Tests cited by MS prove flaws in Linux study – Linux Today

MS says PC Week tests support Mindcraft, but the numbers undermine Mindcraft

An article published earlier today by Linux Today fires back at Microsoft's challenge to Linux. Author Larry J Blunk points out that the PC Week benchmark cited by Microsoft in its challenge in fact supports the Linux community's objections to the earlier Mindcraft study. Mindcraft had been employed by Microsoft to conduct a test, and the results showed NT outperforming Linux. This immediately drew accusations that the tests hadn't been set up properly, and didn't provide an accurate representation of Linux's performance. Blunk argues that the PC Week numbers support these accusations. "PC WeekPC Week article, we find that the NT server achieved a meagre 150Mbps of throughput when tested against WinNT clients. In others words, the Samba/Linux combo outperformed Windows NT server by a very healthy 31 per cent when tested against Microsoft's self-proclaimed business class desktop product. PC Week goes on to state, 'More importantly, Samba had minimal performance degradation at higher client loads. In tests with 60 clients, Windows NT managed only 110M-bps throughput compared with 183Mbps for Samba.'" The key point Blunk makes is that these numbers effectively blow Microsoft's demand that the Linux community "withdraw their criticisms of the initial Mindcraft report" out of the water. The issue is clouded somewhat by the fact that the ZD benchmarks cited by Microsoft in its challenge are favourable to NT, but Microsoft is clearly going a FUD too far by implying that they agree with the Mindcraft test - they don't. That however doesn't make the challenge go away. The newer and (so far) undisputed test results must give Microsoft some confidence that if its challenge is taken up, NT can be made to come out on top. But note how cunningly the spinmeisters have folded two separate issues into one, blurring the join as they went. Oh, and a thought about level playing field tests set up by Microsoft. It doesn't take that long a memory to remember Jim Allchin plus supporting video crew and assembled attorneys trying and failing to claw his way out of the mire. That test strongly suggested that Microsoft was utterly incapable of differentiating between impartial tests and demos, so maybe the answer this time around is to just let Microsoft set things up. ® Related Stories:MS declares war on Linux Can Linux avoid MS NT trap? MS marketing spins and respins

Sponsored: HPC in the cloud: A solution for varied computing needs