I would say that the ati made sense but it was bad timing correct me if I am wrong but wasn't ati purchased right before or after the phenom flopped because of the erratum bug. Fast forward amd is only now catching up to the time it lost they have a talented development team especially in the graphics processor wing and up until core 2 duo they could meet or beat intel at any price point. The phenom 2 is what phenom should have been so if amd can make a jump with its processors like it did with its graphics chips when going from the x3000 series to the x4000 series intel would have a real competitor on its hands again.
- Status
grieve
Distinguished
- Apr 19, 2004
- 2,709
- 0
- 20,790
[citation][nom]superhighperf[/nom]amd should invest more in marketing. most people think that amd is just a cheep substitute for intel. run some adds like the apple vs windows stuff. make people realize that amd is a great computing solution. focus on the needs of an average user and show how little difference there is between intell vs amd. there are only a few percentage of users who really benefit from using intell. i think most users could benefit more from saving a few bucks than a few seconds encoding a mp3 etc.[/citation]
I don’t agree with spending anything in marketing…
I’m guessing but I think the average Joe goes to Bestbuy or dell to purchase his PC. I think the days of “Pentium” are gone and most people don’t even give a crap what’s running there machine because they just want it to run. They are not running benchmarks they want to cruise the net and look at emails and are just looking for a good deal. Just like most people don’t know anything about their car engine, simple things too.. Like how much HP they have.
Just my thoughts… The only people that even know AMD is behind Intel are techs and computer geeks, 10% of the population (total guess).
You do have a valid point with the Apple ads, I do think it sucks the morons in but Apple must spend INSANE amounts of cash for what type of returns?
I don’t agree with spending anything in marketing…
I’m guessing but I think the average Joe goes to Bestbuy or dell to purchase his PC. I think the days of “Pentium” are gone and most people don’t even give a crap what’s running there machine because they just want it to run. They are not running benchmarks they want to cruise the net and look at emails and are just looking for a good deal. Just like most people don’t know anything about their car engine, simple things too.. Like how much HP they have.
Just my thoughts… The only people that even know AMD is behind Intel are techs and computer geeks, 10% of the population (total guess).
You do have a valid point with the Apple ads, I do think it sucks the morons in but Apple must spend INSANE amounts of cash for what type of returns?
[citation][nom]deregtx[/nom]I would have bought VIA Technologies instead of Ati.So they whould have the Via Chipset business, S3 Graphics and its cpu business.I think S3 is able to make highend GPU cards but they focus on cheap ones..Via had Intel and Amd chipsets at the time(just like Ati) so that would work out.And they would have a good product against Intel's Atom, via's Nano cpu.And the biggest pro is that S3 Graphics is located in California(just like AMD's cpu design)and not in canada like ATI, so you could merge the two design labs to make one design lab, and make a real Fusion between a Gpu and Cpu like Intel is doing with Larrabee.The Fusion they are making now is just a normal GPU core added to a cpu.Another plus is that Via has some great patents that intel needs[/citation]
ATI's headquarters is indeed in Toronto (or specifically in Markham which is just 5 blocks south of me) but they have also a design facility in Silicon Valley as well.
What I would do different if I could with AMD is to further nurture co-operation with IBM to get faster die shrinks, press down prices to undercut Intel as much as possible (it was one reason that dragged me from Intel to AMD in the first place). Focus on OEM partners and give them reasons to build machines with AMD processors. Remember that these OEM versions are the ones that end up with the most consumers. Focus on building a lot more efficient notebook processors. The Turions still lose out compared to the Intel Core 2 notebook processors in terms of energy efficiency. It's not all about raw power as people here keeps talking about. It's important to hit the key markets and sadly it isn't what most computer enthusiasts would like which is not as fast as possible.
So, in summary to get AMD back on its feet, continue working with partners like IBM. Push reasons for more OEM support to build machines with AMD processors and to push for newer and more efficient notebook designs. Notebooks are getting more popular than desktops so this will be a big thing to go after which Intel is in the lead.
ATI's headquarters is indeed in Toronto (or specifically in Markham which is just 5 blocks south of me) but they have also a design facility in Silicon Valley as well.
What I would do different if I could with AMD is to further nurture co-operation with IBM to get faster die shrinks, press down prices to undercut Intel as much as possible (it was one reason that dragged me from Intel to AMD in the first place). Focus on OEM partners and give them reasons to build machines with AMD processors. Remember that these OEM versions are the ones that end up with the most consumers. Focus on building a lot more efficient notebook processors. The Turions still lose out compared to the Intel Core 2 notebook processors in terms of energy efficiency. It's not all about raw power as people here keeps talking about. It's important to hit the key markets and sadly it isn't what most computer enthusiasts would like which is not as fast as possible.
So, in summary to get AMD back on its feet, continue working with partners like IBM. Push reasons for more OEM support to build machines with AMD processors and to push for newer and more efficient notebook designs. Notebooks are getting more popular than desktops so this will be a big thing to go after which Intel is in the lead.
I thought it was a little ballsy but now it seems like a good move and might be responsible for intel moving into the discrete graphics card market. (more competition)
ATI graphics cards have been a huge success for AMD since the HD3x series and their onboard chips are superior to nvidias and have taken a huge share of the chipset market away from nvidia.
I wonder what would AMD would be right now if they hadn't acquired ATI.
After the release of the core2duo AMD probably had a huge sales slump on the CPU market but ATI was able to do quite well. Would AMD be in more financial trouble? If so could they have borrowed enough money to stay afloat?
masterasia
Distinguished
- Feb 9, 2009
- 1,128
- 0
- 19,360
AMD is second best right now, so they need to lower prices than Intel. Phenom II 940 is around $200. The i7 920 is $230. For $30 more, I would totally get the i7. If the Phenom II 940 was around $160, then I would be having second thoughts about the the i7.
AMD's AM3 socket processors look like they have great potential and might overtake Intel's C2D and C2Q line. Can't wait to get my hands on a Phenom II 990.
AMD's AM3 socket processors look like they have great potential and might overtake Intel's C2D and C2Q line. Can't wait to get my hands on a Phenom II 990.
Which patents does Via has that Intel would want?
I think that i would change 3 things only:
1) Redesign K8 at 65nm (forget quad cores for a while) and launch it at K9. We have almost same athlon since Slot A. They were very competitive against Pentium III and best clock per clock against Netburst. But Core is better then P-III, so AMD should change more then just Macro-architecture from K8 to K9. Then do the same at 45nm. The quad cores at 65nm would be two K9, each one with one memmory channel (unfortunately isn't as easy as intel because memmory controller is at core, so this is best way i could think xD). So we would have K8 at 90 and 65nm, K9 at 65 and 45nm and K10 at 45 and 32nm.
2) Instead create a new corporation lice TFC with ATIC, should lay-off new stock to ATIC ad avoid problems with Intel. Or sell 180nm and 130nm, and perhaps 90nm. This would make some cash.
3) Focus on mobile and mainstream... Spend less money forgetting the performance leaders. Make it's brand more known as better choice for office and home use. Intel, with more expensive chips, could take the performance hungry and overclocker users. Most of the market doesn't need $300 or $400 chips. So smaller CPUs able to do daily needs would cost less to make and give more profit.
I think that i would change 3 things only:
1) Redesign K8 at 65nm (forget quad cores for a while) and launch it at K9. We have almost same athlon since Slot A. They were very competitive against Pentium III and best clock per clock against Netburst. But Core is better then P-III, so AMD should change more then just Macro-architecture from K8 to K9. Then do the same at 45nm. The quad cores at 65nm would be two K9, each one with one memmory channel (unfortunately isn't as easy as intel because memmory controller is at core, so this is best way i could think xD). So we would have K8 at 90 and 65nm, K9 at 65 and 45nm and K10 at 45 and 32nm.
2) Instead create a new corporation lice TFC with ATIC, should lay-off new stock to ATIC ad avoid problems with Intel. Or sell 180nm and 130nm, and perhaps 90nm. This would make some cash.
3) Focus on mobile and mainstream... Spend less money forgetting the performance leaders. Make it's brand more known as better choice for office and home use. Intel, with more expensive chips, could take the performance hungry and overclocker users. Most of the market doesn't need $300 or $400 chips. So smaller CPUs able to do daily needs would cost less to make and give more profit.
shadowmaster625
Distinguished
- Mar 27, 2007
- 352
- 0
- 18,780
This is a no brainer: Integrated SSD Controller
It's a pretty simple concept. Picture your typical motherboard with its DIMM slots for dual channel memory. Next to those, picture a couple similar slots for flash DIMMS. The DIMMs would only cost $30-100 because all the flash controller logic would be embedded into the CPU. The potential bandwidth is only limited by imagination. Imagine Fusion-IO type performance out of a $200 motherboard/cpu/memory combo! If AMD can get this done, they render Intel's expensive SSDs obsolete.
It's a pretty simple concept. Picture your typical motherboard with its DIMM slots for dual channel memory. Next to those, picture a couple similar slots for flash DIMMS. The DIMMs would only cost $30-100 because all the flash controller logic would be embedded into the CPU. The potential bandwidth is only limited by imagination. Imagine Fusion-IO type performance out of a $200 motherboard/cpu/memory combo! If AMD can get this done, they render Intel's expensive SSDs obsolete.
I would not do anything but R&D for 2 years and conserve money in the CPU market while concentrating on the GPU market. By the time a chip comes out from the R&D, there would be a couple years difference. As Intel would have no competition in the CPU market for these years, in about 1 year they will get complacent and overpriced. The perfect oppurtunity to push a better chip on the market.
I would not have spun off global foundries. Instead I would have looked for foreign partners that want to pay to have an AMD designed and run Fabrication plant in their country. The biggest advantage Intel has over AMD is capacity. Fab plants are expensive. But many countries want to get into the high tech game.
apache_lives
Splendid
- Dec 10, 2002
- 6,873
- 1
- 26,160
I would target SOC (system on a chip) designs where Intel is looking - AMD would have a similar but more powerful platform thanks to ATi, combining a ~256k K7 core @ 45nm (previously 130nm etc) with chipset etc - enabling smaller devices etc
Perhaps even buy VIA out at the right price to jump right into that Netbook market, and stop Nvidia in its tracks
Also ATi needs more console deals - there always a winner, and any issues go back to the console maker not the hardware manufacturers etc.
Hmmmmmmm logic chipsets for iPhone and similar devices?
Perhaps even buy VIA out at the right price to jump right into that Netbook market, and stop Nvidia in its tracks
Also ATi needs more console deals - there always a winner, and any issues go back to the console maker not the hardware manufacturers etc.
Hmmmmmmm logic chipsets for iPhone and similar devices?
MagicPants
Distinguished
- Jun 16, 2006
- 1,315
- 0
- 19,660
I'd concentrate solely on performance/price and performance/watts. Since Intel has become serious about keeping the performance crown AMD just can't keep up. AMD should continue with their goal of designing an entire platform rather than just the CPU/GPU. Get the whole platform price down while giving consumers better features like better integrated video. Make Intel waste money on the bleeding edge.
For ATI work with Nvidia, you need to position GPUs as a replacement for multicore CPUs. The way to do this is to standardize, right now people don't know whether to go Nvidia or ATI for non-graphics GPU applications so the end up going with Intel.
For ATI work with Nvidia, you need to position GPUs as a replacement for multicore CPUs. The way to do this is to standardize, right now people don't know whether to go Nvidia or ATI for non-graphics GPU applications so the end up going with Intel.
1) I would focus on winning the server segment. This can be won because of Intel's outrageous price premiums. (also forcing them to lower server CPU prices is good, because this is what feeds their R&D) AMD doesn't even need uber powers, they just need to be cost effective. This means a whole server platform for under a grand that is fast enough with all the fixings. eg: As long as it has a ton of cache, 6 cores (two triple core opterons would be neat to claime 6 cores) and can carry a whack load of Ram, people will buy it if it is cheap enough.
2) All stock CPUs unlocked, until significant market penetration is achieved. This should get all the enthusiasts to try it out.
3) High end integrated graphics. If I could buy a motherboard that included powerful integrated crossfire for $100 cheaper, I would probably buy it over mixing Intel and Nvidia.
4) Reduction of SKUs. There should be
1) Dirt cheap for netbooks (dual or crippled triple core from #2) [needs HDMI and Highdef and can be made into CHEAP media centers]
2) Entry level workhorse for running real apps (triple core) (that can overclock performance past #3)
3) Medium powered workstation (quad) (that can overclock past #4)
4) Enthiuast Quad (Can overclock to keep up with or do better than high end Intel)
[full solutions beyond this point]
5) Entry Server (1 CPU on mobo + quad or triple core opteron with extra cache) SOHO targeted, I'm talking $600 for the solution with all server features.
6) Medium Server [2 Quad or 2 triple core opterons] $1000, overclockable to keep up with highend Intel Xeons.
7) Uber Server [3 or 4 sockets on mobo] same CPUs as #6, just more cores since people are stupid.
Seriously, face it, people buy AMD for dirt cheap entry, or for overclocking performance effectiveness. Reduce the # of SKUs, you are wasting R&D.
ATI needs an answer to larrabee and needs to research ray tracing in case Intel pushes that way to kill Nvidia.
2) All stock CPUs unlocked, until significant market penetration is achieved. This should get all the enthusiasts to try it out.
3) High end integrated graphics. If I could buy a motherboard that included powerful integrated crossfire for $100 cheaper, I would probably buy it over mixing Intel and Nvidia.
4) Reduction of SKUs. There should be
1) Dirt cheap for netbooks (dual or crippled triple core from #2) [needs HDMI and Highdef and can be made into CHEAP media centers]
2) Entry level workhorse for running real apps (triple core) (that can overclock performance past #3)
3) Medium powered workstation (quad) (that can overclock past #4)
4) Enthiuast Quad (Can overclock to keep up with or do better than high end Intel)
[full solutions beyond this point]
5) Entry Server (1 CPU on mobo + quad or triple core opteron with extra cache) SOHO targeted, I'm talking $600 for the solution with all server features.
6) Medium Server [2 Quad or 2 triple core opterons] $1000, overclockable to keep up with highend Intel Xeons.
7) Uber Server [3 or 4 sockets on mobo] same CPUs as #6, just more cores since people are stupid.
Seriously, face it, people buy AMD for dirt cheap entry, or for overclocking performance effectiveness. Reduce the # of SKUs, you are wasting R&D.
ATI needs an answer to larrabee and needs to research ray tracing in case Intel pushes that way to kill Nvidia.
Ok. you asked, so there is:
notebooks motherboard. Today they are expensive, scarce, and hard to find. I did not see reviews of motherboards on enthusiast sites.
cellphone/netbook motherboards. See above.
notebooks components. Video cards, etc. To assemble DIY notebooks, we need all the components. ¿why to choose between 3, 6, or 9 cells? I want to buy 6 cells, and later add 6 more. And I want a standard on sizes and connectors. Today, parts are rarely interchangeable between different vendors. It sucks.
multi CPU motherboard, as ATI X2 cards give a run for the money to nvidia hugue chips, If the user can buy cheap multisocket Phenom X4 (Or X6/X8), a multisocket motherboard would allow to compete with i7.
Skulltrail was a shame. Came on, AMD you did it with 64 bit, Dual processors, and integrated memory controller... shows us that you can do a better platform than sucking skulltrail!
massive cache memory. I want to buy 1 Gb of extra caché, and add it to my system.
GPGPU RAID 0 support. Don't know if possible, but it would open the way to a GPGPU without CPU.
Simultaneous chipset support for RAID 0 AND Raid 1 in different partitions as intel does. (And Intel RAID performs faster than nvidia or ATI, so please, enhance it).
faster phenoms with DDR3. Why the fasters Phenoms are DDR3 only? It makes nonsense.
PCI cards to put the older, discarded CPUs, and his memory. I hate wasting out an Atlhon 5200 AM2 only because I want an AM3 Phenom X4 with DDR3.
The X2 5200 with DDR2 800 still have a lot of horsepower.
Better coolers. factory coolers suck.
Some kind of hyperthreading, recognizable with software, so it can enable/disable it at convenience.
Enable/disable hyperthreading in ATI drivers profiles, adjusted for each game.
OF course, backward-compatible Havoc drivers. Not like useless PhysX, which don't work with older games.
... and in the fantasy wish list:
internal working with DDRX technologies. Why internal data on processors should transport only one bit by clock, when DDR memory transport many bits in each clock?
it needs multistage transistors, capable of switching between more than two states with low latency.
such "polinary" transistors could store more bits in the same cache size, and multiplicate number of CPUs in the same space.
notebooks motherboard. Today they are expensive, scarce, and hard to find. I did not see reviews of motherboards on enthusiast sites.
cellphone/netbook motherboards. See above.
notebooks components. Video cards, etc. To assemble DIY notebooks, we need all the components. ¿why to choose between 3, 6, or 9 cells? I want to buy 6 cells, and later add 6 more. And I want a standard on sizes and connectors. Today, parts are rarely interchangeable between different vendors. It sucks.
multi CPU motherboard, as ATI X2 cards give a run for the money to nvidia hugue chips, If the user can buy cheap multisocket Phenom X4 (Or X6/X8), a multisocket motherboard would allow to compete with i7.
Skulltrail was a shame. Came on, AMD you did it with 64 bit, Dual processors, and integrated memory controller... shows us that you can do a better platform than sucking skulltrail!
massive cache memory. I want to buy 1 Gb of extra caché, and add it to my system.
GPGPU RAID 0 support. Don't know if possible, but it would open the way to a GPGPU without CPU.
Simultaneous chipset support for RAID 0 AND Raid 1 in different partitions as intel does. (And Intel RAID performs faster than nvidia or ATI, so please, enhance it).
faster phenoms with DDR3. Why the fasters Phenoms are DDR3 only? It makes nonsense.
PCI cards to put the older, discarded CPUs, and his memory. I hate wasting out an Atlhon 5200 AM2 only because I want an AM3 Phenom X4 with DDR3.
The X2 5200 with DDR2 800 still have a lot of horsepower.
Better coolers. factory coolers suck.
Some kind of hyperthreading, recognizable with software, so it can enable/disable it at convenience.
Enable/disable hyperthreading in ATI drivers profiles, adjusted for each game.
OF course, backward-compatible Havoc drivers. Not like useless PhysX, which don't work with older games.
... and in the fantasy wish list:
internal working with DDRX technologies. Why internal data on processors should transport only one bit by clock, when DDR memory transport many bits in each clock?
it needs multistage transistors, capable of switching between more than two states with low latency.
such "polinary" transistors could store more bits in the same cache size, and multiplicate number of CPUs in the same space.
[citation][nom]apache_lives[/nom]I would target SOC (system on a chip) designs where Intel is looking - AMD would have a similar but more powerful platform thanks to ATi, combining a ~256k K7 core @ 45nm (previously 130nm etc) with chipset etc - enabling smaller devices etcPerhaps even buy VIA out at the right price to jump right into that Netbook market, and stop Nvidia in its tracks Also ATi needs more console deals - there always a winner, and any issues go back to the console maker not the hardware manufacturers etc.Hmmmmmmm logic chipsets for iPhone and similar devices?[/citation]
Intel Atom without Nvidia video is the stupidest thing Intel had does.
Bite!, AMD
Intel Atom without Nvidia video is the stupidest thing Intel had does.
Bite!, AMD
ram is dirt cheap these days. I think amd should get out of making video cards and just tell oem's to put multiple gpu sockets on the motherboard(instead of pci-e) along with a faster hypertransport standard. This would lower the cost of pcs along with more value to the consumer.
QEFX
Distinguished
- Jun 12, 2007
- 258
- 0
- 18,790
Boost performance on moblie CPUs & GPUs.
Inexpensive dual socket MBs would be nice as well.
If possible a desktop chipset with a 4850 class GPU would make inexpensive SFF systems quick and easy to build.
And finally a System on Chip (or at least on Package) low end CPU/GPU/Audio/Chipset would be nice.
Inexpensive dual socket MBs would be nice as well.
If possible a desktop chipset with a 4850 class GPU would make inexpensive SFF systems quick and easy to build.
And finally a System on Chip (or at least on Package) low end CPU/GPU/Audio/Chipset would be nice.
one more thing regarding my
ptroen 04/01/2009 4:38 AM post. AMD could also standardize the socket like the AM2 for both GPU and CPU. Then the new user would simply get X amount of sockets on the mobo. Mix and match how many gpus and cpu the mobo should have and have the bios bootup with the details. PC's in general would be alot smaller since the video cards take up alot of space. This would also allow space for PPU which AMD should consider getting into.
ptroen 04/01/2009 4:38 AM post. AMD could also standardize the socket like the AM2 for both GPU and CPU. Then the new user would simply get X amount of sockets on the mobo. Mix and match how many gpus and cpu the mobo should have and have the bios bootup with the details. PC's in general would be alot smaller since the video cards take up alot of space. This would also allow space for PPU which AMD should consider getting into.
sonofliberty08
Distinguished
- Mar 17, 2009
- 658
- 0
- 18,980
well ...... but the Intel are followed in AMD's shadows since the Presscott core failed , Intel just start to making low power chips , chips with intergrated memory controler , Intel not allow to overclock thier chips lasttime , the overclocking function on Intel desktop board are allways missing lasttime , and the i7 chips are as big as the k6 chips
Advertising. Advertising. Advertising.
It was so sad when Athlon had the speed crown, and nobody but the geeky-est geeks knew. Brand is a huge reason Intel is where they are at.
It didn't have to be superbowl ads. But well targeted advertisements could have been a huge difference maker over the last 8 years in building brand loyalty.
It was so sad when Athlon had the speed crown, and nobody but the geeky-est geeks knew. Brand is a huge reason Intel is where they are at.
It didn't have to be superbowl ads. But well targeted advertisements could have been a huge difference maker over the last 8 years in building brand loyalty.
Buying ATI was a good move, spinning off their Fabs was catastrophically bad. Real men have fabs, that's a quote from Jerry Sanders, who kept the company alive, and created the successful AMD.
They desperately need to produce something after the K7, which is still what they have with the Phenom. It's a bad design, and pretty much always was, and only was successful when competing against an old design like the P6 (and even then, not always well) or a really bad design like the Prescott. Living off the success of your competitor's failures is a mistake, especially when it's Intel. Sooner or later, they get it right, and they have.
AMD is in this situation because of the failure of Hector Ruiz, and his inability to significantly change the K7. Tweaks are not enough, and they have to create something almost from scratch. Things like a powerful x87 unit are almost completely useless, and consume power and size, and add cost for no benefit. Supporting 3D Now! is useless, and wastes transistors (not many, but some). These are just examples of waste, and obviously there are serious design issues with the K7 that they have not addressed. The fact that the Phenom is roughly the same size as the i7, yet so much slower is a first for AMD. It's the first time they had such a significantly slower chip the same size as the Intel chip, and more than that, is significantly slower than the previous generation Intel chip. The current design, 10 years later, being so close to the K7 was catastrophic for AMD, and thank goodness for that. Anything else would have pointed to a lack of progress in the industry.
The AMD/ATI synergy is a good thing, and naturally it has some bumps. But, how else could you possibly recommend an AMD solution were it not for the 790GX? The processors aren't very competitive except when they don't matter, but the IGP is significantly better than the Intel IGP, thus giving AMD a significant advantage to offset Intel's in processors, in certain situations.
I doubt AMD has the resources, since they can't seem to develop a decent x86 processor, but I'd love to see them develop a new RISC chip, without the x86 instruction set impediment. I'd like to see them focus almost entirely on the server space with it, since getting a new instruction set for the desktop would be nearly impossible. A partnership with Dell would make it perfect (IBM wouldn't work because it would compete with POWER, HP wouldn't work since they designed Itanium, and Sun is being shopped around so has little relevance). Since decoupled processors are slower (more stages, greater misprediction penalty), use more power (all the extra transistors needed to decode variable length x86 instructions to RISC type instructions), generate more heat, and are larger and thus more expensive (except economy of scale helps x86 a lot), a new design could be appealing in the server space. The big "plus" for x86 is greater code density, and thus greater cache effectiveness, but this is untrue for the K7 since it pads the L1 cache to deal with the variable x86 instructions. If they can do it, they should. But, I doubt they will, they have been very conservative since Jerry Sanders left, much to their detriment.
They desperately need to produce something after the K7, which is still what they have with the Phenom. It's a bad design, and pretty much always was, and only was successful when competing against an old design like the P6 (and even then, not always well) or a really bad design like the Prescott. Living off the success of your competitor's failures is a mistake, especially when it's Intel. Sooner or later, they get it right, and they have.
AMD is in this situation because of the failure of Hector Ruiz, and his inability to significantly change the K7. Tweaks are not enough, and they have to create something almost from scratch. Things like a powerful x87 unit are almost completely useless, and consume power and size, and add cost for no benefit. Supporting 3D Now! is useless, and wastes transistors (not many, but some). These are just examples of waste, and obviously there are serious design issues with the K7 that they have not addressed. The fact that the Phenom is roughly the same size as the i7, yet so much slower is a first for AMD. It's the first time they had such a significantly slower chip the same size as the Intel chip, and more than that, is significantly slower than the previous generation Intel chip. The current design, 10 years later, being so close to the K7 was catastrophic for AMD, and thank goodness for that. Anything else would have pointed to a lack of progress in the industry.
The AMD/ATI synergy is a good thing, and naturally it has some bumps. But, how else could you possibly recommend an AMD solution were it not for the 790GX? The processors aren't very competitive except when they don't matter, but the IGP is significantly better than the Intel IGP, thus giving AMD a significant advantage to offset Intel's in processors, in certain situations.
I doubt AMD has the resources, since they can't seem to develop a decent x86 processor, but I'd love to see them develop a new RISC chip, without the x86 instruction set impediment. I'd like to see them focus almost entirely on the server space with it, since getting a new instruction set for the desktop would be nearly impossible. A partnership with Dell would make it perfect (IBM wouldn't work because it would compete with POWER, HP wouldn't work since they designed Itanium, and Sun is being shopped around so has little relevance). Since decoupled processors are slower (more stages, greater misprediction penalty), use more power (all the extra transistors needed to decode variable length x86 instructions to RISC type instructions), generate more heat, and are larger and thus more expensive (except economy of scale helps x86 a lot), a new design could be appealing in the server space. The big "plus" for x86 is greater code density, and thus greater cache effectiveness, but this is untrue for the K7 since it pads the L1 cache to deal with the variable x86 instructions. If they can do it, they should. But, I doubt they will, they have been very conservative since Jerry Sanders left, much to their detriment.
my_name_is_earl
Distinguished
- Feb 22, 2009
- 110
- 0
- 18,680
- Status
TRENDING THREADS
-
-
News Microsoft updates Windows 11 24H2 requirements, CPU must support SSE4.2 or the OS will not boot- Started by Admin
- Replies: 14
-
-
-
Question New Build GPU help- ASRock Steel Legend OC Radeon RX 7900 GRE - Keeps turning off
- Started by bp240
- Replies: 6
-
Question How can I make my AMD GPU scale *all* resolutions to 1080p, even resolutions that are "supported" by my TV?
- Started by Sol33t303
- Replies: 7
Latest posts
-
-
-
Question New Build, Fans Running, RGB working, no mouse, no keyboard, no display- Latest: drivinfast247
-
-
-
Facebook
Twitter
Instagram