Hi, great video, and thanks for calling the 5x86 fastest skt 3 cpu! That legal trouble with intel you mentioned, took up almost their entire budget to win against intel. This left cyrix with few funds for R&D going forwards. Since their 486 maths co-processor was reasonable, they stuck with it. Thus, significantly reducing dev cost of the 6x86. The 5x86 you mentioned was actually significantly faster than an intel 486 at same clock rate. But it came with most of the optional performance enhancements OFF by default, as many motherboards didn't support these features well. In a fully compatible m/b though I found the 5x86 was 20% faster than intel at same clock rate with optimisations off. And about 33% faster with them ALL on. So my 5x86-120Mhz performed like a 160Mhz 486, or a P90. I loved that chip so much I bought 3 of them, still got one of them now for a keep sake!
Hi! Congratulations for the video! I remember the P1 and P1-MMX (P54C and P55C) had 16KB and 32KB of L1 cache, respectively; half was data and the other instructions. Recently, I also played around with the M2-300 in Win98 and WinXP and was quite pleasantly surprised by some of the results. Also, I have overclocked it and the P54C and P55C that I have.
Hi, thankyou! yes I probabaly didn't mention the cache is 2x in the pentiums, it is also in the mII though the instruction cache is tiny, and the main cache can switch between ooeratiins I think . I was thinking about trying the mII in xp as someone mentioned it should meet the minimum spec, should be a fun project for another vid sometime😁
@@66mhzbrain Indeed! Anyways, if you want to see the M2 in WinXP with a Radeon SDR or something like that, one of my retro videos would show you what to expect. I need to find the time to put the Voodoo3 2000 PCI on that M2 and give it a go. With a nice motherboard and enough RAM, that M2 would suprise you (pleasantly), especially if you clock it 2.5x100 MHz. A small clock bump for silicone, a giant leap for 3Dmark.
As I stated on your initial community post about this PC, I looked forward to this, and what you produced was well above my expectations! I've also always wondered what effect a major graphics card upgrade would have on a budget machine of the era like this, so thanks for this as it's finally answered that wonderment. I'm surprised Packard Bell went with the 75MHz FSB version of the MII 300 - I always thought they used the 66MHz FSB (with 3.5x multiplier) version in the Club 30/30b/40/40b. I sometimes wonder what effect that has on performance and stability. Not sure if you take requests, but another video I'd love to see in future on this machine would be to return it to the integrated graphics, upgrading the RAM to at least 128MB, and testing out Windows XP (no service packs initially) on it. 300MHz was stated as the "RECOMMENDED" processor speed for Win XP, which of course this processor claims to mimic the performance of, however it (more or less) runs at the "MINIMUM" speed of 233MHz (or in this case 225MHz since it's the 75MHz FSB version, but anyways) which adds a bit of a twist. The Club 30/30b/40/40b were essentially amongst the lowest spec, budget oriented, Windows 98 machines on sale after the launch of Windows 98, and I've always wondered how 'useable' such a machine would actually have been after 3 to 4 years with a RAM upgrade and XP installed - obviously not as a gaming machine of course.
Thankyou for watching it! There was a version of this MII that runs on a 66 mhz bus so not sure, I think phils conputer labs took a look at the same cpu and his was the 66mhz version. I'm not sure if there was a reason for cyrix miving to 75mhz as this chip runs slightly slower than the 66mhz version. I guess pb just went with their spec. Yes that is a cool idea, never really thought about min specs for XP. It will be fun to see how it does! I'll add that one to the list find some sensible maybe more office related benchmarks and do it some time in the coming months.
@@66mhzbrain Thanks for commenting back! It appears Packard Bell chopped and changed a bit between these two versions of the MII 300 - I've seen ppsts on forums and articles where their versions had the 66MHx FSB version. I guess they just went with whatever was best available perhaps at the time for the cheapest bulk purchase? If you are able to do the "XP test" in the future that would be amazing.
Your numbers seem about right to me. At the time (mid 1998) I recall the Socket 7 Intel processors started to be fire-saled by local retailers, as the AMD K6-II was released. Based on the results it looks like the MII might be more close to an Intel P166MMX in performance... An interesting comparison might be the cheapest of the cheap processors at that particular time (summer 1998). An early K6-II, a Cyrix MII and the overclocking champ of the time, the Mendocino Celeron at 300 mhz.
Cool thankyou, I wasn't too sure. Yes I got a cyrix 6x86 machine a while back and have been slowly collecting socket 7 cpus to do a bigger compare, winchips and the like so will see how this fits into the big picture at some point.
Hey, did you install the drivers for that? I have a similar motherboard with sis chipset. It’s a IN530 motherboard I own. I noticed that frame rates were terrible when I didn’t load any of the drivers and better when I did. Especially the one of them that has to do with agp even though it doesn’t have agp but it includes a few other things that it needs. Sis is a weird chipset but I thought I let you know. 530 with 5595
Cool thankyou, everything seems to be installed in device manager so I kind of assumed that the packard bell install had put on everything it needed. I will check though!
Ah, and I maybe missed this info in a video (as I was watching it as a background vid while having a cup of coffee) but had you checked/enabled extra CPU features that were built in but possibly disabled in Cyrix CPUs of that era as at some cases these features were causing some software compatibility issues? If not - you're leaving some performance off the table. I belive that in recent Bits und Bolts video there was an extensive mention of what to tune and where to read/look/get the tools required. Vogons that is. Older Phil's videos might be also of a great help here. And CPU Galaxy IIRC.
Cool, thanks for the info! Someone else mentioned this. I was happy to test it out of the box but will definately take a look to see if it can be tweaked😁
Does the M2 need a tweaker program to get the most out of it. Also, AMD K6-2 / K6-3 were the absolute kings of SS7. 75 / 83MHz were kind of halfway to SS7 - with some motherboards doing an overclocked 75 and maybe 83, without reaching the full 100MHz of a true SS7 board
I usually went for the underdog CPUs as well, always hoping that there would be some truth to the "equivalent to Intel XYZ" claims. It was usually marketing BS... The biggest letdown in my opinion was definitely the "Transmeta Crusoe" CPU.
The problem with "alternative" CPUs was that the PR ratings were based on "office" performance. I've always been an AMD supporter (80486, K5, K6, K6-II and so on), always happy with their performance. However, I'm an "office user"/developer and I don't care much about gaming performance, and this is why, for instance, K6-II 500 MHz felt really close to Pentium II 500 MHz to me. But I can remember gamers regretting buying their Cyrix or AMD CPUs, as they were no match to Intels in many 3D games.
@@peachgrush I think that's the crux of it, somewhat cynically, no one was making it clear that the pr rating related to office software at the time. Only those that read mags/reviews etc. knew and apart from maybe when we get to super socket 7 most buyers of these kind of systems werent in the know but woukd expect the oc to do everything (maybe).
The M2 was my first chip .. Came with the Presario 2262 I used to have.. The machine came with the crappy SiS 5597/5598 VGA.. The fond memories I have of this crappy computer beats the best I have on my current beast rig lol
I always looked at these things (and still do) differently to the way presented here. I set a budget for a chip, and see what the various options can do at that budget. And the Cyrix options at the time were hugely cheaper than the Intel products. ISTR that the Cyrix 300s in 1998-ish were selling for similar money to P133s. Looking at them that way, you were getting a chip that performed on-par with the Intel at a similar price for gaming, but absolutely destroyed it for everyday work stuff. And so Intel never got a sniff. Now if I was after gaming specific performance then the Cyrix was a poor option - there the AMD chips were a better balance of price vs performance. The only time Intel won was if price was no object and you were a hardcore PC gamer.
That makes sense and probably how I would have thought back in the day. But🤔 its a great idea for a video approach. Given an amount of cash cash in 1996 what could you get for a cpu upgrade. Watch this space.
"Confusing and stupid" PR thingy AMD and Cyrix were using back in the day actually wes a thing of necessity. Intel were happy with the fact that the average Joe knows nothing about IPC and only looks at MHz amount as a measure of performance, literally thinking the bigger the number the better. Thus it wasn't an apples to apples comparison between Pentiums and K5s/K6s/6x86 and companies at disadvantage were trying to fight it in a whatever way possible.
Debatable I would say, the fact that people always want things with a bigger number as a result of marketing isnt a good thing imho, it perpetuates ignorance. It was clearly also untrue in some cases as with this mII pr300.
@@lexatwo I'm probably nieve but think people should be straight up. In some ways knowing that an amd 166 can beat an intel 200 is a much better marketing tool than a pr number. Imagine an ad campaign, ours is slower but still beats theirs 😁.
@@66mhzbrain Yep, for me it was always the same. Still we have to admit that we're the minority here being techsavy nerds knowing all in-and-outs. I've spent 10+ years being a self-employed "business owner" providing IT consulting services to small companies and rich-er people. It was always an impossible task to educate. Stereotypes like "Intel is always better" and "MHz are all the same across vendors and are the only factor that matters" can't be eliminated. Thus I get it why they were trying to employ all these Performance Rating shenanigans. Trouble is that it fell of deaf ears of average Joe and at the same time wasn't welcome by knowledgeable minority, especially when PRs were inflated next up to being a straight lie. (PR300 turd, e.t.c.)
Haha yeah gaming is just as bad today as it was back then Though if you just needed to get online or disappoint the kids with a homework machine, Cyrix and AMD ran Windows better for less money than Intel
Haha yes. I remember when me and a friend had pcs from the late 80s and gamed and later had 3dfx cards etc. Then all of a sudden everyone had a pc and people I know from what I remember all had machines like this, k5s, k6s, cyrix, ati all in wonders etc.
![[UA] NAVI vs MOUZ | BO5 | IEM Rio 2024](http://i.ytimg.com/vi/CMBYk0cbGoA/mqdefault.jpg)
Hi, great video, and thanks for calling the 5x86 fastest skt 3 cpu! That legal trouble with intel you mentioned, took up almost their entire budget to win against intel. This left cyrix with few funds for R&D going forwards. Since their 486 maths co-processor was reasonable, they stuck with it. Thus, significantly reducing dev cost of the 6x86. The 5x86 you mentioned was actually significantly faster than an intel 486 at same clock rate. But it came with most of the optional performance enhancements OFF by default, as many motherboards didn't support these features well. In a fully compatible m/b though I found the 5x86 was 20% faster than intel at same clock rate with optimisations off. And about 33% faster with them ALL on. So my 5x86-120Mhz performed like a 160Mhz 486, or a P90. I loved that chip so much I bought 3 of them, still got one of them now for a keep sake!
Thankyou for checking it out! Need to do something on 5x86
I was just searching for something to watch & you uploaded this, many thanks
Cool😁, thankyou for taking a look!
Hi! Congratulations for the video! I remember the P1 and P1-MMX (P54C and P55C) had 16KB and 32KB of L1 cache, respectively; half was data and the other instructions.
Recently, I also played around with the M2-300 in Win98 and WinXP and was quite pleasantly surprised by some of the results. Also, I have overclocked it and the P54C and P55C that I have.
Hi, thankyou! yes I probabaly didn't mention the cache is 2x in the pentiums, it is also in the mII though the instruction cache is tiny, and the main cache can switch between ooeratiins I think . I was thinking about trying the mII in xp as someone mentioned it should meet the minimum spec, should be a fun project for another vid sometime😁
@@66mhzbrain Indeed! Anyways, if you want to see the M2 in WinXP with a Radeon SDR or something like that, one of my retro videos would show you what to expect. I need to find the time to put the Voodoo3 2000 PCI on that M2 and give it a go. With a nice motherboard and enough RAM, that M2 would suprise you (pleasantly), especially if you clock it 2.5x100 MHz. A small clock bump for silicone, a giant leap for 3Dmark.
Cool, I will check it out!
thank you very much for this video!
Thankyou for watching it!
Interesting benchmarks. Like!
Thankyou😁
As I stated on your initial community post about this PC, I looked forward to this, and what you produced was well above my expectations! I've also always wondered what effect a major graphics card upgrade would have on a budget machine of the era like this, so thanks for this as it's finally answered that wonderment.
I'm surprised Packard Bell went with the 75MHz FSB version of the MII 300 - I always thought they used the 66MHz FSB (with 3.5x multiplier) version in the Club 30/30b/40/40b. I sometimes wonder what effect that has on performance and stability.
Not sure if you take requests, but another video I'd love to see in future on this machine would be to return it to the integrated graphics, upgrading the RAM to at least 128MB, and testing out Windows XP (no service packs initially) on it. 300MHz was stated as the "RECOMMENDED" processor speed for Win XP, which of course this processor claims to mimic the performance of, however it (more or less) runs at the "MINIMUM" speed of 233MHz (or in this case 225MHz since it's the 75MHz FSB version, but anyways) which adds a bit of a twist.
The Club 30/30b/40/40b were essentially amongst the lowest spec, budget oriented, Windows 98 machines on sale after the launch of Windows 98, and I've always wondered how 'useable' such a machine would actually have been after 3 to 4 years with a RAM upgrade and XP installed - obviously not as a gaming machine of course.
Thankyou for watching it! There was a version of this MII that runs on a 66 mhz bus so not sure, I think phils conputer labs took a look at the same cpu and his was the 66mhz version. I'm not sure if there was a reason for cyrix miving to 75mhz as this chip runs slightly slower than the 66mhz version. I guess pb just went with their spec. Yes that is a cool idea, never really thought about min specs for XP. It will be fun to see how it does! I'll add that one to the list find some sensible maybe more office related benchmarks and do it some time in the coming months.
@@66mhzbrain Thanks for commenting back! It appears Packard Bell chopped and changed a bit between these two versions of the MII 300 - I've seen ppsts on forums and articles where their versions had the 66MHx FSB version. I guess they just went with whatever was best available perhaps at the time for the cheapest bulk purchase? If you are able to do the "XP test" in the future that would be amazing.
@orion1983uk will definately do the xp thing. Ideas for videos are always welcome!
Your numbers seem about right to me. At the time (mid 1998) I recall the Socket 7 Intel processors started to be fire-saled by local retailers, as the AMD K6-II was released. Based on the results it looks like the MII might be more close to an Intel P166MMX in performance... An interesting comparison might be the cheapest of the cheap processors at that particular time (summer 1998). An early K6-II, a Cyrix MII and the overclocking champ of the time, the Mendocino Celeron at 300 mhz.
Cool thankyou, I wasn't too sure. Yes I got a cyrix 6x86 machine a while back and have been slowly collecting socket 7 cpus to do a bigger compare, winchips and the like so will see how this fits into the big picture at some point.
Hey, did you install the drivers for that? I have a similar motherboard with sis chipset. It’s a IN530 motherboard I own. I noticed that frame rates were terrible when I didn’t load any of the drivers and better when I did. Especially the one of them that has to do with agp even though it doesn’t have agp but it includes a few other things that it needs. Sis is a weird chipset but I thought I let you know. 530 with 5595
Cool thankyou, everything seems to be installed in device manager so I kind of assumed that the packard bell install had put on everything it needed. I will check though!
@@66mhzbrain your welcome. Which sis chipset is it?
@@mOddEdLiKeHeLL 5598
Ah, and I maybe missed this info in a video (as I was watching it as a background vid while having a cup of coffee) but had you checked/enabled extra CPU features that were built in but possibly disabled in Cyrix CPUs of that era as at some cases these features were causing some software compatibility issues? If not - you're leaving some performance off the table. I belive that in recent Bits und Bolts video there was an extensive mention of what to tune and where to read/look/get the tools required. Vogons that is. Older Phil's videos might be also of a great help here. And CPU Galaxy IIRC.
Cool, thanks for the info! Someone else mentioned this. I was happy to test it out of the box but will definately take a look to see if it can be tweaked😁
Does the M2 need a tweaker program to get the most out of it. Also, AMD K6-2 / K6-3 were the absolute kings of SS7.
75 / 83MHz were kind of halfway to SS7 - with some motherboards doing an overclocked 75 and maybe 83, without reaching the full 100MHz of a true SS7 board
Not that I'm aware of. Cyrix isn't king of anything on socket 7.
I usually went for the underdog CPUs as well, always hoping that there would be some truth to the "equivalent to Intel XYZ" claims. It was usually marketing BS... The biggest letdown in my opinion was definitely the "Transmeta Crusoe" CPU.
🤔 Never even heard of it before, just looked it up!
The problem with "alternative" CPUs was that the PR ratings were based on "office" performance. I've always been an AMD supporter (80486, K5, K6, K6-II and so on), always happy with their performance. However, I'm an "office user"/developer and I don't care much about gaming performance, and this is why, for instance, K6-II 500 MHz felt really close to Pentium II 500 MHz to me. But I can remember gamers regretting buying their Cyrix or AMD CPUs, as they were no match to Intels in many 3D games.
@@peachgrush I think that's the crux of it, somewhat cynically, no one was making it clear that the pr rating related to office software at the time. Only those that read mags/reviews etc. knew and apart from maybe when we get to super socket 7 most buyers of these kind of systems werent in the know but woukd expect the oc to do everything (maybe).
@@66mhzbrainI have a Compaq TC1000 Laptop with that CPU. It was a very clever piece of kit with the optional docking station, but terribly slow.
@JonathanMcCormack sounds nice, a laptop with one would be cool
Pentium had TWO TIMES 8kB L1 cache, Pentium MMX TWO TIMES 16kB cache.
The M2 was my first chip .. Came with the Presario 2262 I used to have.. The machine came with the crappy SiS 5597/5598 VGA.. The fond memories I have of this crappy computer beats the best I have on my current beast rig lol
Haha, same here, my memories of some fo the rubbish machines I had back then arebthe best, why I do this now!
Court case against AMD was stopped because you couldn't copyright a number, hence Intel using the Pentium name.
any luck you might have used the cyrix disabled features ?
Disabled features?
check a video made by cpu galaxy about its features ;)@@66mhzbrain
I always looked at these things (and still do) differently to the way presented here. I set a budget for a chip, and see what the various options can do at that budget.
And the Cyrix options at the time were hugely cheaper than the Intel products. ISTR that the Cyrix 300s in 1998-ish were selling for similar money to P133s.
Looking at them that way, you were getting a chip that performed on-par with the Intel at a similar price for gaming, but absolutely destroyed it for everyday work stuff. And so Intel never got a sniff.
Now if I was after gaming specific performance then the Cyrix was a poor option - there the AMD chips were a better balance of price vs performance. The only time Intel won was if price was no object and you were a hardcore PC gamer.
That makes sense and probably how I would have thought back in the day. But🤔 its a great idea for a video approach. Given an amount of cash cash in 1996 what could you get for a cpu upgrade. Watch this space.
"Confusing and stupid" PR thingy AMD and Cyrix were using back in the day actually wes a thing of necessity. Intel were happy with the fact that the average Joe knows nothing about IPC and only looks at MHz amount as a measure of performance, literally thinking the bigger the number the better. Thus it wasn't an apples to apples comparison between Pentiums and K5s/K6s/6x86 and companies at disadvantage were trying to fight it in a whatever way possible.
Debatable I would say, the fact that people always want things with a bigger number as a result of marketing isnt a good thing imho, it perpetuates ignorance. It was clearly also untrue in some cases as with this mII pr300.
@@66mhzbrain oh yeah, marketing keeps being marketing no matter what company it is. Usual bulls*t you know :-/
@@lexatwo I'm probably nieve but think people should be straight up. In some ways knowing that an amd 166 can beat an intel 200 is a much better marketing tool than a pr number. Imagine an ad campaign, ours is slower but still beats theirs 😁.
@@66mhzbrain Yep, for me it was always the same. Still we have to admit that we're the minority here being techsavy nerds knowing all in-and-outs. I've spent 10+ years being a self-employed "business owner" providing IT consulting services to small companies and rich-er people. It was always an impossible task to educate. Stereotypes like "Intel is always better" and "MHz are all the same across vendors and are the only factor that matters" can't be eliminated. Thus I get it why they were trying to employ all these Performance Rating shenanigans. Trouble is that it fell of deaf ears of average Joe and at the same time wasn't welcome by knowledgeable minority, especially when PRs were inflated next up to being a straight lie. (PR300 turd, e.t.c.)
Like
Haha yeah gaming is just as bad today as it was back then
Though if you just needed to get online or disappoint the kids with a homework machine, Cyrix and AMD ran Windows better for less money than Intel
Haha yes. I remember when me and a friend had pcs from the late 80s and gamed and later had 3dfx cards etc. Then all of a sudden everyone had a pc and people I know from what I remember all had machines like this, k5s, k6s, cyrix, ati all in wonders etc.
just please talk louder. or edit the video so it's louder. i have my laptop on 100% and yt on 100% and it's still too quiet. thanks