Sign in with
Sign up | Sign in
Your question
Solved

S1g2 vs S1g3

Tags:
Last response: in CPUs
Share
October 21, 2010 8:31:03 AM

I bought a Gateway NV53 off ebay and it has an Athlon II M300 that's S1g3. I was wondering if a Turion X2 Ultra ZM-84, which is S1g2, would be compatible? If not, is it due to a physical difference such as pin number or is it strictly an architectural issue. Otherwise, I'd just be upgrading to an M600, unless you guys can provide a better alternative that I'm not aware of.

Thanks

More about : s1g2 s1g3

October 22, 2010 2:57:10 AM

123 views, 0 replies ???

I read somewhere that S1g3's are backwards compatible, but S1g2's are not forward compatible. So does this mean an S1g3 would work in an S1g2, but not vice versa? If so, I answered my own question. Can someone please confirm this? I get my NV53 tomorrow and don't want to miss out on a CPU if it'll work.
m
0
l
October 22, 2010 7:46:23 PM

Nevermind, this is useless
m
0
l
Related resources
Can't find your answer ? Ask !

Best solution

a c 102 à CPUs
October 23, 2010 2:35:58 AM

theignition34 said:
I bought a Gateway NV53 off ebay and it has an Athlon II M300 that's S1g3. I was wondering if a Turion X2 Ultra ZM-84, which is S1g2, would be compatible? If not, is it due to a physical difference such as pin number or is it strictly an architectural issue. Otherwise, I'd just be upgrading to an M600, unless you guys can provide a better alternative that I'm not aware of.

Thanks


Socket S1 is apparently physically identical across all four of its revisions- S1, S1r2, S1r3, S1r4. The differences lie with the specifications of the buses that connect to the socket. For example, S1g4 uses DDR3 while the others are DDR2. The original Socket S1 is limited to an 800 MHz HT2 link speed, while the others are 2 GHz+ HT3. I am not 100% sure about the differences between S1g2 and S1g3, but I do know that the S1g3 units are K10-based while the S1g2 units are K8-based. I think that the difference may be in the power planes of the sockets, since both S1g2 and S1g3 support DDR2 and HT3. I would guess that the ZM-84 would work unless Gateway didn't include BIOS code for the K8-based S1g2 chips, but I can't guarantee that. AMD is a little behind in publishing information on Socket S1 to their website, so I can't say for sure if it is in fact compatible or not.
Share
October 23, 2010 11:20:27 AM

Best answer selected by theignition34.
m
0
l
April 25, 2011 1:25:42 PM

MU_Engineer said:
Socket S1 is apparently physically identical across all four of its revisions- S1, S1r2, S1r3, S1r4. The differences lie with the specifications of the buses that connect to the socket. For example, S1g4 uses DDR3 while the others are DDR2. The original Socket S1 is limited to an 800 MHz HT2 link speed, while the others are 2 GHz+ HT3. I am not 100% sure about the differences between S1g2 and S1g3, but I do know that the S1g3 units are K10-based while the S1g2 units are K8-based. I think that the difference may be in the power planes of the sockets, since both S1g2 and S1g3 support DDR2 and HT3. I would guess that the ZM-84 would work unless Gateway didn't include BIOS code for the K8-based S1g2 chips, but I can't guarantee that. AMD is a little behind in publishing information on Socket S1 to their website, so I can't say for sure if it is in fact compatible or not.



I have an interest in upgrading a machine that was originally configured with an S1g1 socket/CPU so that it would have an S1g3-designed CPU. (The S1g3 CPUs can sometimes be a small fraction of the cost of the earlier ones.)

In your comments above, you mentioned that the "S1g1 is limited to an 800 MHz HT2 link speed, while the others are 2 GHz+ HT3". What does this mean in practical terms? Is the HT speed something that is limited by the socket, or by the CPU operating parameters? If it is limited by the socket, does it matter to the system as a whole, or does the CPU simply have to "wait" a lot? If it is simply a "waiting" problem, it does not seem to be a discouraging influence?

I don't know the significance of "K10-based" vs "K-8 based" etc., what does this imply?

It appears to me that you are correct in noting that the difference in the different S1 generations lies in the power planes, but as the voltages on the S1g1 sockets are nominally higher those on the S1g2 sockets (...and by extension possibly also on S1g3 sockets...??), I'm thinking that this simply means that using an S1g3 CPU in an S1g1 socket should tend to "over-clock" the S1g3 CPU slightly in the S1g1 socket. Does this seem reasonable to you, or am I simply not understanding the dynamics of the situation?


According to Wikipedia ( http://en.wikipedia.org/wiki/List_of_AMD_Turion_micropr... ):

Voltages for the S1g1 socket are: 1.075/1.10/1.125 V

Voltages for the S1g2 socket are: 0.75-1.2V for the Turion Ultra ZM, and RM, series chips, and 0.95-1.1V for the Athlon QL series chips.

I have no specific information on the S1g3 socket, voltage specs.


My gut tells me that the CPU chip is "in charge" of the operating parameters, and that the system depends on its ability to access the resources it demands to achieve optimal functionality. In other words, as long as it has enough voltage, it will run things as fast as the data arrives, up to the limits of the CPU chip itself?

My question then is this: If I insert a later generation S1g3 CPU in the S1g1 socket. What (if any) 'bad' things will happen to the CPU thus inserted?


Thank you for any information you can provide.


P.S. I have chosen to stop at S1g3, because S1g4 "requires" DDR3 memory. While interesting, does this have any impact on the situation? In other words, does the S1g4-designed CPU "see" anything that would cause it to have difficulty in a DDR2 system, even though it is engineered for DDR3?
m
0
l
April 25, 2011 1:29:20 PM

MU_Engineer said:
Socket S1 is apparently physically identical across all four of its revisions- S1, S1r2, S1r3, S1r4. The differences lie with the specifications of the buses that connect to the socket. For example, S1g4 uses DDR3 while the others are DDR2. The original Socket S1 is limited to an 800 MHz HT2 link speed, while the others are 2 GHz+ HT3. I am not 100% sure about the differences between S1g2 and S1g3, but I do know that the S1g3 units are K10-based while the S1g2 units are K8-based. I think that the difference may be in the power planes of the sockets, since both S1g2 and S1g3 support DDR2 and HT3. I would guess that the ZM-84 would work unless Gateway didn't include BIOS code for the K8-based S1g2 chips, but I can't guarantee that. AMD is a little behind in publishing information on Socket S1 to their website, so I can't say for sure if it is in fact compatible or not.



I have an interest in upgrading a machine that was originally configured with an S1g1 socket/CPU so that it would have an S1g3-designed CPU. (The S1g3 CPUs can sometimes be a small fraction of the cost of the earlier ones.)

In your comments above, you mentioned that the "S1g1 is limited to an 800 MHz HT2 link speed, while the others are 2 GHz+ HT3". What does this mean in practical terms? Is the HT speed something that is limited by the socket, or by the CPU operating parameters? If it is limited by the socket, does it matter to the system as a whole, or does the CPU simply have to "wait" a lot? If it is simply a "waiting" problem, it does not seem to be a discouraging influence?

I don't know the significance of "K10-based" vs "K-8 based" etc., what does this imply?

It appears to me that you are correct in noting that the difference in the different S1 generations lies in the power planes, but as the voltages on the S1g1 sockets are nominally higher those on the S1g2 sockets (...and by extension possibly also on S1g3 sockets...??), I'm thinking that this simply means that using an S1g3 CPU in an S1g1 socket should tend to "over-clock" the S1g3 CPU slightly in the S1g1 socket. Does this seem reasonable to you, or am I simply not understanding the dynamics of the situation?


According to Wikipedia ( http://en.wikipedia.org/wiki/List_of_AMD_Turion_micropr... ):

Voltages for the S1g1 socket are: 1.075/1.10/1.125 V

Voltages for the S1g2 socket are: 0.75-1.2V for the Turion Ultra ZM, and RM, series chips, and 0.95-1.1V for the Athlon QL series chips.

I have no specific information on the S1g3 socket, voltage specs.


My gut tells me that the CPU chip is "in charge" of the operating parameters, and that the system depends on its ability to access the resources it demands to achieve optimal functionality. In other words, as long as it has enough voltage, it will run things as fast as the data arrives, up to the limits of the CPU chip itself?

My question then is this: If I insert a later generation S1g3 CPU in the S1g1 socket. What (if any) 'bad' things will happen to the CPU thus inserted?


Thank you for any information you can provide.


P.S. I have chosen to stop at S1g3, because S1g4 "requires" DDR3 memory. While interesting, does this have any impact on the situation? In other words, does the S1g4-designed CPU "see" anything that would cause it to have difficulty in a DDR2 system, even though it is engineered for DDR3?
m
0
l
April 25, 2011 1:29:22 PM

MU_Engineer said:
Socket S1 is apparently physically identical across all four of its revisions- S1, S1r2, S1r3, S1r4. The differences lie with the specifications of the buses that connect to the socket. For example, S1g4 uses DDR3 while the others are DDR2. The original Socket S1 is limited to an 800 MHz HT2 link speed, while the others are 2 GHz+ HT3. I am not 100% sure about the differences between S1g2 and S1g3, but I do know that the S1g3 units are K10-based while the S1g2 units are K8-based. I think that the difference may be in the power planes of the sockets, since both S1g2 and S1g3 support DDR2 and HT3. I would guess that the ZM-84 would work unless Gateway didn't include BIOS code for the K8-based S1g2 chips, but I can't guarantee that. AMD is a little behind in publishing information on Socket S1 to their website, so I can't say for sure if it is in fact compatible or not.



I have an interest in upgrading a machine that was originally configured with an S1g1 socket/CPU so that it would have an S1g3-designed CPU. (The S1g3 CPUs can sometimes be a small fraction of the cost of the earlier ones.)

In your comments above, you mentioned that the "S1g1 is limited to an 800 MHz HT2 link speed, while the others are 2 GHz+ HT3". What does this mean in practical terms? Is the HT speed something that is limited by the socket, or by the CPU operating parameters? If it is limited by the socket, does it matter to the system as a whole, or does the CPU simply have to "wait" a lot? If it is simply a "waiting" problem, it does not seem to be a discouraging influence?

I don't know the significance of "K10-based" vs "K-8 based" etc., what does this imply?

It appears to me that you are correct in noting that the difference in the different S1 generations lies in the power planes, but as the voltages on the S1g1 sockets are nominally higher those on the S1g2 sockets (...and by extension possibly also on S1g3 sockets...??), I'm thinking that this simply means that using an S1g3 CPU in an S1g1 socket should tend to "over-clock" the S1g3 CPU slightly in the S1g1 socket. Does this seem reasonable to you, or am I simply not understanding the dynamics of the situation?


According to Wikipedia ( http://en.wikipedia.org/wiki/List_of_AMD_Turion_micropr... ):

Voltages for the S1g1 socket are: 1.075/1.10/1.125 V

Voltages for the S1g2 socket are: 0.75-1.2V for the Turion Ultra ZM, and RM, series chips, and 0.95-1.1V for the Athlon QL series chips.

I have no specific information on the S1g3 socket, voltage specs.


My gut tells me that the CPU chip is "in charge" of the operating parameters, and that the system depends on its ability to access the resources it demands to achieve optimal functionality. In other words, as long as it has enough voltage, it will run things as fast as the data arrives, up to the limits of the CPU chip itself?

My question then is this: If I insert a later generation S1g3 CPU in the S1g1 socket. What (if any) 'bad' things will happen to the CPU thus inserted?


Thank you for any information you can provide.


P.S. I have chosen to stop at S1g3, because S1g4 "requires" DDR3 memory. While interesting, does this have any impact on the situation? In other words, does the S1g4-designed CPU "see" anything that would cause it to have difficulty in a DDR2 system, even though it is engineered for DDR3?
m
0
l
a c 102 à CPUs
April 25, 2011 1:37:06 PM

richscho said:
I have an interest in upgrading a machine that was originally configured with an S1g1 socket/CPU so that it would have an S1g3-designed CPU. (The S1g3 CPUs can sometimes be a small fraction of the cost of the earlier ones.)

In your comments above, you mentioned that the "S1g1 is limited to an 800 MHz HT2 link speed, while the others are 2 GHz+ HT3". What does this mean in practical terms? Is the HT speed something that is limited by the socket, or by the CPU operating parameters? If it is limited by the socket, does it matter to the system as a whole, or does the CPU simply have to "wait" a lot? If it is simply a "waiting" problem, it does not seem to be a discouraging influence?


In practical terms, the most it would mean is that your integrated graphics would run a little slower with a slower HT link speed, since the IGP gets access to memory over the HT link. CPU performance in single-CPU systems is basically unaffected by HT link speed.

Quote:
I don't know the significance of "K10-based" vs "K-8 based" etc., what does this imply?


K8-based refers to any CPU whose design is based on the original Athlon 64. K10 is an updated version of K8 and is what underpins Phenom-class CPUs. The thing you probably care about is that K10 is newer and faster than K8 at identical clock speeds.

Quote:
It appears to me that you are correct in noting that the difference in the different S1 generations lies in the power planes, but as the voltages on the S1g1 sockets are nominally higher those on the S1g2 sockets (...and by extension possibly also on S1g3 sockets...??), I'm thinking that this simply means that using an S1g3 CPU in an S1g1 socket should tend to "over-clock" the S1g3 CPU slightly in the S1g1 socket. Does this seem reasonable to you, or am I simply not understanding the dynamics of the situation?


An S1g3 CPU in an S1g1 motherboard will almost certainly refuse to work. The BIOS and such will not recognize the new CPU and its features and thus do nothing or at most give some beeps and nothing more when you try to start it up.


Quote:
According to Wikipedia ( http://en.wikipedia.org/wiki/List_of_AMD_Turion_micropr... ):

Voltages for the S1g1 socket are: 1.075/1.10/1.125 V

Voltages for the S1g2 socket are: 0.75-1.2V for the Turion Ultra ZM, and RM, series chips, and 0.95-1.1V for the Athlon QL series chips.

I have no specific information on the S1g3 socket, voltage specs.


My gut tells me that the CPU chip is "in charge" of the operating parameters, and that the system depends on its ability to access the resources it demands to achieve optimal functionality. In other words, as long as it has enough voltage, it will run things as fast as the data arrives, up to the limits of the CPU chip itself?

My question then is this: If I insert a later generation S1g3 CPU in the S1g1 socket. What (if any) 'bad' things will happen to the CPU thus inserted?


Probably nothing bad will happen, but the unit will refuse to work.

Quote:
P.S. I have chosen to stop at S1g3, because S1g4 "requires" DDR3 memory. While interesting, does this have any impact on the situation? In other words, does the S1g4-designed CPU "see" anything that would cause it to have difficulty in a DDR2 system, even though it is engineered for DDR3?


AMD CPUs have the memory controller built into the CPU. Some DDR3-using AMD CPUs can use DDR2 as well due to tweaks being made to their memory controllers, some cannot. I would hazard a guess that an S1g4 CPU is DDR3-only and will not work with DDR2.
m
0
l
April 25, 2011 2:04:44 PM

Thanks very much for the reply. Sadly it means that I will need to go with the older style processor, but it is great comfort to have information that relieves the massive uncertainty, that accompanies all such decisions.

Thanks again.
m
0
l
!