Sometimes, you need to break the mold—shatter barriers and do something that hasn't been done before. But truly innovative products are really hard to make. If it were easy, everyone would be doing it, right? The more established a market becomes, the more difficult it is to innovate, but that doesn't mean companies can't go back to the drawing board and start fresh. For AMD, that's effectively what they're doing with the upcoming Zen processors.
To be clear, AMD is not starting from scratch. AMD has decades of experience building x86 processors, and everything they've done and learned in the past influences future designs. But when AMD launched the Bulldozer line of processors back in 2011, it was their first major architecture overhaul since the K8 came out in 2003—K10 being largely derived from K8 when it released in 2007. Six years later, AMD has overhauled the architecture in a major way.
That's good news, because if you're using a midrange ($1000) or higher PC to play games, you're likely using an Intel Core i5/i7 processor of some form. The CPU scene has been pretty stagnant for several years now, with Intel's regular tick-tock cadence (now called process-architecture-optimization) giving incremental improvements each generation. AMD has tried to keep pace, but Bulldozer started with a performance deficit, and if you want the fastest and most efficient CPU, Intel wins, period.
AMD had the unenviable task of being the only x86 alternative to Intel, but their manufacturing process fell well behind, with GlobalFoundries being split off from AMD in 2009. Relegated to primarily competing on price, AMD has done quite well at retail markets where people just want an inexpensive PC (think Walmart). But once you get out of the budget sector and into the more lucrative midrange (and business) PC space, Intel dominates sales. And with good reason: take any Core i5 processor and it will generally outperform any AMD CPU of the same generation in the majority of applications and workloads.
This is bad for innovation, bad for the industry, and bad for both AMD and Intel users. We need competition, and it's not too surprising that some of Intel's biggest advancements in processor technology (the Core architecture) came after AMD took the performance crown from them in the Athlon 64 era. The good news is that Zen should shake things up again, with AMD taking another shot at the performance CPU market. Even if AMD can't claim the outright performance crown, having a real performance alternative to Core i5 will be great—especially if it still comes at a lower price.
Designed to scale from mobile products up through desktops, workstations, and servers, Zen looks promising, but will it be enough? Modern processors are complex beasts, with many opportunities for things to go wrong. Here's what we know about Zen, what we expect it will do, and our meditative musings on what we actually want it to do.
What we know: Zen's Architecture
We've had variations of AMD's Bulldozer (Piledriver, Steamroller, and Excavator) for five years, all built using the same building block of a CMT (clustered multi-threading) module that consists of two integer cores with a shared floating-point unit. (It's technically two 128-bit FMAC units that can also work as a single 256-bit FP unit.) All of that changes with Zen.
From a high level, Zen looks a lot like Intel's Core architecture. Gone is the CMT module and in its place AMD appears be moving toward a 4-core/8-thread SMT (symmetric multi-threading) building block. AMD will likely scale down to a 2-core/4-thread module as well, but all indications are that the CPU variants of Zen will launch with 4C/8T and 8C/16T options. AMD has also demonstrated a Naples server chip with 32-cores/64-threads.
Current rumors are that the 4C/8T parts will be sold under the SR5 brand, with 8C/16T selling as SR7, and a lower-tier SR3 to follow. Those may be code names or they may end up being retail names (similar to Intel's i3/i5/i7 nomenclature), but for now we'll stick with calling the 8C/16T part SR7 and the 4C/8T part SR5.
Along with SMT support, the pipeline and various other elements of the architecture have also been reworked. The L1 cache is a faster write-back design, and L2 cache is also up to twice the bandwidth. L3 cache meanwhile will deliver up to five times the bandwidth. There's a new micro-op cache, and each core can issue up to six micro-ops (or four fp-ops) per cycle—similar to Skylake's 6-wide issue width and 50 percent higher than the 4-wide design of the Bulldozer 'heavy equipment' family of CPUs.
Zen has an improved 'perceptron' branch prediction algorithm, now decoupled from the fetch stage, which again helps performance. We don't actually know the pipeline length for Zen (Bulldozer is estimated at a 20-stage pipeline), but better branch prediction can help mitigate having more stages. Notice for example that Intel's NetBurst pipeline was nominally a 20-stage design, which was 'too long' back in the day, and yet all of Intel's recent designs going back at least to Sandy Bridge are around the same length. And not to downplay these aspects, but Zen also features larger load, store, and retire buffers, along with improved clock gating.
Then there's the platform. Zen will use a new AM4 socket, with one of several chipsets, A320, B350, and X370. Regardless of chipset, the platform will stick with a dual-channel DDR4 setup, and the CPU socket has 1331 pins. That's a good number because it's more than Intel's LGA1151, and gives sufficient pins for the rumored 36 PCIe Gen3 lanes on the CPU—that would be 32 lanes for graphics cards, with another four lanes likely used to connect with the chipset. However, slides indicate X370 will be required for SLI/CF setups, so we could end up with more PCIe lanes linked to the chipset, which would in turn connect to the PCIe slots. Sticking with dual-channel makes sense as well, as it keeps motherboard costs in check, and it allows for up to 64GB max memory.
I can't emphasize enough how big of a fundamental change all of this represents, and it means everything we know about AMD's CPU performance from the past may no longer apply. AMD has stated a performance target of 40 percent better IPC (instructions per clock) with Zen versus Bulldozer, and these architecture changes should provide some excellent per-clock performance improvements. A 3.0GHz Zen core should be 40 percent faster than a 3.0GHz Bulldozer core, based on AMD's claims. But there's a catch: we don't know the final clock speeds for Zen. I'll get back to that in a moment, but first let's talk about process technology.
14nm FinFET manufacturing process
In the Bulldozer and earlier timeframe, AMD was similar to Intel in that they did all of the processor design, chip fabrication, and manufacturing in-house for their CPUs. Running a chip foundry—and keeping it up to date—is an expensive proposition, however, and if the facility isn't fully utilized it can be a huge money sink. AMD sold off their fabrication facilities in 2009 and GlobalFoundries (GF) was born, but the separation of the two companies took quite a few years before it was fully realized.
Today, GlobalFoundries is taking orders from other major companies beyond AMD, and AMD has fully divested the last of their GF stock (back in 2012). That means AMD is no longer beholden to GF and can pursue manufacturing agreements with other facilities, focusing on chip design rather than the foundry business. GF is likewise free to take orders for all of their available manufacturing capacity, and GF has also licensed Samsung's 14n FinFET production process has plans to move to 7nm FinFET as their next major process node.
AMD is using GF's 14nm FinFET node for Zen, and moving to a competitive process is a huge jump from Vishera's 32nm SOI process. (Vishera is the codename for the FX-8300/6300/4300 family.) Recent AMD APUs have been using 28nm SOI, so again this will be a big change in feature size. It's similar to the jump graphics chips saw this past year, which ultimately helped double the efficiency of GPUs.
What we expect: Performance
The combination of a new higher performance architecture with an up-to-date manufacturing process basically means anything can happen when it comes to the final product. Is the GF 14nm FinFET fully ready, or are yields still a bit iffy? We don't really know. Does the Zen architecture live up to the hype? Again, we don't know for sure, though early indications are at least promising. And what will the final clock speeds be on retail parts? We've seen reports of engineering samples clocked at up to 3.2GHz floating around, but early ES chips don't tell us much about the final clocks.
Let's start at AMD's IPC claim of being 40 percent higher than Excavator. If maximum clock speed drops from Vishera's 4.3GHz to 3.6GHz, we'll end up with parts that are faster but not amazingly so. On the other hand, we might see 4.4GHz Zen chips that easily outperform AMD's previous offerings and should manage to put some fear into the heart of Intel.
I've run numbers on current AMD and Intel APUs/CPUs (including gaming performance), so let's put things into perspective. A single Excavator core (A10-7890K) runs at up to 4.3GHz and nets 97 points in the single-threaded Cinebench 15 test, while a Piledriver core (FX-8370) running at up to 4.3GHz nets 99 in the same test. A Haswell/Devil's Canyon Core i7-4790K (4.4GHz) gets 173 and a Skylake i7-6700K (4.2GHz) gets 182 in the same test, while the Broadwell-E i7-6900K (3.7GHz) gets 153. Assuming AMD can get Zen to a similar 4.3GHz turbo clock, we're looking at a single-threaded score of around 140. Hmm. That's better, but still well short of Intel's performance, and that's assuming the Zen chips actually ship at 4.3GHz.
Estimating multi-threaded performance is more difficult, but it's a really important factor. Comparing i7-6700K to i5-6600K at the same clock speed, SMT (Hyper-Threading) and a slightly larger L3 cache improves the Core i7-6700K performance by about 30 percent over the i5-6600K. If AMD gets a similar improvement from SMT with Zen, an SR5 part should give Intel's Core i5 parts something to chew on, and an SR7 part—even running at a conservative clock speed—should beat the i7-6700K. But that's only for pure number-crunching performance.
What about games, which don't usually scale super well with CPU core counts? Games are a wild card for Zen, because there are times where AMD's current CPUs are a noticeable bottleneck, at least with a fast GPU like a GTX 1080 running at 1080p. Zen could see larger than 40 percent gains in performance over Vishera in games, and without testing it's impossible to say, but anyone gaming on a 1440p or 4K display should find that a quad-core Zen will likely close any remaining gap between the FX-8370 and the i5-6600K.
And there's still the multi-core factor. DX12 in theory allows games to scale performance with more CPU cores, but so far we've seen little indication of scaling beyond SR5. Ashes of the Singularity is really the only game that pushes beyond a 4C/8T chip, making use of 6-core and even 8-core processors. Other DX12 games haven't hit the CPU nearly as hard, and unless they're using tons of units like AotS, I don't expect this to change that much. A straight quad-core chip will likely prove a bit of a handicap going forward, but quad-core plus SMT isn't close to hitting the end of the road.
We don't have to rely solely on guesstimates, though. WCCFTech (take with a grain of salt) found some leaked performance for the SR7 Zen in Ashes of the Singularity. In the CPU test, Zen beat out an i5-4670K (stock I think), 58 fps to 52.6 fps. That's a win over what will shortly be a three-generation-old Core i5 part, though the newer generation Core i5 CPUs aren't significantly faster, at least not at stock. More perplexing is the Core i7-4790K result, which is still faster than Zen at 65.4 fps. Given Ashes actually does scale beyond four CPU cores, this should be the best-case result, and at least on early samples AMD is well below where I want them to be.
There was also a leaked GeekBench result, which PC Perspective dissected. AMD's extrapolated performance was generally worse than an Ivy Bridge core, but the setup was a bit odd as it used Linux with a 32-core Naples Zen ES chip running at 1.44GHz. That suggests really early silicon (proof that AMD is up and running), so I wouldn't put much stock in those numbers.
[Update: April Fool's jokes are supposed to be pulled down or clearly labeled when it's over. Now let me go wipe that egg off my face....] And another potential leak (dating clear back to April) shows a few select AIDA64 results at Bits and Chips. The problem is having only three of the results doesn't tell us a lot, as AIDA64 tends to be very low-level, so there's not much data there.
What we want from Zen
This is the first time in nearly a decade that a new AMD chip has a chance to go toe-to-toe with Intel's latest. I don't actually expect them to win against Core i7 parts, at least not in overall performance across an array of testing, but I would love to be proven wrong here. AMD is 'still tuning' Zen performance, which in the past has often meant that performance and efficiency weren't quite where they needed to be. Let's hope that's not the case, and that the early leaks have been using sub-optimal silicon and firmware revisions. (As a side note, when I first tested Intel's i7-6700K and Z170, my motherboard BIOS wasn't properly tuned. An updated BIOS a couple of weeks after the launch gave me a 10-15 percent boost in performance, so it's possible AMD could still surprise us—we should know more soon enough.)
Assuming everything else goes as planned, clock speeds will determine a lot of how Zen ranks against Intel. AMD showed an 8-core Zen chip taking on a Broadwell-E i7-6900K, with both running at identical 3.0GHz clocks, and Zen came out just a hair faster in a Blender benchmark. There are many legitimate concerns with that specific test, like whether it was heavily optimized for Zen's architecture—meaning we can't just take that result and unilaterally apply it to all benchmarks. The i7-6900K can also hit 4.5GHz with good cooling, and while AMD was using an engineering sample, we don't know how much higher the chips will actually clock. Looking at the clock speed of RX 480, at least so far the Samsung/GF 14nm FinFET process doesn't appear to clock quite as high as other processes (TSMC 16nm FinFET and Intel 14nm FinFET), but that could be due to the various chip designs rather than the manufacturing process.
I don't see SR7 processors as being 'mainstream' products (just like Intel's X99 platform isn't 'mainstream'), so efficiency isn't my main concern there, but current rumors (based on the engineering samples) are that AMD's 8-core parts will be 95W TDP, with the 4-core part at 65W. That looks good if it's accurate, but the ES chips are also clocked at 3.2GHz and below, which is not so good. What I want to see is the SR5 Zen coming out of the gates with 4.0GHz clocks at a minimum, and I want to be able to slap on a good liquid cooling solution and overclock at least another 10-15 percent. I'd also really love to see AMD's SR7 part ship at clocks that match Intel's i7-6900K, so 3.2GHz base and 3.7GHz turbo, which seems possible if they bump the TDP beyond 95W (note that the i7-6900K is a 130W TDP).
Further out, while taking on the i7-6900K would be great news for Zen, even better would be a quad-core APU with performance rivaling the i5-6600K, paired with a GPU equal to the RX 460. I'd still want discrete graphics, but having a compelling alternative to Intel would be great. Okay, to be honest I really want to see an APU with performance closer to the RX 480, but I doubt that's happening anytime soon, since the 480 has up to 8GB of GDDR5 providing 256GB/s of bandwidth. Even the RX 460 has 112GB/s bandwidth, which a Zen APU can't hope to match with a shared 38.4GB/s memory setup (DDR4-2400). On that note, let's hope the official DDR4-2400 support is merely the baseline and that, like Intel's X99 and Z170 platforms, we see overclocked memory support for stuff like DDR4-3200 (51.2GB/s). Higher performance memory really doesn't matter much for a typical desktop system, but for an APU it can help graphics performance a lot.
Finally, there's the almighty question of the price. This is where AMD has shown their ability to beat Intel for years, because Intel basically refuses to sell higher performance processors for anything less than $175—that's been the approximate entry price for the least expensive quad-core i5 chips going back to the first generation Core i5 chips. With Zen, AMD will have a few options, depending on where performance falls.
If AMD can clearly match or beat Intel's Core i5 Skylake parts with their SR5 Zen chips, we will likely see prices in the $150-$200 range—and if they're competitive with Core i7 Skylake parts AMD could even go for $200-$250. SR7 meanwhile probably isn't going to flirt with Intel's obscene $1000-$1750 prices at the top, and depending on clocks and performance the current guess is $300-$500. Assuming AMD can come close to (or beat) Intel's Broadwell-E performance, we could end up with better prices for everyone, and that would be great to see.
There's a lot of 'ifs' and other qualifiers right now. What will Zen actually deliver? AMD appears more confident in their CPU department than we've seen in a long time, which is good to see. We'll find out if Zen can actually top Intel's parts or merely close the gap in the coming weeks. But if you want the short summary of what I want from Zen: I want it to live up to the hype.
Tidak ada komentar:
Posting Komentar