AMD Quad Core Notebooks showing up?

BN:

 

I hope the black edition makes it into some good gaming laptops. I'd like the i7-920xm, but the price is exorbitant.

 

Megatasking?
Nice word. haha

Cant wait to see actual reviews.

 

i still doubt its going to be near a i7-720qm unless overclocked a lot... the new hexa cores from AMD are already being beaten by i5 quads and i7 desktop quads... i don't expect much more from mobile... however , we just have to wait and see and so far not looking good for AMD.. lack of hyperthreading is really killing the processors..

 

the phenom II was never made to compete against any of the i7s.ever. As for the x6 being beaten by i5s and i7s, in most benchmarks ive seen it only trails spintels big 2.Also AMDs lack of hyperthreading is fine. the average PC user will never use/need more than 4 threads at this point in time.which is alot of people.

 

What does "increasing preference with clear usage model up-sell path" even mean?
I don't really like the whole AMD Vision idea, though; I think I'm the one in the best position to decide what suits my needs. Still, I guess plenty of people aren't in the same position, so it makes sense to target them with this kind of marketing.

 

any one find a review yet?

 

here is a review of the acer aspire 5553

 

yah I saw that one thanks! but I was looking for one that actually addressed heat and battery life.

 

Looks like in cinebench the AMD 930n only reaches i5-430, not too great but at least usable.........

 

That's 32-bit Cinebench though, right? 64-bit seems like a better benchmark to me.

 

64-bit programs tend to pressure the cache more than their 32-bit counterparts so I suspect a transition to 64-bit would hurt these low-cache Phenoms IIs.

 

Meh. Synthetic benchmarks have always favored Intel if you ask me.

 

I think benchmarks always favor the winner? That is just me though. We still need more than one benchmark... to make a fair comparison.

 

intel always has and always will be a shady company so it wouldnt surprise me if they gave millions for a intel slant in benchmarks.

 

I don't really know, this is similar to what you're were saying, but I think it was because the benchmark writers enable certain code paths for certain processor architectures (ie AMD processors cannot use Intel's new AES instructions). I don't think it was necessarily bribery or other shady practices, just optimization bias for the big competitors.

 

It's sloppy coding and it can be even worse in applications for which it's quite possible that the developers only ever looked at one path (the benchmark creators at least have to try every brand to see that there is no obvious silliness). I once had to run a series of tasks each of which was taking 20 minutes on my work desktop (2.8GHz quad-core Opteron). There is no way those tasks should have taken that long, so I duplicated the setup on my laptop (2.5GHz Core2Duo) and each task took around 3.5 minutes. Sometimes people forget how different the processors can be.

That said, I don't think there is anything fishy going on with Cinebench. If there was, we'd see it in the reviews of the desktop Phenom IIs compared to the desktop Core i#s. AMD just ran into the thermal wall of what's possible at 45nm and had to cut both clock speeds and caches to the point where performance is fairly lousy.

 

I'd like to point out that is isn't necessary sloppy coding... Intel's C/C++ compiler uses a CPU dispatcher that does a vendor ID check. This means that even if the processor supports the underlying architecture extensions (lets say SSE3) it will not use that code path unless the vendor ID matches "GenuineIntel". Developers who had paid for Intel's compiler have been complaining for years, and I still don't know if Intel fixed it. So yes, if you want a fair benchmark, make sure it was compiled with gcc and is open source.

 

Surely nobody would be stupid enough to use Intel's (or AMD's, for that matter) compiler for benchmarks, right? That's just asking for problems. Although this might explain some of the application issues (not the one I described above -- for that one, everything was compiled with gcc).

 

any new benchmark sightings?

 

Nothing yet.

Cant wait to see a cheap AMD Phenom x3 with ATI Mobility HD 5670.
It would be the perfect fit for my little brother as he only plays Garry's Mod and other Source based games.
Maybe also some C&C. xD

 

Found a Passmark result for Phenom X3 N830




N830 is better than Core i3-350m and some top-of-the-line Core2Duo mobile.

Interesting result... ^_^

 

About what I expected. Competes with the i3 and low end i5 but is slightly cheaper.

 

How many extra AMD Cores does it take to beat Intel Cores: 3-2

 

That's beside the point considering they have comparable tdps.

 

Looks like 3 physical AMD cores = 2 physical + 2 virtual Intel cores.

 

Not quite. The difference is that in applications that are not heavily multi-threaded (i.e. the overwhelming majority what is currently on the market), the Core iX stuff will annihilate the mobile Phenom IIs, winning by over 50%.

More on topic: where are the laptops with these things? There's the one used in that Bulgarian review and presumably whatever went into this Passmark, but I have not seen the pricing. Weren't they supposed to be out by the end of May?

 

50% in what? SuperPI

That is too exaggerated IMHO.

 

Now looky here... i5-540m eating dust ^_^

 

Saw initial review of Acer Aspire-5625g-p924g50mi

Featuring Phenom II X4 P920 (1.6Ghz) and ATI Switchable graphics

RETERA.RU - ????? ??????????????? ???????? ACER Aspire 5625G-P924G50Mi


Cinebench R10 (Render Multi)

Whetstone

3DMark 2006 (CPU)

wPrime 32M

wPrime 1024M

WinRar

BatteryEater (classical)

BatteryEater (reader)

 

More N930 and P920 results...


Test results Cinebench R10 (Render Multi)

Test results Whetstone

Results of tests in 3DMark 2006 (CPU)

Test results wPrime 32M

Test results wPrime 1024M

Test results in BatteryEater (classical)

Test results in BatteryEater (reader)

 

i shrug my head.. i'm not impressed at all.. even my old T9400 is up there with the AMD quads.. its like a joke to be frank.

 

The T9400 is up there with the i3's and i5's as well, which is what the AMD quads are competing against. No one is claiming that they are competing with the i7's and Intel quad cores. For example, you could get a dv6 with N930 for $30 cheaper than one with a 430M, all other specs equal. Seeing as, according to these benchmarks, they will perform roughly equally then why not do it?

 

It all comes down to price vs performance.

When comparing the HP dv6t and dv6z here are the current price listings on this date.

dv6t base CPU i5-430 priced at $849
+$50 upgrade to i5-450
+$100 upgrade to i5-520
+$150 upgrade to i5-540
+$300 upgrade to i7-720
+$350 upgrade to i7-740
+$550 upgrade to i7-820
+$600 upgrade to i7-840

dv6z base CPU Quad P20 priced at $799
+$20 upgrade to Quad N930

So performance wise, the N930 is around the i5-520 to i5-540 in benchmarks for $80 to $120 less in price. Not too bad.

Also take into account that the i7-720 is around 20%to 30% better in performance for about 15 times the "upgrade" cost of an AMD N930.

Benchmarks are nice but they don't show the full picture.

 

15 times the cost? That seems like a mistake

 

True. I meant the upgrade cost.

On the dv6z the cost to upgrade from the base P920 to a N930 CPU is $20.

On the dv6t the cost to upgrade from the base i5-430 to a i7-720 is $300.

$20 x 15 = $300

 

Intel's upgrade costs are clearly very poor compared to AMD's. I guess that's because Intel can get away with crazy prices for their higher-end parts due to lack of competition.

As long as the N930 still has decent performance when using only one or two cores, and can compete with Intel on battery life, this is exactly what we needed to see. Hopefully Intel will respond by lowering prices on many of their Core i5s, though I guess the i7s will continue to be overpriced.

 

i'm not impressed.. if a so called quad core can't beat a 2 year old processor , IMO it is a joke and not worth it.. i can't say much for AMD really... hopefully they improve but after all the hype.. its ended like italy and france in world cup.

 

Once again, you're putting too much stock by the term "quad-core". What matters is price, power consumption and performance, and AMD seems to be competitive with Intel when you look at all of them together.

 

I hope you realize that the core i's crap out against some Core 2 Duos from "2 years ago" as well right?

Nobody's denying that for pure performance, AMD's CPUs are subpar compared to Intel, but the average Joe doesn't really need a quad core. Hell a dual core is already overkill for 50% of the computing population. Office and internet isn't going to need a dual core before years from now let alone a quad. Anyways, point is, pure performance doesn't matter to the average schmuck. However, price, heat and power consumption do as lackofcheese mentioned. If AMD managed to fix the last 2 aspects(seeing as they more or less already had the first one covered), then they can still play competitive. They're not really aiming to uproot Intel from the "most powerful CPU" throne anyways.

 

Seriously. The higher end C2D's are still just as powerful as the i3s and i5s. As melody and cheese have said, it is not about performance alone. AMD competes because they offer good performance, more than most people need, for a much lower price than Intel just like Rustican showed. What possible reason would you pay $100+ more for an i5-520/540M when you could get the same performance in the N930?

 

+1 melody and abaddon

in addition, the important thing is that amd cpu appears in some new, well-made and premium laptops while all NBR-er know that dont judge a notebook through its spec only

 

take an example of improvements from AMD:
@45nm, the N620 + HD4250 can beat the p8600 + 4500MHD in both CPU performance, graphic and battery life

 

Even if that is true (and I would like to see some benchmarks as the P8600 is a 25W part while the N620 is 35W), the P8600 is more than 2 years old. That's a full cycle of die shrink + architecture change so it would hardly be surprising if the 2 year old hardware is inferior in all respects.

 

A better comparison would be the i3-330/350M, which the N620 is equal to, but at a more expensive price.

 

Actually P8600's close competitor is Turion II Ultra M620 (Caspian processor of Tigris Platform which was released around Sep 2009). Or the
current Turion II N530.

But even Turion II Ultra M600 can beat the hell out of P8600.

Here is your P8600

Geekbench Score 2699

Here is the result of Turion II M600 (2.4Ghz also)

Geekbench Score 2965

 

Try watch this in dual core cpu
YouTube - DSLR film-look (Canon EOS 550D/Rebel T2i)

 

Okay. What are you getting at?

 

Plays fine on my P9600...........

 

Works 100% with a P7450... While streaming music+a video (low res) in background.