tecchannel.de claims to have tested the ClawHammer and has benchmarks from it. Mainly Q3 and some memory throughput benchmarks.
The memory bench results are quite at odds with the established conclusion that initial ClawHammer's used for development purposes came with 1MB L2 cache. Indded, judging by the results the Hammer seems to have inclusive 256K of L2 cache, not even the 512KB many expect the release ClawHammer to contain.
Q3 benchmarks are somewhat better then I would have expected though, quite interesting if true.
The unusual memory throughput data puts the article into much doubt however... but TecChannel.de is historically quite reliable when it comes to claiming benchmarks of pre-release hardware.
Make up your own mind whther you wish to trust it or not... interesting though.
Linky to start of article
Linky to benchmarks
Article is not in english, and the various net-translators seem to make a mess of the article so I've no clue what the article actually says... all I have to judge upon in the benchmark data.
Edit: It seems Adul posted this a minute before I.
Edited Again: Oops, it wasnt Adul to post before I, it was Athlon4All.
Apologies.
The memory bench results are quite at odds with the established conclusion that initial ClawHammer's used for development purposes came with 1MB L2 cache. Indded, judging by the results the Hammer seems to have inclusive 256K of L2 cache, not even the 512KB many expect the release ClawHammer to contain.
Q3 benchmarks are somewhat better then I would have expected though, quite interesting if true.
The unusual memory throughput data puts the article into much doubt however... but TecChannel.de is historically quite reliable when it comes to claiming benchmarks of pre-release hardware.
Make up your own mind whther you wish to trust it or not... interesting though.
Linky to start of article
Linky to benchmarks
Article is not in english, and the various net-translators seem to make a mess of the article so I've no clue what the article actually says... all I have to judge upon in the benchmark data.
Edit: It seems Adul posted this a minute before I.
Edited Again: Oops, it wasnt Adul to post before I, it was Athlon4All.
Apologies.