• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Brisbane?

Well what did you expect? It isn't like the performance is a secret. we pretty much know how it will perform.

i personally think AMD is foolish to go with these low cache models.

I also think 115 is a good price for what it is worth. I mean it would get beat by the E4300 and E6300 and when they take drops to 133-163 respectively yet still likely 15-20% faster...the price seems right

FOR ME these cheap boxes would only be used to build folding boxes, but since they perform so utterly poorly (with the pathetic L2 cache) compared to C2Ds I wouldn't waste the money.

 
A64s with 512KB of L2 perform almost identically to 1MB L2 A64s. There is no reason to add to the cost of the chip for no performance gain.

AMD needs to light a fire under the k8L engineering team though.
 
more than 512K L2 cache for the AMD chips didn't do a huge amount for them, a little but not huge. AMD chips are not nearly so dependent on cache size as intel chips - what do they have now, up to 4MB or something like that? I'd like to see a C2D perform with 512K cache 😉 At any rate, below 512K cache on AMD's does start to show some difference. 256K isn't too bad a drop but noticeable sometimes, 128K can really be felt.
 
Well that may be because not much takes advantage of cache that small....Larger the cache the more and more programs that we find can load things into the cache. The 4 vs 2 of the C2D is about 3-5% in most apps. I have seen a 3.5ghz c2D E6300 is equal to a E6600@3.33ghz to 3.5ghz in a collection of apps I run.



My Intel 4mb cores are 25% faster then the 2mb cores in folding at home when doing 1 instance of a 149x unit. If I run 2 instances the 4mb cache chip slows 25% to that of the speed of the 2mb core. 2 instances on the 2mb C2D are now twice as slow versus 1 instance...Now the E6600 box doubles the output on the E6400 box eventhough speed of cpu clock and memory clock are the same. AMD lose big in these units. Often times my 2.5ghz AMDs were 2x slower then my 3.2ghz C2Ds....with these units that can load into the cache a certain amount of the data they outperform the AMDs closer to 5x now....

You guys can convince yourself whatever you want. reviews sites have done a poor job at testing this anyways.

i did say "I personally" think the low cache is a mistake. It does help the intel C2Ds share the cache in 1 pool. though my QX6700 has 8mb of cache it is really (2) 4mb cache cores so they wont combine or share them.
 
I'm not trying to convince myself of anything. I just pointed out AMD just doesn't use/need as much cache as intel. Different chip designs after all 🙂
 
Originally posted by: YoungGun21
Was it already released, just silently?

http://www.newegg.com/Product/Product.asp?Item=N82E16819103036

I don't remember anything about an official release...
Click to expand...

These were released back in December. Anandtech did a review here:

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2889

and added a followup here:

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2893

Bottom line: Rev G1 of Brisbane seems to have process-related issues. The L2 cache is also slower than Windsor's L2 cache, which makes Brisbane slower per clock. Rev G2 might overclock well or it might not, but nobody seems to have tested one yet. With the launch of the E4300, it doesn't seem to matter anyway.
 
Originally posted by: SparkyJJO
I'm not trying to convince myself of anything. I just pointed out AMD just doesn't use/need as much cache as intel. Different chip designs after all 🙂
Click to expand...

Well that is incorrect. It is often the app that decides whther to use the cache or not. in the case I laid out above there isn't a big enough block of it for the app to use so it doesn't...hence the AMD becomes a dinosaur in performance....Nice move AMD...I see serious folding farms dumping AMD systems left and right...
 
And the K8 is latency dependent which we could all agree too. Doesn't say much when you think ahead with DD2 that has lower voltages, higher bandwidth but higher latency.
Then DDR3 just down the road with yet more bandwidth and higher latency.

So lets just hope AMD thought things through when they were building their Barcelona. Latency impact, I think, will be the sole factor in judging if its a new processor or just another revision in my opinion. Because I don't think the Barcelona will last long paired against Intel and their "archaic" FSB when DDR3 comes around.

Brisbane has higher L2 cache latency then it's 90nm counterparts. Which is likely the sole cause of the Barcelona to arriving Q3/Q4 other than first quarter of this year.
 
The low cache makes plenty of sense for AMD, it lets them crank out more cheap processors to sell to Dell. They know they have the slower chip right now and aren't making any in roads in our tiny enthusiast slice of the market.
 
Originally posted by: aka1nas
The low cache makes plenty of sense for AMD, it lets them crank out more cheap processors to sell to Dell. They know they have the slower chip right now and aren't making any in roads in our tiny enthusiast slice of the market.
Click to expand...



I agree if they are not trying to market it to the upper end of the desktop market....Cheap budget systems for Dell? i agree....

For me? Not. I wont use any AMD core until the cache gets up or the core is fast enough to compensate it. Right Now i am not encouraged. May be the L3 cache pools may help later with the K8L derivatives. So for the conceivable future my folding farm will stay Intel...
 
Originally posted by: Duvie
Originally posted by: aka1nas
The low cache makes plenty of sense for AMD, it lets them crank out more cheap processors to sell to Dell.
Click to expand...
I agree if they are not trying to market it to the upper end of the desktop market....Cheap budget systems for Dell? i agree....
Click to expand...

Hey, the budget consumer wins. What was the option in the past if you wanted the lowest end advertised Dell? Celeron? Now the big winners are people in the $200-300 price bracket. I'm talking about a whole system, not just a CPU! 😛
 
Originally posted by: Zap
Originally posted by: Duvie
Originally posted by: aka1nas
The low cache makes plenty of sense for AMD, it lets them crank out more cheap processors to sell to Dell.
Click to expand...
I agree if they are not trying to market it to the upper end of the desktop market....Cheap budget systems for Dell? i agree....
Click to expand...

Hey, the budget consumer wins. What was the option in the past if you wanted the lowest end advertised Dell? Celeron? Now the big winners are people in the $200-300 price bracket. I'm talking about a whole system, not just a CPU! 😛
Click to expand...

I can agree with that rationale...

The only problem is their whole line is rather weak at L2 cache. Even the ones that based on pricing are geared for the uber expensive enthusiast market. Now it could be said they have conceded this platform and we will have to wait until the K8L's arrive. talk is similar size L2 cache pools per core but 1 large shared pool of L3...we will have to see how that works... The L3 is undoubtedly slower then L2 Sram cache
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top