Athlon64 1066MHz HyperTransport?

ta2

Member
Jan 21, 2004
56
0
0
Does anyone have any idea if the...

AMD Athlon 64 3700+
2.4GHz
512KB Cache
Releaseed Q2 '04
Socket 939

...will have a Hypertransport speed of 1066MHz? Looking at the clock speed, it is possible (2400/533 = 4.5). Since if it will have 1066 Hypertransport the DDR ram fast enough to support it will have extremely high latencies and be very expensive, and DDR2 will be even more expensive.

So, any ideas?



Cheers
 

ta2

Member
Jan 21, 2004
56
0
0
I am pretty sure it is 800MHz right now, since everyone is complaining that nForce3 only has 600Mhz.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
26,129
15,274
136
I think it is 800x2 for FX51 and 800 for the regular Athlon64, and heard that it was going up to 1000 soon. (or 2000 for the FX-51 ??)
 

andreasl

Senior member
Aug 25, 2000
419
0
0
I believe it is 1000MHz (5*200) HT planned for Socket 939 later this year. Probably to ensure that PCI-Express 16x has all the bandwidth it needs.
 

Jeff7181

Lifer
Aug 21, 2002
18,368
11
81
Via and Sis chipsets have 800 Mhz HT links right now... the nForce3 150 has a 600 Mhz HT link right now... the nForce3 250 is supposed to have a 1000 Mhz HT link. The maximum speed the HT link can run at is 1.6 Ghz.
 

JBT

Lifer
Nov 28, 2001
12,094
1
81
I thought 1.6 had something to do with it apparently its the max thanks
 

Spearhawk

Member
Dec 27, 1999
75
0
0
Anyone know of a good article that explains the HyperTransport, how it relates to a regular FSB and what effect it has one memory bandwidth and so on (for example, is HT 800 MHz as good as Intels 800 MHz FSB?).
 

ta2

Member
Jan 21, 2004
56
0
0
So for the new Socket 939 Processors you will need what type of RAM?

PC4000? PC4200?

Extremely slow latencies there, would the faster HyperTransport actually offset this?
 

InlineFive

Diamond Member
Sep 20, 2003
9,599
2
0
You are all CONFUSED!

Each K8 processor has exactly two HyperTransport tunnels running at 800Mhz going to and from the processor. Each channel is 32-Bits wide so the grand total of bandwidth is 6.4GB/s. In this sense though you are correct that the total bandwidth could be considered 1.6GHz.

The only difference like someone said above is that the FX has dual-channel memory not HT tunnels.

-Por
 

Jeff7181

Lifer
Aug 21, 2002
18,368
11
81
Originally posted by: PorBleemo
You are all CONFUSED!

Each K8 processor has exactly two HyperTransport tunnels running at 800Mhz going to and from the processor. Each channel is 32-Bits wide so the grand total of bandwidth is 6.4GB/s. In this sense though you are correct that the total bandwidth could be considered 1.6GHz.

The only difference like someone said above is that the FX has dual-channel memory not HT tunnels.

-Por

Socket 754 processors only have 1 HT link.
Socket 940 processors have 3 HT links.
 

InlineFive

Diamond Member
Sep 20, 2003
9,599
2
0
Originally posted by: Jeff7181
Originally posted by: PorBleemo
You are all CONFUSED!

Each K8 processor has exactly two HyperTransport tunnels running at 800Mhz going to and from the processor. Each channel is 32-Bits wide so the grand total of bandwidth is 6.4GB/s. In this sense though you are correct that the total bandwidth could be considered 1.6GHz.

The only difference like someone said above is that the FX has dual-channel memory not HT tunnels.

-Por

Socket 754 processors only have 1 HT link.
Socket 940 processors have 3 HT links.

No I thought they all had 2 HT links according to AMD. *Me goes off to do research...* Links?

-Por
 

Stunt

Diamond Member
Jul 17, 2002
9,717
2
0
Herre's a link to a site that should help a bit.

Hypertransport

if hammer supported the 32 bit width and peak frequency, the bus would support 25600Mb/s on hypertransport 1, instead of the 6400Mb/s it supports now.

Then see what Hypertransport 2 brings to the table.
 

ta2

Member
Jan 21, 2004
56
0
0
Originally posted by: PorBleemo
You are all CONFUSED!

Each K8 processor has exactly two HyperTransport tunnels running at 800Mhz going to and from the processor. Each channel is 32-Bits wide so the grand total of bandwidth is 6.4GB/s. In this sense though you are correct that the total bandwidth could be considered 1.6GHz.

The only difference like someone said above is that the FX has dual-channel memory not HT tunnels.

-Por

So you are saying that the 939 processors will still use PC3200 RAM?
 

andreasl

Senior member
Aug 25, 2000
419
0
0
Originally posted by: PorBleemo
You are all CONFUSED!

Each K8 processor has exactly two HyperTransport tunnels running at 800Mhz going to and from the processor. Each channel is 32-Bits wide so the grand total of bandwidth is 6.4GB/s. In this sense though you are correct that the total bandwidth could be considered 1.6GHz.

The only difference like someone said above is that the FX has dual-channel memory not HT tunnels.

-Por

Talk about being confused

First the K8 has 3 HT links (Socket 754 has only 1). Each link is bi-directional, 16 bit wide and runs at 800MHz DDR (1600MHz effective). This gives it a bandwidth of 3.2GB/s in each direction for the combined 6.4GB/s.
 

ta2

Member
Jan 21, 2004
56
0
0
I heard that the 939 processors will use 1066MHz/link or 1000MHz/link... So then they will all have 3*1066MHz hypertransport?
 

andreasl

Senior member
Aug 25, 2000
419
0
0
Originally posted by: ta2
I heard that the 939 processors will use 1066MHz/link or 1000MHz/link... So then they will all have 3*1066MHz hypertransport?

Well only one will be used. I'm not sure how many HT links socket 939 will have, but considering it's their desktop socket it might only have 1.

But basically the speed of the HT links has hardly any influence on the performance of the system. For example look at the difference between the Nforce3 150 and the VIA K8 chipset. In most cases the difference is negligable and that's with a massive difference in bandwidth (2.4GB/s vs 6.4GB/s). Don't expect much improvement from the 1GHz HT speed.
 

Cerb

Elite Member
Aug 26, 2000
17,484
33
86
Originally posted by: ta2
Does anyone have any idea if the...

AMD Athlon 64 3700+
2.4GHz
512KB Cache
Releaseed Q2 '04
Socket 939

...will have a Hypertransport speed of 1066MHz? Looking at the clock speed, it is possible (2400/533 = 4.5). Since if it will have 1066 Hypertransport the DDR ram fast enough to support it will have extremely high latencies and be very expensive, and DDR2 will be even more expensive.

So, any ideas?



Cheers
1. No reason not to just go for 1000MHz....but doubtful.
2. 2400/533? WTF? It's 2400/200 (12), maybe it'll be 250 (~9.5), 266 (9) or 300 (8). Have you been smoking the Intel Crack Pipe again? Remember, a P4A 2.0 was 100x20, not 400x5.
3. DDR fast enough to support it? Huh? As it is they will use as low as 100MHz DDR...that's what...oh yeah, 1.6GB/s, where the current HT link is 3.2GB/s each way. DC DDR now officially supports up to 6.4GB/s on the RAM (PC3200x2), and there's no reason to increase that because the HT link speeds up. Marketting and demand are other issues entirely, and are a large part of what is governing the move to Socket939 this soon.
4. DDR-II will have to come down, unless it can seriously increase in clock speeds enough to leave DC DDR in the dust. But oh, if it can do that, it will be beneficial for it to come down, as then the volume of sales will severely increase.

The current HT links, much like the current P4 FSB, are not being bottlenecks. increasing them offers little or no performance gain. As it is, any increase, probably until at least this fall, but probably winter, will be purely for marketting.
 

alexruiz

Platinum Member
Sep 21, 2001
2,836
556
126
Let's clear the confussion. The Hypertransport tunnel speed has NOTHING to do with the kind of memory used. You are still thinking OLD CPU architecture, where the only data exchange path from the CPU to the world was the FSB. Now, for the AMD64 chips, you have 2 paths to exchange data:

One of those paths is the memory controller, integrated into the CPU. The other path is the HT link, intended mainly for peripherial access and video card access (AGP, and soon PCI express). Both numbers are independent from each other. The fastest memory that can be used depends on the CPU itself, based on how high the integrated memory controller can scale.

Both numbers seem to be related because some motherboard implementations link them through a multiplier (WRONG approach, but I assume it saves money.... ). Both numbers should be totally independent, and I hope NForce3 250 fixes it (yes, having the choice to change the memory speed without touching the HT link speed)

Hence, you can have combinations of a extremely fast HT (1 GHz each direction, 2 GHz total, 8 GB/sec) coupled with low memory bandwidth (mobile Athlon DTR socket 754, DDR333 single channel, 2.7 GB/sec) On the other hand, you could have a very "slow" HT tunnel (ftake the speed of the nVidia NForce 3 Go 120, used for the transmeta efficeon, carrying no more than 1.6 GB/sec... 200 Mhz maybe?) coupled with a CPU that has potent memory access (Athlon 64 FX55, socket 939, DDR500 dual channel, 8 GB/sec).......

The memory can go as high as the CPU memory controller allows, and the BIOS must be aware of such capability. The HT tunnels is used ONLY for peripherials and AGP access... that is why the K8T800 performs better than the NForce 3 150 in workstation applications (intensive on the AGP bus).

Alex
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |