This is the first time EVER that I've read that Intel Platform uses the Socket 940!

What the HELL are they smoking?

Not to mention that the entire article only has ONE benchmark that actually compares the difference between dual Opterons (single core) vs. one dual-core single Opteron. (Sandra memory). Everything else is all CPU computationally dependent, which make very little difference because the cores are clocked at the same speed, and the FLOP counts are nearly identical.

(reference: http://www.tomshardware.com/cpu/20051107/index.html)

Can't believe that they a) post this crap, and b) let this crap through!?

They also had another article that was supposed to compare SATA to SCSI RAID, yet they completely leave out the SCSI RAID stuff and do not even bother to explain what the hell the graphs were.

Are they comparing to iSCSI? Ultra160 RAID? Ultra320 RAID? SAS? ??? And then they said:

Quote: Originally Posted by THG Benchmark Results Due to the lack of a serious benchmarking utility that would properly support either the large array size of 2 TB or RAID arrays in general, we decided to primarily focus on I/O performance

well...freak - if you can't test it, then why are you writing about it? And have they not heard of a program from Atto Technologies called ExpressPro Tools (or at least it used to be called that) which can be used to test bandwidth/throughput??? THG - GET A CLUE!

(update: the new benchmarking tool is called "Windows HBA Utilities").

Attached Thumbnails

 

Quote: Originally Posted by alpha754293 This is the first time EVER that I've read that Intel Platform uses the Socket 940! What the HELL are they smoking? Not to mention that the entire article only has ONE benchmark that actually compares the difference between dual Opterons (single core) vs. one dual-core single Opteron. (Sandra memory). Everything else is all CPU computationally dependent, which make very little difference because the cores are clocked at the same speed, and the FLOP counts are nearly identical. (reference: http://www.tomshardware.com/cpu/20051107/index.html) Can't believe that they a) post this crap, and b) let this crap through!? They also had another article that was supposed to compare SATA to SCSI RAID, yet they completely leave out the SCSI RAID stuff and do not even bother to explain what the hell the graphs were. Are they comparing to iSCSI? Ultra160 RAID? Ultra320 RAID? SAS? ??? And then they said: well...freak - if you can't test it, then why are you writing about it? And have they not heard of a program from Atto Technologies called ExpressPro Tools (or at least it used to be called that) which can be used to test bandwidth/throughput??? THG - GET A CLUE! (update: the new benchmarking tool is called "Windows HBA Utilities").

im confused... the article says intel uses socket 604 ( same one xeon uses...)

never mind i cant read... i was looking at the specs earlier in the article

so what motherboard did they use to test the intel platform on?

Quote: Originally Posted by THRiLL KiLL im confused... the article says intel uses socket 604 ( same one xeon uses...)

EXACTLY!!!! LOL....AND that in the benchmark results - there's NOTHING for the Intel stuff!!! Hence, my point exactly!

Must be a "Friday" job.

WOW....
Speechless...

okay even the reivew is wrong.....

correct me if i am wrong,

800 series (842,846,850) are for 8 way opteron supported motherboards

400 series (442,446,450) are for quad opteron supported motherboards

200 series (242,246,250) are for dual opteron supported motherboards

100 series (142,146,150) are for single opteron supported motherboards

so why are they comparing 2 processors from the same series?

...Im just wondering why he isnt using firefox or mozilla...

but yea, dont trust THG for that kinda crap O.o they lie..Ive seen benchmarks of the X2 where the pentium came out ahead wtf?

Beating a dead horse...

Quote: Originally Posted by lAnonymousl Beating a dead horse...

Ditto that....

I haven't been on THG in ages. Good to know that i haven't missed much. Sent out four flamers this morning though.

I'm not using Mozilla/Firefox because it's not supported by the school's IT department *rolls eyes*. (i.e. I am not the admin on that system)

All the more reason for me to get a craptop.

I read that article when it came out - errmm...as far as I can remember ( not going back to read it again), the intention was to compare 2 single core hammers and one dual core Italy to see if a dual core cpu would be physically faster than having 2 separate cores.

Yes there is a part in that article suggesting an Intel comparison - maybe its just there for comparison? I agree THG are like that sometimes, and you have to know the benchmark programs to explain the graphs yourself - THG are going the way of ArsTechnica - no more Mr Easy-To-Understand-Explanations.

Mind you - they were right about Socket 423...contrary to common belief, but I have to agree in some cases there is a somewhat confussing method of testing.