Turion 64 X2 Released

tsuehpsyde said:
I'll milk you for all you're worth. :D

But yeah, who said an Opteron 165 was $300? They're $277 shipped, so that's like a purchase price of $270, and it has 2x the cache. The pricing isn't terrible, but it's not fantastic either.

Energy efficiency comes at a price. To be more comparable:

Athlon 64x2 3800+ Windsor 303US 85W
Athlon 64x2 3800+ Windsor 323US 65W
Athlon 64x2 3800+ Windsor 364US 35W

Turion 64x2 TL-60 Taylor 354US 35W

Seems about right to me.
 
pxc said:
These are 65nm processors AFAIK.
No the Taylor core is 90nm, the die size is enormous. AMD 65nm fabrication technology is not ready till the end of 2006 at the earliest.

Where they will shrink this core to the Tyler core which is the 65nm revision.
 
Mmmmmm dual core lappy with a 12" screen, 100gb drive, and a gig of memory would own! (I much prefer *very* small and light/portable to large/heavy as I use laptops solely for movies and work type stuff)
 
Hito Bahadur said:
I can't presume that the Turion X2s will be slower than Core Duos, I'll have to see benches before I can decide on that as I think their specs should put them fairly even with their different architectures giving wins to both products on price however I see this lineup:



I know I'm mixing two different types of prices (retail versus 1k lots) but it's the best I have.

So the match-ups are:
T2600 - TL-60 = AMD much lower price
T2400 - TL-56 = Price competitive
T2300 - TL-52/TL-50 = price competitive/Much lower

Seems like AMD is targeting the high performance $$ limited crowd (young enthusiasts) and the budget audience based on their price comparisons..... just my 2 cents

?! Turion X2 has 512KB L2 and the weakest one has 256 KB I can't presume how it would at least march a Core Duo that has 2MB L2 , shared from the highest to the lowest part.Your match-up is very thin air.You should have at least put equal MHz parts on both sides if you tried to show Turion in a good light , because a T2600/2700 would run circles around a 2GHz/512KB Turion.

Core Duo has also lower TDP.

Nope , X2 is a wee cheaper , but also slower and hotter.
 
NulloModo said:
Wow, I am very surprised that the Turion X2 kept up as well as it did, and even managed to best the Yonah in a couple tests. For a budget dual-core laptop, these might be nice chips to look into.

I found this quote funny:

I mean, I guess it is well known that in many Asian societies the elder matriarch runs the homestead, but who knew Asus, Abit, et al had them making decisions regarding sockets? ;)

If ScienceMark is the only thing I would run , I would take the Turion.Otherwise it is a no brainer.
 
savantu said:
?! Turion X2 has 512KB L2 and the weakest one has 256 KB I can't presume how it would at least march a Core Duo that has 2MB L2 , shared from the highest to the lowest part.Your match-up is very thin air.You should have at least put equal MHz parts on both sides if you tried to show Turion in a good light , because a T2600/2700 would run circles around a 2GHz/512KB Turion.

Core Duo has also lower TDP.

Nope , X2 is a wee cheaper , but also slower and hotter.


Okay, you are wrong on one part. Turion X2, except for the lowest end, is 2x512KB Cache vs 1x2MB shared for the Core duo. It has half not 1/4 of the cache.

http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_13909_13911,00.html

edit: for typo
 
Hito Bahadur said:
Okay, you are wrong on one part. Turion X2, except for the lowest end, is 2x512MB Cache vs 1x2MB shared for the Core duo. It has half not 1/4 of the cache.

No need to be pedantic , I know that very well.I assume people have a minimal knowledge base and don't need to be told every last bit ...

Let me be pedantic : It's KB not MB.


;)
 
Not being pendantic. There is a big difference between saing 1 processor is 512KB and the other is 2MB when that is not the case. I think it is a pretty big assumption that most people know that the Core Duo has a shared Cache of 2MB and that the Turion X2's have 2x512KB cache or 1MB combined L2 cache.

If you are going to present your information that way, say that the Turion X2 has 1MB cache and the Core Duo has 2MB cache which is more representative.

If you want to pick on me for typos, please do it offline in a PM.
 
savantu said:
If ScienceMark is the only thing I would run , I would take the Turion.Otherwise it is a no brainer.

I just meant it in that considering a P-M can stay slightly ahead of an equally clocked single core Turion (and lose in a couple places by a little bit), and that the Yonah is noticeably faster clock for clock than the P-M, I would have expected the Core Duo to destroy the Turion X2 a lot more than it did.
 
When Intel drops those Core Duo prices in a little while Core Duos will be an even better value than the Turion X2s.....

FYI, just so you guys know, from here on out, Intel is going to ALWAYS have better price/value than AMD.... it all comes down to the Fab capacity/tech and amount of cash the company has... and Intel is ruining AMD in those two areas

Merom, with similar price points to the Core Duo, lower power consumption, and way higher performance (even tho it has 64 bit, it isn't much of a feature, still) will only make the mobile raping of AMD even worse.....
 
NulloModo said:
I just meant it in that considering a P-M can stay slightly ahead of an equally clocked single core Turion (and lose in a couple places by a little bit), and that the Yonah is noticeably faster clock for clock than the P-M, I would have expected the Core Duo to destroy the Turion X2 a lot more than it did.
something tells me that ddr2 is gonna do more than people have been thinking it will ;)
 
I'd really be interested in testing different memory with the Turions. Specifically, I'm wondering if CAS 3 DDR2 400 would be better than CAS 5 DDR2 667. I'm betting that it would.
 
3 clks * 5ns vs 5 clks * 3ns... ;)

access latency will be similar, with the ddr2-667 winning due to higher overall mhz.. and then it will also have more bandwidth due to higher mhz

i bet you anything the ddr2-667 at 5-5-5 will win over ddr2-400 at 3-3-3
(btw- cas latency has traditionally had a very small effect on K8's performance, it's trcd and trp that influence performance the most)
 
I don't think bandwidth is an issue with the Turion X2s though. They obviously have way more than their predecessors or even that of the current desktop X2s. I would think CAS Latency would have more of an effect on lower cache parts like the Semprons or the low end turion X2. I think this is the first Dual Channel part with 256x2 cache.
 
but what i was trying to point out is that the access latency for the ic itself will be identical in both cases, but actual latency will be lower due to faster routing speeds and higher bandwidth

consider this.. ddr1 at 200mhz, 2-2-2 vs 300mhz at 3-3-3. a similar thing to what you're saying, just 2clks * 5ns vs 3clks * 3.33ns.. both turn out to be 10ns for each timing.
one came in at 51ns, the other at 43ns. guess which was which ;)
 
Back
Top