve image quality and enhance power efficiency are much more important than raw x86 performance in determining how good a consumer’s experience is with a particular PC."

He writes further: "AMD’s “Llano” and Intel’s “Sandy Bridge” are roughly equal in size and transistor count...But...An analysis of the two components’ die area shows that AMD has invested much more heavily in graphics, parallel compute and video whereas Intel has invested much more of its silicon area in improving classic x86 performance."  That's where this video comes in.


We can't speak to whether or not the two systems were comparably configured and the representativeness of AMD's chosen workload is, as always, open to debate. One thing we did note, however, is an interesting difference in the execution speed of the Microsoft Excel benchmark. (This may be the first time the words "interesting" and "Excel benchmark" have ever been used in the same sentence without the inclusion of a "not" --Ed). We calculated our results by taking multiple screenshots per second and comparing both the displayed graphs and the distribution of color in the data columns linked to those graphs. This last is essential—some of the early graphs are very similar.

When the Excel benchmark 'starts' at 1:07, the Intel 2630QM is running roughly two seconds behind AMD's 3510MX. By the time the additional video workload is introduced at 1:18, that gap has shrunk to just a fraction of a second. This implies that we'd see Intel overtake AMD were the test left to run a few seconds longer and that Intel's investment in traditional x86 areas does pay off in those types of computation tests.

The really interesting bit, however, is that once the CyberLink video starts at 1:19, the Intel Excel window redisplays the same graph it displayed at 1:09 and does nothing thereafter. From 1:19 - 1:25, the Intel Excel window doesn't update, while the AMD system goes on drawing a set of graphs we've never seen before. When we see the Excel graphs again at 1:41, the Intel graph is the same as the AMD graph from 1:12. When AMD adds a 3D rendering component, the Intel side of the Excel graph freezes even longer (10-11 seconds).


Our Llano reference test platform from late last year.

It's entirely possible that these discrepancies were caused by Sandy Bridge-related driver issues, a specific application performance issue, or a bit of visual skullduggery from AMD (the video sequences could have been timed to show the worst side of Sandy Bridge.)  If none of these issues exist, it suggests that Sandy Bridge runs into trouble fairly quickly when handling CPU and GPU workloads simultaneously.

The question of which CPU is faster may come down to home much multithreading 'typical' users really need. We saw good reason to suspect that Sandy Bridge would've caught Llano in Excel when juggling the 3D demo—we'd be very curious to see this test repeated with 720P-1080P flash content standing in for CyberLink PowerDVD. Much of the  video users watch these days, including YouTube, Hulu, and most news sites, is played back using Flash inside a browser environment, not via disc-based media.
0
+ -

Hmm, I7 is at 2Ghz, What about llano? This is great for AMD but honestly it would be nice to see exactly what kind of cpu performance we're looking at on the AMD side.

Aside from that, this means decent gaming performance at the price of a CPU! No discrete GPU needed.... oh boy the future looks bright.

0
+ -

You know, even if the performance is under 20% of the equilvalent intel.... These still look pretty good..... Simply because at the same pricepoint, you get more cores.... sure the clock speeds arent THAT fast and raw performance isnt that good... but for the average user like me who doesnt game, and video edits once in a while....i'd take more cores and threads over raw power.... I'm slowly coming back to AMD now..... cant wait to some real world benchmarks and not these skewed ones.

I hope HH gets some of these cpu's sometime soon.... a late valentines gift from AMD hahha

0
+ -

Coolice,

Single periods rock. They really do. Try them sometime. ;)

Pwr,

We don't know. In the past AMD has indicated that Llano would launch above 3GHz but the company has never specifically said that all Llano chips would be 3GHz+. We don't know if AMD compared equally-priced processors (relative to its own SKU plans). The 2630 QM is a low-frequency, HT-enabled chip—we don't know if HT was enabled (or if it even helps). HT doesn't *always* improve performance; this type of scenario could actually hurt total platform numbers.

0
+ -

hahaha, sorry Joel! I wrote that comment literally minutes before my exam today, and i was jumping from slide to slide (ppt) while i was on this website posting a comment. lol.

Interpret the ... as 2-3 second pauses i guess?

sorry for using one "...." in this comment!

i'll do me best to use them less frequently (lies, i tell ya)

0
+ -

Sounds interesting, and the greatest thing I take from this is it is not even Bulldozer, but Llano. The reason I say that is because Bulldozer is more of a performance chip than Llano is. So if this CPU/GPU combination does better then what will a Bulldozer do even with a relatively inexpensive series 5 or 6 GPU thrown in as a singular device rather than a combo!

0
+ -

Correct me if I'm wrong, but it sounds like Llano is more geared towards the typical desktop user's experience, and Sandy Bridge is more geared toward workstations.

@Coolice: The ellipsis (...) is a recognized part of speech, and used in written form, adds a conversational tone to your missive. So pooh on the critics!

0
+ -

Christopher,

(From Wikipedia): An ellipsis can also be used to indicate a pause in speech, an unfinished thought, or, at the end of a sentence, a trailing off into silence (aposiopesis) (apostrophe and ellipsis mixed). When placed at the end of a sentence, the ellipsis can also inspire a feeling of melancholy longing. The ellipsis calls for a slight pause in speech.

Apply that to what Coolice wrote (I hear you Coolice--I'm just using this as an example, not calling you out.) It's one thing to use a single ellipses to convey a pause or an emotion, as mentioned in the definition above. You end up with a very long statement with no clear breaks, interludes, or complete thoughts. I've had a few professors in my day that actually talked as the above; it's absolutely maddening to try to track what's important vs. what isn't.

Although being from KY, I suppose someone might write all the ellipses out to indicate when they're pausing to spit tobacco. ;)

0
+ -

Gents, thank you for defending and correcting me... (see what i just did there)

Its fascinating the things you learn apart from just electronics online! and Meh, its all cool calling me out and what not. Its how we all learn!

+1
+ -

with APUs going how they are won't greater graphics power mean greater number crunching? I don't know maybe I'm crazy...but my vision of the future is graphics on chip, with discreet cards only being for "hardcore" gamers. At this point I *expect* transparent openCL everywhere applicable, so the winning chip will ultimately be the one with better graphics.

0
+ -

CFeldman: Not really, at least not for the foreseeable future (which I'll define as five years). I spent some time thinking about how best to explain this today, so hopefully the following analogy will be helpful.

Imagine that you own a lumber processing facility equipped with four heavy-duty saws that can process lumber of any size, up to and including things like redwoods. (We're ignoring Hyper-Threading here). These mills can process smaller trees, of course, but that's really not the best way to use them—ideally you want to keep them focused on the really big stuff.

In order to do this, you construct an array of smaller mills. These facilities can't handle the really big logs; they're built to process smaller, more common trees. Here's the catch: The big mills (aka your four conventional x86 cores) can process smaller logs faster than the small mills--but if all four of the big blades are busy cutting small logs when a bundle of big logs arrives, there's nothing to do but wait until they're finished.

Here's how this maps to your question: The cores inside a GPU are much simpler than the x86 cores they're linked too and can therefore only handle certain kinds of work with any degree of efficiency. It's still essential to improve the performance and speed of the primary x86 cores, especially since the operating system that assigns and handles workload allocation is running on the CPU.

If you're dealing with workloads that can only be handled effectively by the traditional x86 cores, it doesn't matter how many GPU processors are on-die. The fact that GPUs are now programmable isn't going to change this.

If I had to guess, I'd bet that the GPU cores of five years hence will *not* overly resemble modern CPU cores. The gap will shrink to some degree on both sides, but there are some tasks that don't translate well from one to the other. Unless/until we successfully reinvent the way we think about programming, that's not going to change.

0
+ -

So "to home" is where the heart is? As in CPU???? Grammah is fun!

The question of which CPU is faster may come down to home much multithreading 'typical' users really need.

0
+ -

The issue is, putting a mid-range Nvidia GPU in the Intel machine will just blow the Llano system away. But that was not quite the point here. AMD seems better balanced in complex user scenarios. I would like to see better power usage on the AMD part considering they have had plenty of time to solve that problem. We shall see where BullDozer can so soon.

0
+ -

@fteoOpty64 At what point did Nvidia get thrown into the conversation? As for power usage emmm, didn't they just show they use considerably less power than the competition, whilst providing smoother frames.

@CFeldmanHaim Couldn't agree more! When people say it will take years to develop. Take a loot at Apple emmm already here. Companies are fast coming round to the idea and i think with the introduction of AMD's APU's and also software development support from AMD this will be more like in the next year or so.

This is just the beginning, in 1Q 2012 AMD will introduce next gen Bulldozer APU's, with AMD HD 7000 series GPUs. I really do think that this is the most significant leap in the x86 platform. The future is bright for the PC market.

Login or Register to Comment
Post a Comment
Username:   Password: