• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Do other Tbred B behave like the 1700+ DLT3C at 1.5V ?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

hitechjb1

Senior Member
Joined
Feb 2, 2003
I would like to see how a Tbred B other than a 1700/1800+ DLT3C behave at Vcore 1.5 V, 1.6V.

Can a Tbred B 2100+ or above run stable at 1.5V Vcore at 1.5 GHz ? 1.467 GHz is the rated freq for 1700+ at 1.5V.

For Tbred B, the rated Vcore for 2100/2200+ is 1.60V and the higher ones are 1.65 V.

What is the stable clock a 2100+ or above can run at each of the following Vcore:
1.5V
1.6V
1.7V
1.8V

I have an estimate for Tbred B 1700+ DLT3C, between 1.5V to 1.75V, it is about 900 MHz per 150 mV, or 150 MHz per 25 mV Vcore. For less oc'able one, probably is around 750 MHz per 250 mV, or 75 MHz per 25 mV Vcore. It begins to slow down somewhere between 2.2 - 2.4 GHz, depends on individual chips. And different chip may have slightly different numbers.

If most 2100+ or above can run at 1.5 GHz at 1.5V, then we can say that the other Tbred B scale similarly (in terms of clock/Vcore) like the Tbred B 1700/1800 DLT3C. And the Tbred B 1700/1800 have nothing special in their intrinsic silicon properties and variations than the other siblings in the same 0.13 u process.


Can anyone who has a Tbred B 2100+ or above try and post the results?

If possible, pls obtain the stable clock at more than one Vcore, perferably at 1.5, 1.6, 1.7 and 1.8V.
 
I think right now, 1500mhz on 1.5vcore is really really close to the limit of the .13 micron manufacturing prosess. I think that is why amd has more potential because the amd's get more work done per clock cycly meaning they can have slower clock cycles than the competion and have an advantage because when the competion is maxed out the amd can still go farther PR wise.
 
well, one guy around here had a 1700 1.5vcore at like 1.06v or some rediculous number like that, so ay least some of the chips dont need anywhere near 1.5v to run at stock speeds. id test for ya but i have a volt modded board so the lowest vcore i get is 1.64. but my 2100 soes 2150 stable at that vcore and did 2350 stable at 1.74 vcore
 
What you refer to is the number of instructions executed per clock cycle (or called the IPC), which is not my question about whether the Tbred B XP 2100+ or above have similar silicon behaviour as the Tbred B 1.5V DLT3C.

But anyway, for IPC, ... it measures the amout of instructions using the same benchmark programs to compare CPU at different clock frequency and also to compare different CPU.

For example, if one runs a benchmark for integer calculation such as Dhrystone MIPS (million instructions per sec) using Sandra, that number is pretty constant, around 3.76 instruction/clock over the entire TBred B CPU (including XP 1700+ DLT3C, XP 2100+, even Barton w/ different size of L2).

It is a number of merit for the CPU normalized over a range of clock frequency, e.g. for XP from 1.5 to 2.5+ GHz.

Intel P4 has similar behaviour, the IPC for integer calculation measured by Dhryston MIPS using Sandra is around 2.93 instruction/clock over a wide range of clock frequency from 2 to 3 GHz.

The ratio between XP and P4 is 1.28 to 1.

For now, e.g. P4 does less integer instructions per clock, but it makes up the absolute by being able to be clocked faster. Between XP and P4, which side delivers the higher absolute DMIPS depends on the ratio of max clock speed above or below that magic number 1.28 for integer calculation.

One can also compared the IPC, clock frequency for a given CPU price. For this, comparing at $70 CPU between P4 and XP, definitely no other CPU can beat the Tbred B 1.5V 1700+.

For today, for non-overclocking, highest P4 is 3.06 GHz, highest XP is 3000+ (2.167 GHz). 3.06/2.167 = 1.41 > 1.28 which means P4 is faster in absolute sense for manufacters who do not overclock.

For oc, you have to find out what is the normal highest XP and P4 (e.g. on air) and compare the ratio of the clock with the number 1.28.

Similar IPC can be used for floating point (FLOPS), multimedia benchmarks, memory, cache, ...
 
I still would like to find out can a 2100+ or above run at 1.5 V at 1.5 GHz.

Any info and test results.
 
Lithan said:
I don't think theres a tbred b around that won't do 1.5ghz at 1.5v

I would like to see actual confirmation that a 2100+ can run or cannot run at 1.5V 1.5GHz stable. E.g. set the bios Vcore to 1.5V and see whether it can run 3D mark at 1.5 GHz.


Add: this test is to see whether the 2100+ (and above) have similar transistor characteristics as the 1.5V 1700+ (tbred b).
 
Last edited:
hitechjb1 said:


I would like to see actual confirmation that a 2100+ can run or cannot run at 1.5V 1.5GHz stable. E.g. set the bios Vcore to 1.5V and see whether it can run 3D mark at 1.5 GHz.

Just tested my XP2400+ (AIUHB 0247) - it run 1.47GHz stable at 1.4V (minimum for Epox m/b). It can also run 2.4GHz at 2.05V.
 
I have a 2100 pali and it wont undervolt even to 1.7

my 2100 is a real
icon13.gif
 
lol really? what you got, stock amd cooler? what a joke.
i undervolted my 1800 a little bit, from 1.75 to 1.725(wow!!) and it was stable. @ 1.61ghz
 
josi said:


Just tested my XP2400+ (AIUHB 0247) - it run 1.47GHz stable at 1.4V (minimum for Epox m/b). It can also run 2.4GHz at 2.05V.

It is good to know that a 2400+ can run at 1.5V stable. If more posts about 2100+ and higher can run at 1.5V 1.5GHz stable, we may be able to say that the Tbred B 2100+ and 1700+ are basically having similar transistor property, may be just a sorting and marking of the chips for different rated Vcore and frequency, ...
 
Back