- Apr 17, 2011
- 9,228
AMD cuts 10% of its workforce
- Thread starter McLovin
- Start date
- Status
AyeAyeCaptain
Level 1
- Feb 24, 2011
- 585
Seems they're retreating after the BullDozer did not do/perform as well as expected, I think this is a bad move to just aim to provide low-end budget stuff as sooner or later they're gonna get forced out completely. They should learn from the mistake and refine their high end stuff in the hopes of converting a good chunk of users of the Intel userbase.
Valentin N
Level 2
- Feb 25, 2011
- 1,314
Well said captin 
Bulldozer is not what people thought it would be and it was so anticipated. If they don't begin to act fast amd will end up bad. Intel is already on to make 22nm cpu and so on. it can be that bulldozer is really made for servers
Bulldozer is not what people thought it would be and it was so anticipated. If they don't begin to act fast amd will end up bad. Intel is already on to make 22nm cpu and so on. it can be that bulldozer is really made for servers
AyeAyeCaptain
Level 1
- Feb 24, 2011
- 585
Totally agree, it's sad to hear really as I for one do not want one company to dicate the price/product without a struggle... no rival/competition is not a good thing at all for the end consumer.
Valentin N
Level 2
- Feb 25, 2011
- 1,314
AyeAyeCaptain said:
And intel doesn't have any reason to lower the prices and move faster to new releases.
Valentin N
Level 2
- Feb 25, 2011
- 1,314
according to AMD bulldozer should give the bulldozer module 180% compared to Hyperthreading that give 114-150% depending on how well-optimized the application is for HT.
Bulldozer is still not to compare to Hyperthreading since you don't have 4cores that can handle two threads per cores. so comparing is't comparing 4cores vs octa. i5 is quadcore with for four threads... which FX8 is most close to in many fields.
I believe the lvl1 cache is way to little (only 16KB per core). This cache has to write the info to lvl2 cache which is big but that won't do much good. I am aware the smaller the cache are the more reduces latency is, which results in better performance and more responsive cpu. I thought that 64KB lvl1 cache was low (the phenom II X-series mostly has 128KB (at least the X6 T1000 series do)).
Bulldozer is still not to compare to Hyperthreading since you don't have 4cores that can handle two threads per cores. so comparing is't comparing 4cores vs octa. i5 is quadcore with for four threads... which FX8 is most close to in many fields.
I believe the lvl1 cache is way to little (only 16KB per core). This cache has to write the info to lvl2 cache which is big but that won't do much good. I am aware the smaller the cache are the more reduces latency is, which results in better performance and more responsive cpu. I thought that 64KB lvl1 cache was low (the phenom II X-series mostly has 128KB (at least the X6 T1000 series do)).
- Status
Similar threads
