by Ilya Gavrichenkov
01/07/2009 | 09:16 PM
AMD products haven’t been able to attract the computer enthusiasts’ attention for a while now. Intel with their extremely successful Core microarchitecture pushed AMD into the back seat. Sadly, AMD processors have lost their strengths over the past few years. As a result, the year 2008 end didn’t look pretty for AMD: all their processors were slower, more power-hungry and not very attractive for overclocking community compared against Intel’s offerings. The consequences followed immediately: many computer users turned their back to formerly beloved AMD CPUs. And of course, it couldn’t help affecting the company’s market share that kept getting smaller over the past several quarters. In fact, it is only the aggressive price policy that has been keeping AMD afloat lately: Athlon and Phenom trademarks have long become symbols of affordability rather than high performance.
Unfortunately, Stars (K10) microarchitecture AMD introduced over a year ago failed to remedy the situation. Even though CPUs based on it used quad-core design and boasted a number of useful improvements, their consumer qualities were still not competitive enough. However, the microarchitecture was not the cause for this unsuccessful outcome; from the theoretical positions Stars (K10) microarchitecture is practically as good as Core. AMD’s stumbling stone was 65nm production process: the company fell far behind the competitor in adopting the new manufacturing technology. Therefore, all CPUs on Stars (K10) microarchitecture have so far been manufactured with the unsuccessful 65nm technology. The Athlon processor family is living proof of how fatal the issues of this manufacturing process were in AMD’s implementation: 65nm Brisbane processors couldn’t outperform 90nm Windsor CPUs in terms of supported clock frequencies. As a result, Phenom X4 faced the same challenges. The notorious production process wouldn’t let them raise the clock speeds to acceptable limits, so AMD’s flagship quad-core solutions quickly rolled down to the value market segment.
Therefore we were looking forward to AMD’s transition to new 45nm process, which was expected to become an ultimate remedy against many problems. And finally our patient waiting has been rewarded. A year later than Intel, AMD finally moves desktop CPU production to 45nm process. According to AMD, this milestone opens a new era for the company that is why the new processors using the latest technology will have a new name – Phenom II. The company fans and supporters expect to see Phenom II become a revolutionary breakthrough. But let’s not get ahead of time here and check out these promising new solutions inside out before making any conclusions of our own.
AMD Company uses Phenom II name for their new desktop processors formerly known as Deneb. And although the “II” part following the well-established trademark name may be regarded as an implication of a new microarchitecture spin, the actual differences between Phenom and Phenom II are not that dramatic at all. Even AMD themselves do not deny that Phenom II use the same Stars (K10) microarchitecture as the previous generation Phenom processors.
Therefore, the main distinguishing feature of the new Phenom II CPUs is their transition to new 45nm production process and the use of immersion lithography. Unfortunately, AMD are an entire year behind their main competitor with the production process upgrade. Nevertheless, they use a number of unique techniques, because following in the footsteps of their primary technological partner – IBM – they started using immersion lithography equipment. The specifics of this equipment imply that it uses liquid instead of gas between the projection lenses of the lithography system and the semiconductor wafer. This enhances the system projecting precision and improves the characteristics of the ready semiconductor dies increasing the production yields at the same time. This is a truly progressive approach, because immersion lithography equipment is not just an interesting improvement of contemporary production technologies, but is also a possible alternative to EUV-lithography for future production technologies. This way, 45nm manufacturing process in AMD’s implementation may resolve a number of production and engineering issues all at the same time.
The use of new technologies for processor dies manufacturing affects the frequency potential in the first place. And AMD is in desperate need of higher frequency potential, because their 65nm Phenom processors could only go as high as 2.6GHz. It is low clock frequencies that do not let Phenom X4 processors to successfully compete against Intel Core 2 Quad CPUs. Luckily, the new production process proves up to all expectations. The new Phenom II X4 processors manufactured today use 45nm cores and can go as high as 3GHz clock frequency.
The second important advantage contemporary semiconductor technologies have to offer is the ability to fit more transistors onto the semiconductor die of the same size. And the more transistors there are, the better can certain processor features be improved. Namely, new Phenom II X4 CPUs use a 258sq.mm die and contain 758 million transistors, while previous generation Phenom X4 manufactured with 65nm process had only 450 million transistors on a 285sq.mm die. They managed to increase the number of transistors in the new Phenom II X4 die thanks to microarchitectural improvements aimed at increasing the performance independent of the clock speed.
Overall, AMD promises a 20% performance boost that the new generation CPUs should be able to deliver compared with the previous generation solutions.
Of course, higher clock frequencies of the new Phenom II X4 ensure most of the acceleration. The models announced today seem to be continuing the Phenom X4 lineup, their frequencies are 200MHz and 400MHz higher than those of Phenom X4 9950 – the top product in the 65nm family. However, a 15% frequency increase cannot possibly provide a 20% performance boost.
There are also other factors in the picture. Most added transistors serve to increase the L3 cache memory capacity. The L3 cache of the new Phenom II X4 processors tripled: 6MB vs. 2MB by the previous generation Phenom X4 CPUs. By the way, new processors L3 cache is not only larger; it is also faster than that of the predecessors. Its latency is 2 clocks lower, although that pushed the associativity 1.5 times higher. Phenom II X4 L3 cache has 48-way set associativity, while Phenom X4 has only 32-way set.
As a result, the performance of the L3 cache memory changed diversely, at least according to the Everest Ultimate 4.60 test.
Deneb (45nm) 3.0GHz; 1.8GHz on-die North Bridge
Agena (65nm) 3.0GHz; 1.8GHz on-die North Bridge
I have to add that unfortunately, transition to new production process didn’t let AMD increase the frequency of the on-die North Bridge and it still works at 1.8GHz in the new Phenom II X4 processors. It is really sad, because had this frequency been increased, the L3 cache and memory subsystem performance would have improved much more than after a slight reduction of the L3 cache latency.
These seem to be all the evident advantages of the new Phenom II X4. However, there have also been made a few minor enhancements that also affected the performance. The new 45nm CPUs have improved branch prediction algorithms: now AMD processors can predict indirect branching, just like their competitors. Moreover, the engineers managed to increase the capacity of internal data load and store buffers, as well as FPU buffers. Among other improvements we should definitely mention accelerated floating-point register-to-register instruction moving, LOCK pipelining, accelerated cache coherency during inter-core data transfer.
You can see the general effect from these minor improvements in the Sisoftware Sandra 2009 test results employing simple algorithms not affected by the cache-memory speed and capacity.
The demonstrated gain between 0.5 and 4% is the fruit of these particular “minor improvements”. In other words, the changes AMD engineers made to the computational cores can be described as insignificant. And it means that Phenom II X4 processors will yield seriously to quad-core Intel CPUs working at the same frequencies.
The illustration above shows that they also promise to boost the performance by using DDR3-1333 SDRAM. However, it is not the case for Phenom II X4 processors launching today. Today’s CPUs are designed for Socket AM2+ mainboards, which means they can only work with DDR2 memory, just like their predecessors. Nevertheless, AMD is going to expand the Phenom II lineup in February already by adding Socket AM3 CPUs to the existing models. These processors will be compatible with the existing Socket AM2+ mainboards, as well as with the upcoming platforms with the new CPU socket. Socket AM3 mainboards will be designed to support DDR3 SDRAM, which is also expected to additionally speed up systems with AMD processors.
Although improving performance was one of the most important goals for AMD engineers, the transition to new manufacturing technology helped resolve a few other issues that were typical of the previous generation Phenom CPUs. Here I am talking about heat dissipation and power consumption that have never been the strong sides of Phenom X4 processors built with 65nm process. And even though Phenom II X4 models launched today have the maximum thermal envelope of 125W, just like their predecessors, this number will be revised shortly. The Phenom II X4 models designed for Socket AM3 form-factor that are scheduled to come out in February will have 95W TDP, the same as Intel Core 2 Quad processors.
However, we don’t have to wait until February in order to see the progress made in lowering the power consumption levels. The Socket AM2+ Phenom II X4 CPUs launching today consume about 40% less power in idle mode. AMD has once again revised Cool’n’Quiet technology having provided the version 3.0 implemented in Deneb processors with a new cache flush on halt feature: core's L1 and L2 flush into shared L3 after a core enters a halt state allowing the core to drop to a lower speed and save power. This way you can cut the power to idling cores completely in case there is no workload to deal with.
It is fairly easy to check the efficiency of Cool’n’Quiet 3.0 technology in practical tests. We measured the power consumption in the processor voltage regulator circuitry and thus calculated how much power a CPU needs in case different number of processor cores is actually utilized. Note that the provided numbers do not take into account the efficiency coefficient of the voltage regulator, i.e. they reflect the power consumption of the CPU and VRM together. Despite that, these numbers can suit perfectly for our comparison.
New Phenom II X4 CPUs have really become more economical than their predecessors. You can see it not only in idle mode, when the power consumption dropped 60%, but also during different CPU utilization. Nevertheless, new Phenom II X4 processors are still less power-efficient than 45nm competitor solutions.
Summing up the brief discussion of the Phenom II X4 peculiarities, let’s compare them side by side with the specs of the previous generation quad-core processors from AMD:
Phenom II X4
2.2 – 2.6 GHz
2.8 – 3.0 GHz
4 x [64 KB + 64 KB]
4 x [64 KB + 64 KB]
4 x 512 KB
4 x 512 KB
HyperTransport bus frequency
Built-in North Bridge frequency
C1E, Cool'n'Quiet 2.0
C1E, Cool'n'Quiet 3.0
Today AMD announces two new processors from the Phenom II X4 family: CPUs with 920 and 940 model names. They differ from one another by their clock frequencies, as you can clearly see from the technical specification table below:
As you see, AMD not only transferred their quad-core processors manufacturing to a finer 45nm technological process, but also changes the processor rating system one more time. Now AMD CPU model numbers look very similar to the model numbers of Intel Core i7 CPUs. We believe this is not a simple coincidence, although in February already the analogy will be ruined by the launch of 45nm Socket AM3 processors with smaller L3 cache and tree CPU cores.
In about a month Phenom II family will be expanded with another 6 models:
The upcoming transition to a new processor socket will not only allow using more contemporary memory types, but will also help lower the typical TDP. Therefore, top CPUs in the family are being launched today; they have higher operational frequencies and higher power consumption. Moreover, the top model in the lineup - Phenom II X4 940 – belongs to the Black Edition series, i.e. has an unlocked multiplier. The triple-core Phenom II X3 720 will also be from the same series.
In terms of exterior looks, only the marking makes the new 45nm Socket AM2+ processors different from the predecessors. It is the Socket AM3 CPUs that will differ more evidently: they will have 2 pins less to match the new socket form-factor.
Diagnostic utilities recognize the new Deneb processors just fine:
As you can see from the screenshot, our Phenom II X4 940 has C2 core stepping. Note that it has relatively high core voltage of 1.35V.
The maximum Vcore is set at 1.5V and the CPU package temperature in this case is claimed not to exceed 62°C. This means that Phenom II X4 processors can work at higher voltage settings than their predecessors, however, their heat dissipation and power consumption will remain within the predefined limits. This is peculiar of 45nm Deneb cores: they allow much more aggressive voltage increase than their 65nm predecessors thus determining higher frequency potential.
AMD promised that the new Phenom II X4 processors launching today would be fully compatible with the existing Socket AM2+ mainboards. However, you will need to reflash the BIOS to ensure that the new CPUs work correctly. So, if you are going to upgrade your existing system with a new 45nm quad-core AMD solution, make sure you check there is a corresponding BIOS update with Deneb support for your particular mainboard. In fact, new Phenom II X4 processors can also work with older Socket AM2 mainboards with a few allowances, such as lower HyperTransport bus frequency and no individual core voltage control. However, most mainboard makers do not hurry to update the BIOS of their older platforms with the necessary code. So, in reality only a few Asus Socket AM2 solutions currently support the new Phenom II X4 CPUs.
Nevertheless, it doesn’t prevent AMD from using the continuity of platforms concept that makes Phenom II X4 an attractive upgrade choice. However, AMD seems to be forced to use this argument. The promised 20% performance boost is too small to ensure that the new quad-core processors will compete successfully against Intel Core i7. On the other hand, Phenom II X4 may be a good economical choice: these processors are compatible with many inexpensive Socket AM2+ mainboards and still work with very affordable DDR2 SDRAM. So, even though the price of the top Phenom II X4 CPU is close to that of Core i7-920, the complete platform on the new AMD Phenom processor will cost you $250 less.
Therefore, new 45nm CPUs first of all compete against Core 2 Quad, i.e. offer very attractive price-to-performance ratio in the inexpensive quad-core segment. And if the previous-generation Phenom X4 could only be regarded as a value quad-core CPU, the new Phenom II X4 intend to take some of the market away from Intel solutions.
In order to better understand how the new Phenom II X4 processors can cope with the goals set in front of them, we have to get over to the actual tests.
We are going to compare the performance of the new AMD Phenom II X4 processors against quad-core Intel CPUs from the same price range. As a result, the rivals will be the top CPU from the Core 2 Quad Q8000 family and the junior model from the Core 2 Quad Q9000 family. Of course, we couldn’t help including the hot buy of this season – the junior Nehalem based Intel Core i7-920 CPU. We will also compare the performance of the new Phenom II X4 against that of the top processor from the previous generation Phenom X4.
As a result, we ended up with three test platforms configured as follows:
1. AMD Socket AM2+ platform:
2. Intel LGA775 platform:
3. Intel LGA1366 platform:
Note that although we used a Socket AM2+ mainboard with an integrated graphics core, it was disabled during our test session.
As usual, we started our investigation of Phenom II X4 performance from the Futuremark PCMark Vantage testing suite that reveals the average performance during different types of multi-threaded load.
The results showed that AMD didn’t fully stick to their promises. Phenom II X4 940 performance on average improved only by 10% over the previous generation Phenom X4 9950. Nevertheless, even this improvement is enough for the new 45nm Phenom II X4 to compete against the junior and mainstream Intel Core 2 Quad solutions. However, unfortunately, it is still not enough for them to catch up with the junior Core i7 processors. Even though Phenom II X4 works at 333MHz higher clock frequency than Core i7-920, it is significantly slower than the Intel solution under any type of workload. In other words, contemporary Intel processors can boast the best IPC rate - average number of instructions processed per clock.
Phenom II X4 is dramatically behind Core i7 processors in 3DMark tests. However, there is a logical explanation to that: this test very well splits the load into parallel threads and takes advantage of SMT technology implemented in the new Intel processors. If we compare the Phenom II X4 performance against that of the competitor’s LGA775 solutions, we will see that both models perform as fast as similarly priced Core 2 Quad Q9400 and Core 2 Quad Q8300.
The performance improvement demonstrated by the new 45nm CPUs over their 65nm predecessors is much more significant in games. The top Phenom II X4 processor runs 20-25% faster than Phenom X4 9950. However, we remember very well that the previous generation Phenom processors were actually so slow in games that new quad-core AMD solutions can only compete against junior Intel CPUs from the Core 2 Quad Q8000 series.
Nevertheless, we believe we have to remind you that quad-core processors cannot reveal their potential in most contemporary games yet. However, unfortunately, we will see the dual-core 45nm AMD processor no sooner than in June. Therefore, the best choice for a mainstream gaming system will still be Intel Core 2 Duo.
Phenom II X4 demonstrated relatively high results during media content encoding (compared against its predecessors). In fact, they could have become a serious threat to mainstream Intel processors in video encoding tasks, if Intel hadn’t launched their new Core i7 two months before. The latest Intel solutions are way ahead of all their rivals in media encoding applications.
As for the results demonstrated by the new Phenom II X4 processors in final rendering applications, the verdict will be the same as for the video encoding tasks. Namely, new AMD processors are way faster than their predecessors, but not fast enough to compete successfully against Core i7. They caught up with the junior Core 2 Quad models, but that’s about it. So, the picture is again quite common: even if Phenom II X4 940 outperforms Core 2 Quad Q9400, it still loses to Core i7-920. Therefore, since all three processors belong to the same price range, the only argument in favor of the new AMD solution will be high total cost of the LGA1366 platform.
Quad-core AMD processors of both generations cannot offer the same level of performance as their rivals during digital image processing in Adobe graphics editor. However, HD video rendering speed during non-linear video editing turns out quite high. The new Phenom II X4 processors perform very well even compared against the junior Core i7 during final encoding into Windows Media format.
The popular Mathematica kernel has finally become completely multi-core friendly. Therefore, from now on we are going to use a standard benchmark running within a single kernel.
However, the new testing approach didn’t really change the picture. Quad-core AMD processors still lose to Intel solutions, and Phenom II X4 940 only manages to catch up with Intel Core 2 Quad Q8300 CPU.
AMD processors cope even worse with numeric calculations in Excel. Intel Core i7-920 appears almost twice as fast as the new Phenom II X4 940. Even a cheaper Core 2 Quad Q8300 is still ahead of our hero in this test.
Larger L3 cache memory puts the new Phenom II X4 processors on one level with competitor’s quad-core solutions featuring an L2 cache with 6MB total capacity. However, they still fail to catch up with Core i7, which has an even larger 8MB L3 cache and boasts an extremely fast built-in triple-channel memory controller.
The transition to 45nm production process boosted the performance of the new AMD processors by good 20%, however, it is still not enough for them to demonstrate good results in the chess benchmark. Phenom II X4 920, for instance, is even slower than Core 2 Quad Q8300 – one of Intel’s junior quad-core offerings.
As per our readers’ request we have also included the results obtained in a popular distributed computing project called Folding@Home. We used TOC F@H Bench 0.4.6.0 application.
Unfortunately, the performance of the new Phenom II X4 processors wasn’t up to our expectations. The new generation AMD CPUs could only compete against Core 2 Quad Q8000 series processors, but not the junior Core 2 Quad Q9000 models.
To complete our today’s test session we decided to measure the power consumption of the system built on Phenom II X4 940 processor (without the monitor) and compared the results against the power consumption of similar platforms built around competitor processors. These tests differ from the previously performed power consumption measurements because they take into account the power characteristics of the AMD and Intel chipsets and do not depend on the presence or absence of the memory controller integrated into the processor.
During this test session we loaded the CPUs using Prime95 utility. Besides, we activated all power-saving technologies, such as C1E, Cool’n’Quiet and Enhanced Intel SpeedStep, to make sure that the power readings in idle mode were correct. We also enabled turbo-mode for Core i7-920 processor.
The results demonstrated by the new Phenom II X4 940 processor look absolutely stunning against the background of the previous generation Phenom X4 9950. Due to new more advanced manufacturing process, AMD managed to significantly lower the power consumption of its platforms. However, there is still room for improvement. Contemporary systems with quad-core Intel LGA775 CPUs still boast better “performance-per-watt” ratio.
At the same time, surprising as it is, a system with Phenom II X4 940 processor turned overall more energy efficient than the one built with the latest generation Intel Core i7-920. However, this is no big victory, because Core i7 proved to be a higher –performance solution, according to our tests.
Phenom X4 processors could hardly be considered an overclocker choice. They could rarely be overclocked past 3.2GHz frequency using air-cooling. Therefore, overclocking fans preferred quad-core Core 2 Quad CPUs that could reach much higher speeds with less effort in most cases.
Phenom II X4 seems to be much more interesting in this respect. These processors are made with new 45nm process that improved their frequency potential dramatically and theoretically can push back the overclocking limits, too.
To see if this is in fact the case, we tried overclocking Phenom II X4 940 processor sample available in our lab. This CPU belongs to the Black Edition series, so it features an unlocked clock frequency multiplier. As a result, it is much easier to check out its maximum frequency potential. Having increased the core voltage from the nominal 1.35V to 1.55V and using a Scythe Mugen air cooler we got our CPU running stably at 3.8GHz by simply raising the clock frequency multiplier.
So, new Phenom II X4 processors do have very good frequency potential: the top model overclocked by 26% with air cooling only. So, the new AMD processor seems to be quite fit for overclocking experiments. However, we would like to make sure that potential Phenom II X4 owners keep in mind: 3.8GHz frequency will not let a CPU on Stars (K10) microarchitecture work as fast as overclocked Core 2 Quad from the same price range could.
For example, Core 2 Quad Q9400 processors priced close to Phenom II X4 940 can be easily overclocked to the same 3.8GHz. However, our tests showed that Phenom II X4 CPUs working at 3.0GHz nominal speed are averagely slower than Core 2 Quad Q9400 at the nominal 2.66GHz. As a result, if we increase the processors frequency to the same level, Phenom II X4 will still be noticeably slower.
It means that Phenom II X4 may actually appeal to overclockers only as an alternative to less overclockable Intel CPUs. Among them there is the disappearing Core 2 Quad Q6600 on a 65nm Kentsfield core or 45nm quad-core processors with low multipliers, such as Core 2 Quad Q8200 or Q8300. However, it doesn’t make sense to compare their performance during overclocking against that of an overclocked Phenom II X4 940, because AMD’s solution is in fact more expensive.
Therefore, we assume that a less expensive Phenom II X4 920 may actually be more appealing to overclockers. However, we do not have a CPU like that at this time, so we cannot check out its potential yet.
Frankly speaking we were hoping to finish our today’s review in an optimistic way. It is very sad to realize that over the past few years the competition in the CPU market has practically disappeared. Especially since this situation is hardly good for the end-users: they have no other choice but to decide between the CPUs from one single manufacturer who can joggle the prices any way they want. Besides, the absence of healthy competition slows down the overall technological progress: the lack of high-performance AMD processors inevitably slows down the introduction of new technologies and performance increase in the Intel camp. Therefore, we were hoping until the very last moment that new Phenom II X4 CPUs will start the renaissance era for AMD.
And our expectations came partially true. At least we can say with all certainty that new 45nm manufacturing process proved much better than the previous 65nm technology. The old 65nm process didn’t let different AMD processor families increase their clock speeds for several years. By simply transferring their Stars (K10) based processors to new manufacturing technology, AMD managed to immediately increase their clock speeds by 400MHz – up to 3.0GHz. And this is definitely far not the end of it. We expect AMD to increase their clock frequencies even more over the next few months.
The new manufacturing process also allowed AMD engineers introduce a few improvements in their processors on Stars (K10) microarchitecture: increase their L3 cache memory and make a few changes inside the computational cores. This paid back right away. Our tests showed that the performance of top quad-core AMD processors improved by the good 20%. A few other characteristics of the new Phenom II X4 processors have also improved. Their power consumption lowered and their overclocking frequency potential increased.
However, all the changes in the AMD quad-core processor lineup seem significant enough only when compared against the previous generation Phenom X4, and not against their competitors. It took AMD way too long to switch to 45nm manufacturing technology and launch their Phenom II X4. They missed the window of opportunity and the launch of Phenom II X4 doesn't have the desired effect on the market. The new Phenom II X4 doesn’t look too impressive against the background of contemporary Core 2 Quad and especially Core i7 CPUs. The results of our tests show that the top Phenom II X4 processors can only be worthy rivals to the Core 2 Quad CPUs from the “junior” Q8000 series. Unfortunately, Phenom II X4 cannot yet do better than that.
However, let’s not make any final conclusions at this time. Remember that in February we should see new AMD Socket AM3 processors that will support DDR3 SDRAM. Moreover, we hope that advanced manufacturing process will soon let AMD manufacture faster, more power-efficient and more overclockable processors than the current Phenom II X4 940. So far AMD has proven that they still have hidden reserves to improve the consumer qualities of their solutions on Stars (K10) microarchitecture. And we sincerely hope that these reserves haven’t been exhausted just yet.