P4s are a variable bunch. The weakest are destroyed by Tulatins, the best ones are 6-7x more powerful than Tulatins. That's the weakness of using the same brand name to 4 generations of an architecture.

m1so wrote:

P4s are a variable bunch. The weakest are destroyed by Tulatins, the best ones are 6-7x more powerful than Tulatins. That's the weakness of using the same brand name to 4 generations of an architecture.

I don't have a 3.4+ GHz P4 to play with at the moment, but I do have an A64 3700+ which kicks the snot out of my 2.93GHz P4-516. Yet, I can't even get it to run 6-7x faster than my 1585MHz Tualatin. Not even in SSE2 aware applications where the P3 is stuck using plain old SSE.

Standard Def Steve wrote:

m1so wrote:

P4s are a variable bunch. The weakest are destroyed by Tulatins, the best ones are 6-7x more powerful than Tulatins. That's the weakness of using the same brand name to 4 generations of an architecture.

I don't have a 3.4+ GHz P4 to play with at the moment, but I do have an A64 3700+ which kicks the snot out of my 2.93GHz P4-516. Yet, I can't even get it to run 6-7x faster than my 1585MHz Tualatin. Not even in SSE2 aware applications where the P3 is stuck using plain old SSE.

There is very little that can change my mind about netburst in general and dislike most of the builds in general. The power consumption, board related issues, and almost all the boards were using crappy chipsets. All the same for socket A except for the mobiles.

jwt27 wrote:

Wifi appears to work... somewhat. […]
Show full quote

Wifi appears to work... somewhat.

Dell laptop (10.0.0.157) running mTCP FTPSRV here, connected over WEP encrypted wifi to my sister's laptop (...156), which is connected to the wired network.
On the other side, Pentium 3 (...155) running FTP from X:\ which is a shared folder on my Windows PC (...151) 🤣

The connnection is not very reliable though. Once I load the packet driver, it connects to the other laptop, and I can run one mTCP program. When it's finished it closes the connection and I have to reload the packet driver to connect again. Maybe MS Network Client will keep the connection open.

Good thing is, the device driver and packet driver (2x19kB) both load into UMB without crashing!

swaaye wrote:

I haven't gazed upon any sort of passive matrix notebook in over 10 years. You guys are lucky 🤣 Make some HD YouTube videos for me!

If you really want to see it... I guess I can make a video of some games. My camera is not very HD though.

leileilol wrote:

Who's willing to make a passive matrix LCD shader, complete with fixed native resolution, simulated backlight, dithered 18-bit output, low brightness and crappy scaling? 😁

Please call your psychiatrist.. NOW 🤣

if your running dos idk what to tell you but, if your running 98, Microsoft sold some wifi cards along with the need software to make it work. they don't support newer encryption though.

Standard Def Steve wrote:

... I do have an A64 3700+ which kicks the snot out of my 2.93GHz P4-516. Yet, I can't even get it to run 6-7x faster than my 1585MHz Tualatin. Not even in SSE2 aware applications where the P3 is stuck using plain old SSE.

Could you please pass Lynx 0.6.4 on your A64 3700+ with these settings: Problem size 4031 , Memory 128 , Run 5 times.

It would be interesting to compare GFLOPS with Tuallys and P4s.

Athlon 64 3700+, 2.2GHz, 1MB L2

Thanks! That's ~4.5 times faster than a Tually. For comparison, a non-OCed Preshott 3.4 :

Shit... I need a Prescott.

Jwt27, better disable Hyper-Threading before passing Lynx on your Northwood, HT destroys performance.

It does? I thought the whole point of HT was to improve performance in multi-threaded programs...

Same Preshott system as above with Hyper-Threading enabled:

HT knocks off more than 1 GFLOPS 😵 .

I'll have to try that then. Would this actually increase real-world performance (say, +10% fps in Skyrim), or will I only see a higher number on some benchmarks?

Better test and judge for yourself. But I'd not be surprised if some games, specially ancient ones, run better on your Northwood with HT disabled.

Interesting read at the bottom of this page: http://www.xbitlabs.com/articles/cpu/display/ … 50_7.html#sect0