Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Rambus (RMBS) - Eagle or Penguin -- Ignore unavailable to you. Want to Upgrade?

To: Tenchusatsu who wrote (31363)	10/1/1999 3:48:00 PM
From: kash johal	Read Replies (1) \| Respond to of 93625

Tench, Pretty interesting benchmarks. I noted that you called out the poor performance of the Micron system in the concurrency tests. the key here is that when there was no concurrency the Micron system had a much slower download time than the intel systems. Do you possibly think that this may mean that there was an issue with the ethernet card on the millenia system. Clearly if this is the case then this invalidates the entire conclusion of VC-SDRAM vs RDRAM wrt to the concurrency tests. regards, Kash

To: Tenchusatsu who wrote (31363)	10/1/1999 4:15:00 PM
From: Dan3	Read Replies (1) \| Respond to of 93625

Re: PC Magazine compares 820 and RDRAM vs. VIA and PC133 VC-SDRAM... Interesting test. It would have been more interesting if they had used a 600B instead. The test seems to indicate more of a problem with a network or network DMA driver than a chipset constraint. The micron outperforms the others in disk i/o, and the Maxtor 6800 has an average sustained data streaming rate of 512 * 380 * 7200 / 60 = 23.3 MegaBytes/Sec or something better than 200 mb/sec. Well above any sustained data transfer rate that you can get from a 10/100 network card using an NT, Linux, Novell etc. I took my 380 as a typical sectors/track figure for a system with 262 to 462 sectors/track (you need to weight a little higher than average number since a greater percentage than 50 of the sectors are on the longer tracks) In other words, on the disk i/o test, which is a more significant test of streaming data rates, the VC133 outperformed the other systems. In the network i/o test, which is constrained by the network, network card drivers, and network card DMA drivers, the rambus did better. this was not a very good showing for rambus, but there could easily be a driver problem here, as well as with the network card. PC magazine has a way of letting the mainstream system look good if that system is fairly close. I dropped my subscription after having had it for many years because I felt they had become more of a press release echo than an independent magazine. I would like to see what these systems could do with a 133MHZ memory bus CPU. Dan

To: Tenchusatsu who wrote (31363)	10/1/1999 4:35:00 PM
From: dumbmoney	Read Replies (2) \| Respond to of 93625

The 820 reference system produced a 7 percent better Concurrent Download and Winstone test result than the BX system. Undoubtedly because of a lower cache hit rate (more software running). A 100mBIT/s PCI stream doesn't exactly stress the system. Despite its speedy hard disk, the Micron couldn't match the 820 reference machine on this test, a sign that the Via is less adept at handling concurrent data streams. So much for VC-SDRAM being able to handle concurrent data streams as well as RDRAM. I'm not sure if this is a VIA chipset problem or just a fundamental limitation of VC-SDRAM. But one thing's for sure: This article isn't a very glowing endorsement of PC133 VC-SDRAM's performance. Get real. The machines used different hard drives, among other things. It was a system performance test, not a Rambus vs. SDRAM test.